Oracle® Hierarchical Storage Manager and StorageTek QFS Software samu User Interface GuideRelease 6.0 E48431-03 |
|
 Previous |
 Next |
Oracle® Hierarchical Storage Manager and StorageTek QFS Software samu User Interface GuideRelease 6.0 E48431-03 |
|
Previous |
Next |
This section lists detailed summaries of samu operator commands in alphabetical order.
a (Display Archiver Status Screen)The a command opens the Archiver status screen and displays current archiver activity.
When issued without a parameter, the a command displays archiver status for all file systems. When the command is issued with the name of a file system, the command displays status for the specified file system only.
Ctrl-B (^b)Move back one file system.
Ctrl-F (^f)Move forward one file system.
Ctrl-D (^d)Move down one archive copy.
Ctrl-U (^u)Move up one archive copy.
Ctrl-J (^j)Toggle between binary and decimal units of size.
The following example shows a typical archiver status screen. The archiver is idle:
Archiver Status samuversiontimedatesam-archiverd: Idle sam-arfind: samqfs1 mounted at /export/samqfs1 Files waiting to start 32 schedule 0 archiving 0 Monitoring file system activity.
The next example shows the same file system, but the archiver is actively archiving:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: samqfs1 mounted at /export/samqfs1 Files waiting to start 149 schedule 0 archiving 0 Monitoring file system activity. sam-arcopy: samqfs1.archset1.1.83 li.VOL119 Copying file /export/samqfs1/2014/data/A00876.dat
The screen lists status information for all Oracle HSM file systems by file-system name and mount point.The fields in the detail display are explained below.
sam-archiverdThe sam-archiverd field displays the current status of the Oracle HSM archiver daemon, the software component that copies modified files to backup media:
Idle means that Oracle HSM is waiting for archive-ready files
Archiving means that Oracle HSM is copying files to archival storage.
Waiting for resources means that Oracle HSM is ready to copy files to archival storage, but is waiting for a drive to become available or for a volume to mount.
Various other messages indicate errors.
sam-arfindThe sam-arfind field displays the most recent results reported by the software component that monitors a file system and locates candidates for archiving. The results include:
the name that you assign to the Oracle HSM file system (samqfs1 in the example)
the mount-point directory for the Oracle HSM file system (/export/samqfs1 in the example)
the number of archive-ready Files waiting to start archiving (32 in the example)
the number of archive-ready files currently in the archiving schedule (0 in the example)
the number of archive-ready files that are currently archiving (0 in the example).
sam-arcopyThe sam-arcopy field displays the current activity reported by the software component that copies archive-ready files to archival media. It reports the following:
the name of the file system (samqfs1 in the example)
The files in an archive request file are members of the same archive set and thus share the same archiving characteristics and requirements. The filename consists of the name of the copy directive (the archiving rule that sets the criteria for copying and retaining a file on specified media) and a sequence number set by the archiver.
the name of the archive set (archset1 in the example)
the copy number of the current copy (1 in the example)
a sequence number (83 in the example)
the volume serial number of the media volume that is being used to store the current copy (VOL119 in the example).
the path and file name of the file that is currently being copied (/export/samqfs1/2014/data/A00876.dat in the example).
abr (Enable Application Based Recovery)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the abr command lets Oracle Real Application Clusters (RAC) use the Application Based Recovery feature of Solaris Volume Manager to handle recovery of databases stored in Oracle HSM file systems that use asynchronous I/O with Solaris Volume Manager mirrored volumes.
By default, ABR is enabled.
The following example enables ABR on family-set device 20:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr 14 on /dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:abr20
add (Add Equipment to a Mounted File System)When issued with an Oracle HSM equipment ordinal number, the add command updates the corresponding file system so that it can use disk devices that have been added to the family set definition in the mcf (Oracle HSM Master Configuration File) and set up with the samd config command.
See remove (Move Data and Remove Equipment from a File System).
For example, assume that we need to add an additional disk slice to file system family set qfs1. We start by adding a line to the /etc/opt/SUNWsamfs/mcf file that identifies the slice, /dev/dsk/c5t8d0s3, as Oracle HSM equipment ordinal number 13 (the selected equipment number must be unique within the Oracle HSM system):
root@solaris:~#vi/etc/opt/SUNWsamfs/mcf... # Section 1: File Systems # # Equipment Equipment Equipment Family Device Additional # Identifier Ordinal Type Set State Parameters # ----------------- --------- --------- -------------- ----- ---------- qfs1 10 ma qfs1 on /dev/dsk/c5t8d0s0 11 mm qfs1 on /dev/dsk/c5t8d0s1 12 md qfs1 on/dev/dsk/c5t8d0s3 13mdqfs1on:wqroot@solaris:~#
Then we reconfigure the running Oracle HSM daemon to incorporate the changes in the file:
root@solaris:~# samd config
At this point, the File systems screen shows that the equipment is off and thus not yet available. So we enter the add command with the user-assigned Oracle HSM equipment number of the new slice:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on qfs1 -----2---r- 80% 70% mm 11 on /dev/dsk/c5t8d0s0 md 12 on /dev/dsk/c5t8d0s1 md13 off/dev/dsk/c5t8d0s3:command:add 13
The device state changes to on:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on qfs1 -----2---r- 80% 70% mm 11 on /dev/dsk/c5t8d0s0 md 12 on /dev/dsk/c5t8d0s1 md13 on/dev/dsk/c5t8d0s3
alloc (Enable Storage Allocation on a Device)When issued with an Oracle HSM equipment ordinal number, the alloc command starts storage allocation on the specified data device. See noalloc (Disable Storage Allocation on a Device).
The following example starts storage allocation on disk device 14:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr14 noalloc/dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:alloc 14
The device state changes to on:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr14 on/dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0
aplease (Set Append Lease Expiration Time for a Shared File System)Before appending data to a shared file, a host must request and obtain an append lease that grants it exclusive access to the file. The aplease command defines the number of seconds during which this lease remains valid on the specified device. If the specified time runs out before the append operation is complete, the host must reacquire the lease before proceeding.
The specified number of seconds must fall in the range [15-600]. The default is 30 seconds.
For related information, see rdlease (Set the Read Lease Time for a Shared File System), wrlease (Set the Write Lease Time for a Shared File System), and lease_timeo (Set the Unused Lease Timeout for a Shared File System).
The following example sets the append lease time to 2 minutes (120 seconds) on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:aplease 800 120
aridle (Stop Archiving Gracefully)The aridle command stops archiving at the next logical point (such as the end of the current archive tar file). The command is typically used prior to disruptive maintenance, such as unmounting, growing, or shrinking file systems. Archiving remains inactive until the arrun command is entered.
When entered with no parameters, aridle stops all archiver activity.
When entered with the dk parameter, aridle stops all archiving to disk media.
When entered with the rm parameter, aridle stops all archiving to removable media, such as tape.
When entered with the fs. parameter and a file-system name, aridle stops all archiver activity on the specified file system.
In the first example, aridle stops all archiving:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:aridle
After the aridle command is entered, the archiver daemon is Idle until the arrun command is entered:
Archiver Status samuversiontimedatesam-archiverd:Idlesam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0Waiting for :arrun
In the second example. aridle stops archiving on the sam1 file system only:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:aridle fs.sam1
After the aridle command is entered, the archiver daemon is Waiting... until the arrun command is entered with the file-system directive fs.sam1:
Archiver Status samuversiontimedatesam-archiverd:Waiting for :arrun fs.sam1sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0Waiting for :arrun fs.sam1
In the third example. aridle stops archiving to disk:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:aridle dk
After the aridle command is entered, the archiver daemon is now Waiting... until the arrun command is entered with the dk (disk) directive:
Archiver Status samuversiontimedatesam-archiverd:Waiting for :arrun dksam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0Waiting for :arrun dk
arrerun (Restart the Archiver Non-Disruptively)The arrerun command restarts archiving without disrupting ongoing operations. The archiver daemons are restarted, and all work in progress is recovered. Compare arrerun to arrestart (Restart the Archiver Disruptively).
In the example, arrerun restarts archiving and recovers work in progress:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:arrerun
arrestart (Restart the Archiver Disruptively)The arrestart command stops and restarts the archiver regardless of the state of the archiver. Compare the command with arrerun (Restart the Archiver Non-Disruptively).
Use arrestart with caution, because it stops operations immediately, and any copy operations that fail to complete will need to be repeated. This waste space on archival media.
arrmarchreq (Remove Archive Requests)When issued with a file-system name and either the name of an archive-request file or a wildcard, the arrmarchreq command deletes the specified archive requests.
An archive-request filename consists of the name of the copy directive (the archiving rule that sets the criteria for copying and retaining a file on specified media) and a sequence number set by the archiver.
:arrmarchreqfile-system-name.archive-request-file-name:arrmarchreqfile-system-name.*
In the first example, arrmarchreq deletes archive request file data0.1.83 from the sam1 file system:
Archiver Status samuversiontimedatesam-archiverd: Idle sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 0 archiving 0 Monitoring file system activity.:command:arrmarchreq sam1.data0.1.83
In the second example, arrmarchreq deletes all archive request files from the sam1 file system:
Archiver Status samuversiontimedatesam-archiverd: Idle sam-arfind: sam2 mounted at /sam2 Files waiting to start 0 schedule 0 archiving 0 Monitoring file system activity.:command:arrmarchreq sam2.*
arrun (Start Archiving)The arrun command starts archiving.
When entered with no parameters, arrun starts all archiver activities.
When entered with the dk parameter, arrun starts archiving to disk media.
When entered with the rm parameter, arrun starts archiving to removable media.
When entered with the fs. parameter and a file-system name, arrun starts archiver activity on the specified file system.
In the first example, arrun starts all archiving:
Archiver Status samuversiontimedatesam-archiverd:Idlesam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0Waiting for :arrun:command:arrun
After the arrun command is entered, the archiver daemon resumes Archiving:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename
In the second example. arrun starts archiving on the sam1 file system:
Archiver Status samuversiontimedatesam-archiverd:Waiting for :arrun fs.sam1sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0Waiting for :arrun fs.sam1:command:arrun fs.sam1
After the arrun command is entered, the archiver daemon resumes archiving file system sam1:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename
In the third example. arrun starts archiving to disk media:
Archiver Status samuversiontimedatesam-archiverd:Waiting for :arrun dksam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0:command:arrun dk
After the arrun command is entered, the resumes archiving to disk:
Archiver Status samuversiontimedatesam-archiverd:Archivingsam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename
arscan (Scan a File System for Unarchived Files)When issued with the name of an Oracle HSM file system, the arscan command scans the specified file system for unarchived files.
When entered with no parameters other than the file-system name, arscan recursively scans all directories in the specified file system.
When entered with the file-system name and a directory, arscan recursively scans all directories in the specified file system starting at the specified directory.
When entered with the file-system name and the .inodes key word, arscan scans the .inodes file. If most files in a file system are already archived, this is faster than scanning directories.
Entering an integer representing a number of seconds with any of the other parameters delays the scan for the specified number of seconds.
arstop (Stop All Archiving Immediately)The arstop command stops archiving immediately.
When entered with no parameters, arstop stops all archiver activities.
When entered with the dk parameter, arstop stops archiving to disk media.
When entered with the rm parameter, arstop stops archiving to removable media.
When entered with the fs. parameter and a file-system name, arstop stops archiver activity on the specified file system.
In the first example, arstop stops all archiving:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:arstop
In the second example, the arstop command stops archiving on file system sam1:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:arstop fs.sam1
artrace (Perform Archiver Tracing)The artrace command starts archive tracing.
Tracing writes a detailed log of archiver activity to the sam-archiverd file in the /var/opt/SUNWsamfs/trace/ subdirectory.
When entered with no parameters, artrace traces archiver activities on all file systems.
When entered with the fs. parameter and a file-system name, artrace traces archiver activity on the specified file system.
In the first example, artrace starts archiver tracing for all file systems:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:artrace
In the second example, the artrace command starts archiver tracing for file system sam1:
Archiver Status samuversiontimedatesam-archiverd: Archiving sam-arfind: sam1 mounted at /sam1 Files waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. sam-arcopy: sam1.copy-directive.sequence-numbervol-serCopying filepath/filename:command:artrace fs.sam1
atime (Configure Access Time Updates for a File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device -1, 0, or 1, the atime command controls how and when the access time attribute is updated for files in the corresponding file system.
Setting a 0 (zero) value enables caching and delayed writes up to one minute for access time updates. Access time updates are only written to disk immediately when utilization of an Oracle HSM file system is above the low water mark, when the access time change coincides with updates to the creation time (ctime) or modification time (mtime), or when the file system is unmounted. This setting reduces disk I/O and thus improves performance.
Setting a 1 value immediately updates the access time on disk whenever a file is accessed. This increases disk I/O and reduces performance accordingly.
Setting a -1 value or executing the noatime command disables access time updates. If access times are unimportant, this can significantly reduce I/O and thus improve performance. However, the POSIX standard requires that access times be marked on files. Never use this setting with an archiving, Oracle HSM file system.
The default is value is 0 (caching and delayed writes enabled).
For more information, see the mount_samfs and stat man pages for further information.
The following example disables caching and forces immediate access time updates to disk on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2----d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:atime 100 1
audit (Audit Library or Library Slot)When issued with the Oracle HSM equipment ordinal number of a robotic library and, optionally, a slot identifier, the audit command causes the specified automated library to mount volumes, read the volume serial number (VSN) of each volume, and rebuild the library catalog. If specified, the slot must be in use and physically occupied by media.
If the specified slot contains a tape cartridge, the -e parameter tells the audit to skip to the end of data (EOD) and determine the space available. Note, however, that this process is not interruptible and may take hours to complete.
If the specified slot contains a two-sided optical cartridge, identify the desired side as 1 or 2. If a side is not specified, both sides are audited.
For more information, see the auditslot(1M), mount_samfs and stat man pages for further information.
:auditequipment-number:auditequipment-number:slot:audit -eequipment-number:slot:auditequipment-number:slot:side:audit -eequipment-number:slot:side
The following example audits slot 17 in the robotic library with Oracle HSM equipment ordinal number 900:
Robot VSN catalog by slot : eq 900 samuversiontimedatecount 32 slot access time count use flags ty vsn 0 2013/07/02 12:53 3 96% -il-o-b----f li VOL001 1 2013/07/02 12:53 2 98% -il-o-b----f li VOL002 2 2013/07/02 12:52 1 86% -il-o-b----f li VOL003 3 2013/07/02 12:52 1 95% -il-o-b----f li VOL004 4 2013/07/02 12:51 1 79% -il-o-b----- li VOL005 5 2013/07/02 12:45 0 14% -ilEo-b----f li VOL006 MEDIA ERROR 6 2013/07/02 12:46 0 0% -il-o-b----- li VOL007 7 2013/07/02 12:46 0 0% -il-o-b----- li VOL008 8 2013/07/02 12:47 0 0% -il-o-b----- li VOL009 9 2013/07/02 12:47 0 0% -il-o-b----- li VOL010 10 2013/07/02 12:48 0 0% -il-o-b----- li VOL011 11 2013/07/02 12:48 0 0% -il-o-b----- li VOL012 12 2013/07/02 12:48 0 0% -il-o-b----- li VOL013 13 2013/07/02 12:49 0 0% -il-o-b----- li VOL014 14 2013/07/02 12:49 0 0% -il-o-b----- li VOL015 15 2013/07/02 12:50 0 0% -il-o-b----- li VOL016 16 2013/07/02 12:50 0 0% -il-o-b----- li VOL017 17 2013/07/02 12:51 0 0% -il-o-b----- li VOL018 18 2013/07/02 12:51 0 0% -il-o-b----- li VOL019 19 none 50 0% -il-oCb----- li CLN020:command:audit 900:17
c (Display Device Configuration Screen)The c command opens the Device configuration screen, which lists all device names and equipment ordinal numbers.
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
The following example shows the device configuration screen:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 SL8500 ti 101 down /dev/rmt/0cbn 100 SL8500 ti 102 down /dev/rmt/1cbn 100 SL8500 ti 103 down /dev/rmt/2cbn 100 SL8500 hy 104 on historian 104
tyThe ty field displays the Oracle HSM device type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq field displays the user-assigned Oracle HSM equipment ordinal number of the device.
stateThe state field displays the current operating state of the device, which may include any of the following:
on means that the device is available.
ro means that the device is available read-only.
off means that the device is not available.
down means that the device is available for maintenance only.
idle means that the device is completing operations that are already in progress but is not accepting new connections.
noalloc means that no more storage can be allocated on this device. An administrator has run the noalloc command (see noalloc (Disable Storage Allocation on a Device)). The noalloc state persists until an administrator issues an alloc command (see alloc (Enable Storage Allocation on a Device)).
device_nameThe device_name field displays the path and file name of the device.
In this instance, the family set equipment type, sk, indicates an Oracle StorageTek ACSLS interface to a network-attached library. So the device name is the path to the parameters file for the library, SL8500_T10K_parms. For detailed information on network-attached libraries and parameter files, see the mcf man page.
fsThe fs field displays the user-assigned Oracle HSM equipment ordinal number for the family-set that contains a device.
family_setThe family_set field displays the user-assigned name of the Oracle HSM family set that contains the device. Family sets represent the equipment associated with a Oracle HSM component, such as the group of disk devices associated with an Oracle HSM file system or the tape library associated with archiving.
clear (Clear a Load Request for a Removable Storage Volume)When issued with a volume serial number (VSN) and, optionally, an index value, the clear command cancels pending load requests for the specified VSN, clears any reference to the VSN from the Removable media load requests screen, and aborts any processes that are waiting for the volume to mount.
Optionally, you may specify the value of the index column of the row that corresponds to the VSN on the screen.
See p (Display the Removable Media Load Requests Screen) and load (Load a Cartridge into a Drive) for further information.
d (Display Daemon Trace Controls Screen)The d command opens the Daemon trace controls screen, which displays the events that Oracle HSM is currently tracing, as specified in the defaults.conf file. For more information about enabling trace files, see the defaults.conf man page.
The following example shows trace file information. It includes information about the daemons being traced, the paths to the trace files, the events being traced, and information about the size and age of the trace files.
Daemon trace controls samuversiontimedatesam-amld /var/opt/SUNWsamfs/trace/sam-amld cust err fatal misc proc debug date size 10 age 0 sam-archiverd /var/opt/SUNWsamfs/trace/sam-archiverd cust err fatal misc proc debug date size 10 age 0 sam-catserverd /var/opt/SUNWsamfs/trace/sam-catserverd cust err fatal misc proc debug date size 10 age 0 sam-fsd /var/opt/SUNWsamfs/trace/sam-fsd cust err fatal misc proc debug date size 10 age 0 sam-rftd /var/opt/SUNWsamfs/trace/sam-rftd cust err fatal misc proc debug date size 10 age 0 sam-recycler /var/opt/SUNWsamfs/trace/sam-recycler cust err fatal misc proc debug date size 10 age 0
def_retention (Set Default WORM Retention Time)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a time interval, the def_retention command changes the default retention period for WORM-enabled files that do not have a specified retention period to the specified time interval.
The data and path of a WORM (Write Once Read Many) file cannot be changed or deleted until the retention period expires. Specify the retention period in any of three ways:
To indicate that the file should never be deleted or changed, enter the string permanent or 0.
To specify an interval in years, days, hours, and/or minutes, enter a string of the form
[YEARSy][DAYSd][HOURSh][MINUTESm]
where YEARS is the specified number of years (if any), MONTHS is the specified number of months (if any), DAYS is the specified number of days (if any), HOURS is the specified number of hours (if any), and MINUTES is the specified number of minutes (if any). For example, 5y3d1h4m specifies 5 years, days, 1 hour, and 4 minutes, 10y specifies 10 years, and 60d12h specifies 60 days and 12 hours.
To specify retention as a number of minutes, enter an integer in the range [1-2147483647].
If left unchanged, the default retention period for WORM files is 43,200 minutes (30 days).
For additional information on WORM options, see the mount_samfs and sam_worm man page.
The following example sets the default retention period for WORM files stored in equipment ordinal number 10 (file system samfs1) to 7 years:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on samfs1 m----2----d 90% 70% /samfs1 mm 11 on /dev/dsk/c5t10d0s0 md 12 on /dev/dsk/c5t10d0s1 md 13 on /dev/dsk/c5t11d0s0 md 14 on /dev/dsk/c5t11d0s1:command:def_retention 10 7y
devlog (Set Device Logging Options)When issued with the Oracle HSM equipment ordinal number of a device and a space-delimited list of event-logging options, the devlog command configures logging for the specified device.
Specify event-logging options using one or more of the following key words:
all logs all events.
none disables logging.
default restores the default logging settings.
detail logs events that help track the progress of operations.
err logs error messages.
label logs labeling operations.
mig logs migration toolkit messages.
msg logs thread and process communications.
retry logs retries of device operations.
syserr logs system library errors.
time logs time device operations.
module includes module names and source lines in log messages.
event includes event names in the message in log messages.
date includes event dates in the message.
- keyword (a minus sign preceding one of the keywords above) clears removes the keyword from the current logging specification.
The default logging specification is err retry syserr date. For additional information, see the samset man page.
:devlogequipment-numberoption[next-option]... :devlog alloption[next-option]... :devlogequipment-number-option[-next-option]... :devlog all -option[-next-option]...
The following example logs the event name in addition to the standard options for device 101:
Removable media status: all samuversiontimedatety eq status act use state vsn ti 101 ---------p 0 0% notrdy empty ti 102 ---------p 0 0% notrdy empty:command:devlog 101 event err retry syserr date
dio_rd_consec (Limit the Number of Consecutive Direct Reads)The dio_rd_consec command specifies the number of consecutive reads larger than dio_rd_form_min or dio_rd_ill_min that Oracle HSM should perform using direct I/O.
The default value is 0, which disables direct reads based on I/O size and causes Oracle HSM to ignore dio_rd_form_min or dio_rd_ill_min.
For more information, see dio_rd_form_min (Set the Size of Well-Aligned Direct I/O Reads) and dio_szero (Zero Uninitialized Areas of Sparse Files During Direct I/O).
The following example sets minimum size of well-formed direct reads for family set device 100 to 16:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_rd_consec 100 16
dio_rd_form_min (Set the Size of Well-Aligned Direct I/O Reads)The dio_rd_form_min command makes a specified number of kilobytes the minimum size at which Oracle HSM will automatically switch from paged I/O to direct I/O when reading well-aligned data on a specified family-set disk device (see "Switched I/O").
The default value is 256 kilobytes. Setting the value to 0 disables switched I/O during reads of well-aligned data.
The following example sets minimum size of well-formed direct reads for family set device 100 to 1024 kilobytes:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_rd_form_min 100 1024
dio_rd_ill_min (Set the Size of Misaligned Direct I/O Reads)The dio_rd_ill_min command makes a specified number of kilobytes the minimum size at which Oracle HSM will automatically switch from paged I/O to direct I/O when reading misaligned data on a specified family-set disk device (see "Switched I/O").
The default value is 0 kilobytes which disables switched I/O during reads of misaligned data.
The following example sets minimum size of malformed direct reads for family set device 100 to 4096 kilobytes:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_rd_form_min 100 4096
dio_szero (Zero Uninitialized Areas of Sparse Files During Direct I/O)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the dio_szero command makes direct I/O to the specified device handle sparse files in the same way that paged I/O handles them. Uninitialized areas are zeroed as the areas are accessed. Sparse file behavior the same as that for paged I/O. This reduces performance when writing sparse files directly.
The default is nodio_szero (Do Not Zero Uninitialized Areas of Sparse Files with Direct I/O).
The following example zeroes sparse files on family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_szero 100
dio_wr_consec (Limit the Number of Consecutive Direct Writes)The dio_wr_consec command specifies the number of consecutive writes larger than dio_wr_form_min or dio_wr_ill_min that Oracle HSM should perform using direct I/O.
The default value is 0, which disables direct writes based on I/O size and causes Oracle HSM to ignore dio_wr_form_min or dio_wr_ill_min.
For further details, see dio_wr_form_min (Set the Size of Well-Formed Direct I/O Writes) and dio_wr_ill_min (Set the Size of Misaligned Direct I/O Writes).
The following example sets minimum size of well-formed direct reads for family set device 100 to 16:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_wr_consec 100 16
dio_wr_form_min (Set the Size of Well-Formed Direct I/O Writes)The dio_wr_form_min command makes a specified number of kilobytes the minimum size at which Oracle HSM will automatically switch from paged I/O to direct I/O when writing well-aligned data to a specified family-set disk device (see "Switched I/O").
The default value is 256 kilobytes. Setting the value to 0 disables switched I/O during writes of well-aligned data.
The following example sets minimum size of well-formed direct writes for family set device 100 to 1024 kilobytes:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_wr_form_min 100 1024
dio_wr_ill_min (Set the Size of Misaligned Direct I/O Writes)The dio_wr_ill_min command makes a specified number of kilobytes the minimum size at which Oracle HSM will automatically switch from paged I/O to direct I/O when writing misaligned data to a specified family-set disk device (see "Switched I/O").
The default value is 0 kilobytes which disables switched I/O during reads of misaligned data.
The following example sets minimum size of malformed direct writes for family set device 100 to 4096 kilobytes:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:dio_wr_form_min 100 4096
diskvols (Set/Clear Flags on Disk Volumes Used for Archiving)When issued with the name of an archival disk volume, a plus or a minus, and one of the flags listed below, the diskvols command sets or clears the specified flag on the specified volume. When troubleshooting hardware issues, you may want to change the values of the U and E flags. But leave the remainder alone:
l means that the volume is labeled; a seqnum file has already been created.
r means that the volume is defined on a remote host.
U means that the volume is unavailable.
R means that the volume is read only.
E means that a media error occurred when writing to the disk archive directory.
A means that the volume needs to be audited.
F means that the volume is full.
c means that the volume is ready for recycling.
The following example sets the U (unavailable) flag on the archival storage volume disk01:
Disk volume dictionary samuversion time dateheader version 460 volumes magic 340322 version 9 nkeys 2 ndata 2 index space capacity used flags volume 0 12882411520 12887785472 10291200 ----- disk01 1 6443827200 6443892736 70656 ----- disk02 clients magic 340322 version 9 nkeys 1 ndata 1:command:diskvols disk01 +U
The flag is set:
Disk volume dictionary samuversion time dateheader version 460 volumes magic 340322 version 9 nkeys 2 ndata 2 index space capacity used flags volume 0 12882411520 12887785472 10291200--U--disk01 1 6443827200 6443892736 70656 ----- disk02 clients magic 340322 version 9 nkeys 1 ndata 1
dmr (Enable Directed Mirror Reads of Software Mirrors)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the dmr command lets Oracle Real Application Clusters (RAC) use the Directed Mirror Reads feature of Solaris Volume Manager when databases are stored on Oracle HSM file systems that use asynchronous I/O with Solaris Volume Manager mirrored volumes. See nodmr (Disable Directed Mirror Reads of Software Mirrors).
By default, DMR is enabled.
The following example enables DMR on family-set device 20:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr 14 on /dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:dmr 10
dtrace (Enable/Configure Tracing for Oracle HSM Daemons)When issued with an Oracle HSM daemon specifier plus the keyword on or off or a variable name and value, the dtrace command controls how trace information is written to trace files.
Specify daemons using the keyword all (for all Oracle HSM daemons) or the name of a particular daemon. Valid daemon names include sam-archiverd, sam-catserverd, sam-fsd, sam-rftd, sam-recycler, sam-sharefsd, sam-stagerd, sam-serverd, sam-clientd, and fsmgmt.
Use the on keyword to enable tracing for the specified daemon(s).
Use the off keyword to disable tracing for the specified daemon(s).
Use a variable name with a value to control the tracing process for the specified daemon. Set the variable using the syntax daemon-specifier.variable value.
:dtracedaemon-specifieron :dtracedaemon-specifieroff :dtracedaemon-specifier.filepath-filename:dtracedaemon-specifier.optionsoption_name1option_name2-option_name3 ...
fileThe file variable takes a path or file-name value:
all.file path-filename names the directory that holds the trace files path-filename.
By default, trace files reside in /var/opt/SUNWsamfs/trace.
daemon-name.filepath-filename names the trace file path-filename.
By default, trace files are named for the daemon they trace.
optionsThe options variable takes a space-separated list of trace options, events to be traced or elements to be included in trace output. To deselect an option, use a minus sign with the option (-option_name).
The pre-defined event types are cust, err, fatal, misc, proc, and rft. The message elements program[pid] and time are always included and can't be deselected.
Optional events include the following:
none excludes all event types.
all includes the most useful events: cust, err, fatal, ipc, misc, proc, and rft.
alloc includes memory allocation events.
cust includes customer-notification, syslog and notify file messages.
err includes non-fatal program errors.
fatal includes fatal syslog messages.
files includes file actions.
rft includes file transfer events.
ipc includes inter-process communications.
misc includes miscellaneous, otherwise uncategorized events.
oprmsg includes operator messages.
proc process initiation and completion.
queue archiver queue contents when changed.
Optional message elements include:
date includes the date in the message (the time is always included).
module includes the source file name and line number in the message.
type includes the event type in the message.
ageThe age variable sets the time between trace file rotations to the specified value (rotating trace files keeps their size within manageable limits). Set the value of age to a number of seconds, minutes, hours, days, weeks, and/or years greater than 2 minutes.
A value of two minutes or less disables trace-file rotation.
sizeThe size variable sets the file size at which trace file rotations occur. Set the value of size as an integer with one of the following suffixes specifying the units: b for bytes, k for kilobytes, M for megabytes, G for gigabytes, or T for terabytes.
The following example excludes non-fatal errors from the sam-archiverd trace file and includes the event date and the source file and line number where the error occurred in each trace file entry:
Help information page 1/15 samuversiontimedateDisplays: a Archiver status w Pending stage queue c Device configuration C Memory d Daemon trace controls D Disk volume dictionary f File systems F Optical disk label g Shared clients I Inode h Help information J Preview shared memory l Usage information K Kernel statistics m Mass storage status L Shared memory tables n Staging status M Shared memory o Optical disk status N File system parameters p Removable media load requests P Active Services r Removable media R SAM-Remote s Device status S Sector data t Tape drive status T SCSI sense data u Staging queue U Device table v Robot catalog:command:dtrace sam-archiverd.options -err date module
export (Move Removable Media to the Robotic Library's Mailbox)When issued with the Oracle HSM equipment ordinal number of a robotic library and a slot identifier, the export command moves the cartridge that resides in the specified slot to the mailbox. If specified, the slot must be in use and physically occupied by media. If the slot contains a two-sided optical cartridge and no side is specified, both sides are audited.
When issued with an Oracle HSM media-type identifier and the Volume Serial Number (VSN) of a cartridge, the export command moves the specified cartridge to the mailbox.
If the specified equipment ordinal number identifies a StorageTek network-attached library, the -f parameter exports the cartridge to the library's Cartridge Access Port (CAP) and updates the Oracle HSM catalog. The CAPID parameter must be defined in the library parameters file. See the stk man page for details.
If the specified equipment ordinal number identifies some other network-attached library, the export command updates the Oracle HSM catalog but leaves physical movement of the cartridge to the library's control software.
See also import (Move Removable Media from the Mailbox into the Robotic Library).
Note that, by default, exported volumes are tracked by the Oracle HSM in the historian. The historian acts is a virtual library that can handle staging and archiving requests for volumes that are no longer resident in the library. An operator can then import the required cartridges to satisfy the load requests. Exporting cartridges from the historian itself deletes all information on the exported cartridge. The information about volumes on this cartridge will be lost. See the historian man page for further information.
:exportequipment-number:slot:export -fequipment-number:slot:exportmedia-type.vsn:export -fmedia-type.vsn
In the following example, the first form of the command exports the contents of slot 1 in the robotic library with Oracle HSM equipment ordinal number 900. The second form exports the LTO cartridge with VSN VOL004.
Robot VSN catalog by slot : eq 900 samuversiontimedatecount 32 slot access time count use flags ty vsn 0 2013/07/02 12:53 3 96% -il-o-b----f li VOL001 1 2013/07/02 12:53 2 98% -il-o-b----f li VOL002 2 2013/07/02 12:52 1 86% -il-o-b----f li VOL003 3 2013/07/02 12:52 1 95% -il-o-b----f li VOL004 ...:command:export 900:1... command:export li.VOL004
f (Display File Systems Screen)The f command opens the File systems screen and displays the components of Oracle HSM file systems.
The following example shows the file systems screen. Note that member drives are indented one space and appear directly below the file system to which they belong.
File systems samuversiontimedatety eq state device_name status high low mountpoint server ms 10 on sam1 m----2----d 90% 70% /sam1 md 11 on /dev/dsk/c5t8d0s3 md 12 on /dev/dsk/c5t8d0s4 md 13 on /dev/dsk/c5t8d0s5 md 14 on /dev/dsk/c5t8d0s6 md 15 on /dev/dsk/c5t8d0s7 ms 20 on sam2 m----2----d 90% 70% /sam2 md 21 on /dev/dsk/c5t9d0s3 md 22 on /dev/dsk/c5t9d0s4 md 23 on /dev/dsk/c5t9d0s5 md 24 on /dev/dsk/c5t9d0s6 md 25 on /dev/dsk/c5t9d0s7 ma 30 on qfs1 m----2----d 90% 70% /qfs1 mm 31 on /dev/dsk/c5t10d0s0 md 32 on /dev/dsk/c5t10d0s1 ma 40 on qfs2 m----2----d 90% 70% /qfs2 mm 41 on /dev/dsk/c5t11d0s0 md 42 on /dev/dsk/c5t11d0s1 ma 50 on qfs3 m----2---r- 90% 70% /qfs3 mm 51 on /dev/dsk/c5t12d0s0 mr 52 on /dev/dsk/c5t12d0s1 ma 60 on qfs4 m----2---r- 90% 70% /qfs4 mm 61 on /dev/dsk/c5t13d0s0 mr 62 on /dev/dsk/c5t13d0s1 ma 100 on shareqfs1 m----2c--r- 80% 70% /shareqfs1servermm 101 on /dev/dsk/c6t50020F2300004655d0s0 mr 102 on /dev/dsk/c6t50020F2300004655d0s1 ma 110 on shareqfs2 m----2c--r- 80% 70% /shareqfs2servermm 111 on /dev/dsk/c6t50020F2300004655d0s6 mr 112 on /dev/dsk/c6t50020F2300004655d0s7
tyThe ty field displays the Oracle HSM device type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq field displays the user-assigned Oracle HSM equipment number of the device.
stateThe state field displays the current operating state of the device:
on means that the device is available.
ro means that the device is available read-only.
off means that the device is not available.
down means that the device is available for maintenance only.
idle means that the device is completing operations that are already in progress but is not accepting new connections.
noalloc means that no more storage can be allocated on this device. An administrator has run the noalloc (Disable Storage Allocation on a Device) command. The noalloc state persists until an administrator issues an alloc (Enable Storage Allocation on a Device) command.
device_nameThe device_name field displays the path and file name of the device.
statusThe status field displays an eleven-character string where each character position represents the value of a different file-system attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 11. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | m---------- |
A lowercase m in position 1 means that the file system is mounted. |
| 1 | M---------- |
An uppercase M in position 1 means that the file system is being mounted. |
| 2 | -u--------- |
A lowercase u in position 2 means that the file system is being unmounted. |
| 3 | --A-------- |
An uppercase A in position 3 means that data is being archived (copied to nearline or offline storage). |
| 4 | ---R------- |
An uppercase R in position 4 means that previously archived data is being released from online disk storage to free space for new files. |
| 5 | ----S------ |
An uppercase S in position 5 means that previously released data is being staged (copied from archival media to online disk storage). |
| 6 | -----1----- |
The numeral 1 in position 6 means that this is a version-1 file system. |
| 6 | -----2----- |
The numeral 2 in position 6 means that this is a version-2 file system. |
| 7 | ------c---- |
A lowercase c in position 7 means that this is a shared file system. |
| 8 | -------W--- |
An uppercase W in position 8 means that this is a single-writer file system. |
| 9 | --------R-- |
An uppercase R in position 9 means that this is a multi-reader file system. |
| 10 | ---------r- |
A lowercase r in position 10 means that the file system stores data on single-allocation mr devices. |
| 11 | ----------d |
A lowercase d in position 11 means that the file system stores data on dual-allocation md devices. |
highThe high field displays the high-water mark, a storage utilization threshold expressed as the percentage of the file system's total capacity. When percent utilization exceeds the high-water mark, Oracle HSM starts deleting files that have been archived to backup media, starting with the largest and least recently used files. If users subsequently access the deleted files, Oracle HSM restores them to disk from an archival copy.
lowThe low-water mark, a storage utilization threshold expressed as the percentage of the file system's total capacity. When percent utilization drops below the low-water mark, Oracle HSM stops deleting files from the disk. The low-water mark balances the need to maintain free disk space for new files against file-system performance by keeping files resident on disk and avoiding the need for excessive staging from archival media.
mountpointThe directory where the file system attaches to the host's root file system.
serverThe name of the metadata server (MDS) for a file system that is shared by multiple hosts.
flush_behind (Set the Flush-Behind Parameter)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a number of kilobytes, the flush_behind command configures the corresponding file system to asynchronously write the specified number of kilobytes from dirty (modified) cache memory pages to disk.
As files change in memory, flush-behind insures that the changes are quickly saved to non-volatile storage media. When a process writes modified data to cache memory pages, the flushing follows close behind, copying the changes to disk even as the main write process continues. Unsaved, dirty cache pages are kept to a minimum, which helps to maintain the integrity of the file data.
On the other hand, flush-behind increases overhead and can reduce write performance. The process of flushing a data block may delay subsequent re-writes of that block. So flush-behind is disabled by default.
Set the specified value in the range [0-8192] kilobytes, where 0—the default—disables the flush-behind capability.
The following example sets staging flush-behind for family set device 100 to 1024 kilobytes:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:flush_behind 100 1024
forcedirectio (Use Direct I/O by Default)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the forcedirectio command makes direct I/O the default for all input/output operations on the specified device.
Paged I/O is the default.
For more information, see "Understanding I/O Performance Tuning Objectives and Options", the noforcedirectio (Do Not Use Direct I/O by Default) command, and the directio, setfa, sam_setfa, and sam_advise man pages.
The following example makes direct I/O the default for family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:forcedirectio 100
force_nfs_async (Force Asynchronous NFS)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the force_nfs_async command causes the specified device to override the Oracle HSM default and cache NFS (Network File System) data on the server, even if NFS has requested that the data be written through to disk.
The force_nfs_async command is effective only if the file system is mounted as an NFS server and only if the clients are mounted with the noac NFS mount option. For more information about mounting an NFS file system, see the mount_nfs man page.
|
Caution: Use this command with caution. In the event of a server interruption, data can be lost!The |
The following example sets force_nfs_async on family set 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma100m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:force_nfs_async 100
g (Display Shared File System Clients)The g command opens the Shared clients screen and displays current configuration information.
When issued without a parameter, the a command displays archiver status for all file systems. When the command is issued with the name of a file system, the command displays status for the specified file system only.
Ctrl-I (^I)Toggle between the abbreviated and detailed views.
Ctrl-BMove back one page.
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
ordThe ord field displays the Oracle HSM server ordinal number. The server ordinal distinguishes the currently active metadata server from potential metadata servers and clients:
Ordinal 1 indicates the active metadata server.
Ordinal 2 or higher indicates a potential metadata server.
Ordinal 0 indicates a client.
hostnameThe hostname field displays the name of the server, potential server, or client host.
statusThe status field displays the mount status, host type (server or client), file system type, and distributed I/O (datamover) status. In the abbreviated view, the field is displayed in bit-mask form. In the detailed view, they are expanded as shown in the example.
config and conf1The config and conf1 fields display file system configuration details. In the abbreviated view, the fields are displayed in bit-mask form. In the detailed view, they are expanded as shown in the example.
flagsThe flags field displays abbreviated, textual status information such as mount status (MNT) and host type (SVR or CLI).
The following example shows the default, abbreviated view:
Shared clients samuversiontimedatesamsharefs is shared, server is samsharefs-mds, 2 clients 3 max ord hostname seqno nomsgs status config conf1 flags 1 samsharefs-mds 14 0 8091 808540d 4051 0 MNT SVR 2 samsharefs-client1 71 0 a0a1 808540d 4041 0 MNT CLI
This example shows the detailed view:
Shared clients samuversiontimedatesamsharefs is shared, server is samsharefs-mds, 2 clients 3 max ord hostname seqno nomsgs status config conf1 flags 1 samsharefs-mds 14 0 8091 808540d 4051 0 MNT SVR config : CDEVID ARCHIVE_SCAN GFSID OLD_ARCHIVE_FMT " : SYNC_META TRACE SAM_ENABLED SHARED_MO config1 : NFSV4_ACL MD_DEVICES SMALL_DAUS SHARED_FS flags : status : MOUNTED SERVER SAM DATAMOVER last_msg : Wed Jul 2 10:13:50 2014 2 samsharefs-client1 127 0 a0a1 808540d 4041 0 MNT CLI config : CDEVID ARCHIVE_SCAN GFSID OLD_ARCHIVE_FMT " : SYNC_META TRACE SAM_ENABLED SHARED_MO config1 : NFSV4_ACL MD_DEVICES SHARED_FS flags : status : MOUNTED CLIENT SAM SRVR_BYTEREV " : DATAMOVER last_msg : Wed Jul 2 11:09:04 2014
h (Display the Help Screen)The h command opens the Help information screen. The help screen summarizes samu displays and controls. By default, this is the first display that the system presents when you enter the samu command at the command line.
Ctrl-BMove back one page.
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
There are fifteen Oracle HSM help screens in all, but archive-related displays are not shown if the Oracle Hierarchical Storage Manager software is not installed.
The first example shows the first page that appears by default whenever you launch samu. Subsequent help screens show samu commands.
Help information page 1/15 samuversiontimedateDisplays: a Archiver status w Pending stage queue c Device configuration C Memory d Daemon trace controls D Disk volume dictionary f File systems F Optical disk label h Help information I Inode l Usage information J Preview shared memory m Mass storage status K Kernel statistics n Staging status L Shared memory tables o Optical disk status M Shared memory p Removable media load requests N File system Parameters r Removable media P Active Services s Device status R SAM-Remote t Tape drive status S Sector data u Staging queue T SCSI sense data v Robot catalog U Device table more (ctrl-f)
samu Interface ControlsThe second page of the help shows the hot keys that control the samu interface. Note that some controls are only applicable to certain displays:
Help information page 2/15 samuversiontimedateHot Keys: q Quit : Enter command sp Refresh display ^f Page display forward ^b Page display backward ^d Half-page display forward ^u Half-page display backward ^i Show details (selected displays) ^j Toggle power for size, base 2 or 10 (selected displays) ^k Advance display format ^l Refresh display (clear) ^r Toggle refresh / Search for VSN (v display) % Search for barcode (v display) $ Search for slot (v display) more (ctrl-f)
samu Command PromptThe third page of the help shows display-control commands that can be entered at the same command prompt that control the displays. Note that some controls are only applicable to certain displays:
Help information page 3/15 samuversiontimedateDisplay control commands: a [filesystem] Archiver status n [media] Staging status p [media] Removable media load requests r [media] Removable media u [media] Staging queue v [eq] Robot catalog w [media] Pending stage queue C address Memory I [inode] Inode J [address] Preview shared memory M [address] Shared memory R [eq] SAM-Remote S [address] Sector data T [eq] SCSI sense data U [eq] Device table more (ctrl-f)
To view the archive display for a specified file system, you would enter the command-prompt hot key : (the colon). Then, at the samu command prompt, you would enter a filesystem-equipment-identifier as shown below:
...
T [eq] SCSI sense data
U [eq] Device table
more (ctrl-f)
Command:a filesystem-equipment-identifier
The samu interface then displays a detailed archiving status report for the specified file system:
Archiver status samuversiontimedatesam-archiverd: Waiting for resourcessam-arfind:filesystem-equipment-identifiermounted at /mountpoint-directoryFiles waiting to start 0 schedule 482 archiving 0 Monitoring file system activity. Examine: noscan Interval: 4m Logfile: /var/opt/SUNWsamfs/archiver.log events 12,138 syscalls 41,537 buffers 2 archive 0 change 3 close 7,961 create 3,927 hwm 29 modify 0 rearchive 0 rename 9 remove 209 unarchive 0 idstat 18,101 opendir 28 getdents 28 cached
The fourth page of the help shows various file-system commands that can be entered at the samu command prompt:
Help information page 4/15 samuversiontimedateFile System commands - miscellaneous: stripe eq value Set stripe width suid eq Turn on setuid capability nosuid eq Turn off setuid capability sync_meta eq value Set sync_meta mode atime eq value Set access time (atime) update mode trace eq Turn on file system tracing notrace eq Turn off file system tracing add eq Add eq to mounted file system remove eq Remove eq; copy files to ON eqs release eq Release eq and mark files offline alloc eq Enable allocation on partition noalloc eq Disable allocation on partition def_retention eq interval Set default WORM retention time more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The fifth page of the help shows Oracle HSM file-system commands that can be entered at the samu command prompt:
Help information page 5/15 samuversiontimedateFile System commands - SAM: hwm_archive eq Turn on hwm archiver start nohwm_archive eq Turn off hwm archiver start maxpartial eq value Set maximum partial size in kilobytes partial eq value Set size to remain online in kilobytes partial_stage eq value Set partial stage-ahead point in kilobytes stage_flush_behind eq value Set stage flush behind size in kilobytes stage_n_window eq value Set direct stage size in kilobytes thresh eq high low Set high and low release thresholds more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
hwm_archive (Start Archiving When the High Water Mark is Reached)
nohwm_archive (Do Not Start Archiving at the High Water Mark)
maxpartial (Set the Maximum Partial Release Size for the File System)
maxpartial (Set the Maximum Partial Release Size for the File System)
partial (Set the Default Partial Release Size for the File System)
partial_stage (Set Partial Stage-Ahead Point for the File System)
stage_n_window (Size the Buffer for Reading Directly from Archival Media)
The sixth page of the help summarizes file-system input/output (I/O) commands that can be entered at the samu command prompt:
Help information page 6/15 samuversiontimedateFile System commands - I/O: dio_rd_consec eq value Set number of consecutive dio reads dio_rd_form_min eq value Set size of well-formed dio reads dio_rd_ill_min eq value Set size of ill-formed dio reads dio_wr_consec eq value Set number of consecutive dio writes dio_wr_form_min eq value Set size of well-formed dio writes dio_wr_ill_min eq value Set size of ill-formed dio writes flush_behind eq value Set flush behind value in kilobytes forcedirectio eq Turn on directio mode noforcedirectio eq Turn off directio mode force_nfs_async eq Turn on NFS async noforce_nfs_async eq Turn off NFS async readahead eq value Set maximum readahead in kilobytes writebehind eq value Set maximum writebehind in kilobytes sw_raid eq Turn on software RAID mode nosw_raid eq Turn off software RAID mode wr_throttle eq value Set outstanding write size in kilobytes abr eq Enable Application Based Recovery noabr eq Disable Application Based Recovery dmr eq Enable Directed Mirror Reads20a nodmr eq Disable Directed Mirror Reads
For more information on individual commands, see the full descriptions listed below:
dio_rd_consec (Limit the Number of Consecutive Direct Reads)
dio_rd_form_min (Set the Size of Well-Aligned Direct I/O Reads)
dio_rd_ill_min (Set the Size of Misaligned Direct I/O Reads)
dio_wr_consec (Limit the Number of Consecutive Direct Writes)
dio_wr_form_min (Set the Size of Well-Formed Direct I/O Writes)
dio_wr_ill_min (Set the Size of Misaligned Direct I/O Writes)
nosw_raid (Turn Off Software RAID Mode)wr_throttle (Set the Maximum Size of Outstanding Writes)
The seventh page of the help summarizes QFS file-system commands that can be entered at the samu command prompt:
Help information page 7/15 samuversiontimedateFile System commands - QFS: mm_stripe eq value Set meta stripe width qwrite eq Turn on qwrite mode noqwrite eq Turn off qwrite mode more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The eighth page of the help summarizes multi-reader file-system commands that can be entered at the samu command prompt:
Help information page 8/15 samuversiontimedateFile System commands - multireader: invalid eq value Set multireader invalidate cache delay refresh_at_eof eq Turn on refresh at eof mode norefresh_at_eof eq Turn off refresh at eof mode more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The ninth page of the help summarizes shared file-system commands that can be entered at the samu command prompt:
Help information page 9/15 samuversiontimedateFile System commands - shared fs: minoallocsz eq value Set minimum allocation size maxallocsz eq value Set maximum allocation size meta_timeo eq interval Set shared fs meta cache timeout lease_timeo eq interval Set shared fs lease relinquish timeout min_pool eq value Set shared fs minimum threads count mh_write eq Turn on multihost read/write nomh_write eq Turn off multihost read/write aplease eq interval Set append lease time rdlease eq interval Set read lease time wrlease eq interval Set write lease time more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
minallocsz (Set the Minimum Allocation Size for a Shared File System)
maxallocsz (Set the Maximum Allocation Size for a Shared File System)
meta_timeo (Set the Metadata Cache Timeout for a Shared File System)
lease_timeo (Set the Unused Lease Timeout for a Shared File System)
min_pool (Set the Minimum Number of Shared File System Threads)
aplease (Set Append Lease Expiration Time for a Shared File System)
The tenth page of the help summarizes shared device commands that can be entered at the samu command prompt:
Help information page 10/15 samuversiontimedateDevice commands: devlog eq [option ...] Set device logging options idle eq Idle device off eq Turn off device on eq Turn on device readonly eq Make device read only ro eq Make device read only unavail eq Make device unavailable unload eq Unload device more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The eleventh page of the help summarizes robot commands that can be entered at the samu command prompt:
Help information page 11/15 samuversiontimedateRobot commands: audit [-e] eq[:slot[:side]] Audit slot or library import eq Import cartridge from mailbox export [-f] eq:slot Export cartridge to mailbox export [-f] mt.vsn Export cartridge to mailbox load eq:slot[:side] Load cartridge in drive load [-f] mt.vsn Load cartridge in drive priority pid priority Set priority in preview queue more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The twelfth page of the help summarizes archiver commands that can be entered at the samu command prompt:
Help information page 12/15 samuversiontimedateArchiver commands: aridle [dk | rm | fs.fsname] Idle archiving arrerun Soft restart archiver arrestart Restart archiver arrmarchreq fsname.[* | arname] Remove ArchReq(s) arrun [dk | rm | fs.fsname] Start archiving arscan fsname[.dir | ..inodes][int] Scan filesystem arstop [dk | rm | fs.fsname] Stop archiving artrace [fs.fsname] Trace archiver more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The thirteenth page of the help summarizes stager commands that can be entered at the samu command prompt:
Help information page 13/15 samuversiontimedateStager commands: stclear mt.vsn Clear stage request stidle Idle staging strun Start staging more (ctrl-f)
For more information on individual commands, see the full descriptions listed below:
The fourteenth page of the help summarizes miscellaneous commands that can be entered at the samu command prompt:
Help information page 14/15 samuversiontimedateMiscellaneous commands: clear vsn [index] Clear load request dtrace daemon[.variable] value Daemon trace controls fs [filesystem] Select a filesystem name (ex samfs1) mount mountpoint Select a mount point open eq Open device q Exit from samu refresh [interval] Set display refresh interval in seconds read address Read sector from device snap [filename] Snapshot screen to file diskvols volume +flag | -flag Set or clear disk volume flags !shell-command Run command in a shell more (ctrl-f)
See the full descriptions for more information on the commands listed below:
The fifteenth page of the help lists media type codes recognized by the samu interface:
Help information page 15/15 samuversiontimedateMedia types: all All media types tp tape at Sony AIT tape sa Sony Super AIT tape d2 Ampex DST310 (D2) tape d3 STK SD-3 tape dt 4mm digital tape (DAT) fd Fujitsu M8100 128track tape i7 IBM 3570 tape ib IBM 3590 tape li IBM 3580, Seagate Viper 200 (LTO m2 IBM 3592 tape lt digital linear tape (DLT) se STK 9490 tape sf STK T9940 tape sg STK 9840 tape st STK 3480 tape so Sony DTF tape vt Metrum VHS tape xm Exabyte Mammoth-2 8mm tape xt Exabyte 8mm tape od optical mo 5 1/4 in. erasable optical disk wo 5 1/4 in. WORM optical disk o2 12 in. WORM optical disk
hwm_archive (Start Archiving When the High Water Mark is Reached)When issued with an Oracle HSM family-set equipment ordinal number, the hwm_archive command configures the corresponding file system to start archiving automatically when the total size of the stored files passes the high water mark, the maximum space-utilization threshold specified for the file system. See nohwm_archive (Do Not Start Archiving at the High Water Mark).
The following example configures disk family-set device 20 to automatically start archiving when the corresponding file-system is 80% full:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam/sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:hwm_archive 20
idle (Idle Device)The idle command blocks new connections to the device specified by an Oracle HSM equipment ordinal number. Operations that are already in progress continue.
The example below idles device 103:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/ libraryB_T10K_parms 100 libraryB ti 101 on /dev/rmt/0cbn 100 libraryB ti 102 on /dev/rmt/1cbn 100 libraryB ti103 on/dev/rmt/2cbn 100 libraryB hy 104 on historian 104:command:idle 103
The device state changes to idle:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/ libraryB_T10K_parms 100 libraryB ti 101 on /dev/rmt/0cbn 100 libraryB ti 102 on /dev/rmt/1cbn 100 libraryB ti103 idle/dev/rmt/2cbn 100 libraryB hy 104 on historian 104
import (Move Removable Media from the Mailbox into the Robotic Library)When issued with the Oracle HSM equipment ordinal number of a robotic library, the import command tells the library to move media from the library's mailbox to the first available slot in the library.
See also export (Move Removable Media to the Robotic Library's Mailbox).
The following example below imports the contents of the mailbox in the robotic library with Oracle HSM equipment ordinal number 900:
Robot VSN catalog by slot : eq 900 samuversiontimedatecount 32 slot access time count use flags ty vsn 0 2013/07/02 12:53 3 96% -il-o-b----f li VOL001 1 2013/07/02 12:53 2 98% -il-o-b----f li VOL002 2 2013/07/02 12:52 1 86% -il-o-b----f li VOL003 3 2013/07/02 12:52 1 95% -il-o-b----f li VOL004 4 2013/07/02 12:51 1 79% -il-o-b----- li VOL005 5 2013/07/02 12:45 0 14% -ilEo-b----f li VOL006 MEDIA ERROR ...:command:import 900
invalid (Set the Invalidate-Cache Delay for a Multi-Reader File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device that is mounted read-only in a single-writer/multi-reader file system, the invalid command tells the readers of a single-writer/multi-reader file system to delay invalidating their file-system metadata caches for a specified number of seconds. When files are modified, the host waits the prescribed interval before checking the metadata server for updates.
|
Caution: Setting a non-zero delay value keeps stale file-system metadata cached, so the host may not have a consistent, current view of a read-only file system during the delay. |
The specified number of seconds must fall in the range [0-60]. The default is 0 seconds.
The following example sets the cache invalidation delay to 30 seconds on family set device 900:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 900 on qfs1 m----2--R-d 90% 70% /qfs1 mm 901 on /dev/dsk/c5t10d0s0 md 902 on /dev/dsk/c5t10d0s1 md 903 on /dev/dsk/c5t11d0s0 md 904 on /dev/dsk/c5t11d0s1:command:invalid 900 30
l (Display Utilization Information)The l command opens the Usage information screen. The screen displays storage utilization statistics for the file systems and associated archival storage.
The following example shows a sample usage screen.
Usage information samuversiontimedatehostid =0xIDOS name: SunOS Architecture: SPARC CPUs: 2 (2 online) library 40: capacity 389.3G bytes space 291.1G bytes, usage 25% library 51: capacity 9.5G bytes space 9.5G bytes, usage 0% library 55: capacity 0 bytes space 0 bytes, usage 0% library 56: capacity 10.7G bytes space 10.7G bytes, usage 0% library totals: capacity 409.5G bytes space 311.3G bytes, usage 24% filesystem samfs3: capacity 54.5M bytes space 13.4M bytes, usage 75% filesystem samfs4: capacity 319.5M bytes space 298.0M bytes, usage 7% filesystem samfs7: capacity 96.6M bytes space 69.6M bytes, usage 28% filesystem samfs6: capacity 5.0G bytes space 4.9G bytes, usage 3% filesystem samfs8: capacity 5.0G bytes space 4.9G bytes, usage 2% filesystem totals: capacity 10.5G bytes space 10.2G bytes, usage 3%
lease_timeo (Set the Unused Lease Timeout for a Shared File System)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device and a number, the lease_timeo command configures the way in which the corresponding shared file system manages unused read, write, and append leases.
Specify a number in the range [0-15] to force processes to relinquish leases that remain unused for the specified number of seconds.
Specify the number -1 to disable unused lease timeouts and let processes hold unused leases for the full terms specified by aplease (Set Append Lease Expiration Time for a Shared File System), rdlease (Set the Read Lease Time for a Shared File System), and "wrlease (Set the Write Lease Time for a Shared File System)". The default is 0.
Note that read and write leases are not relinquished if mh_write (Turn on Multi-Host Read/Write) is enabled.
:lease_timeofamily-set-equipment-number0-to-15-seconds:lease_timeofamily-set-equipment-number-1
The following example sets the unused lease timeout for shared file system shareqfs1 on family set device 800 to 5 seconds:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:lease_timeo 800 5
load (Load a Cartridge into a Drive)The load command loads media into the drive specified by either an Oracle HSM equipment ordinal number and slot location or a specified Oracle HSM media type and volume serial number (VSN). Media-type identifiers are listed in Appendix A and on the mcf man page.
The following example loads the LTO (li) with volume serial number VOL001:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 SL8500 li 101 on /dev/rmt/0cbn 100 SL8500 li 102 on /dev/rmt/1cbn 100 SL8500 li 103 on /dev/rmt/2cbn 100 SL8500 hy 104 on historian 104:command:load li.VOL001
m (Display the Mass Storage Status Screen)The m command opens the Mass storage status screen. The screen displays the status of mounted file systems and their member drives.
In the following example, note that member drives are indented one space and appear directly below the file system to which they belong.
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ms 10 m----2----d 1% on 68.354G 68.343G 1M 16 90% 70% md 11 1% on 0 13.669G 13.666G md 12 1% on 1 13.669G 13.667G md 13 1% on 2 13.669G 13.667G md 14 1% on 3 13.674G 13.672G md 15 1% on 4 13.674G 13.672G ms 20 m----2----d 1% on 68.354G 68.344G 1M 16 90% 70% md 21 1% on 0 13.669G 13.667G md 22 1% on 1 13.669G 13.667G md 23 1% on 2 13.669G 13.667G md 24 1% on 3 13.674G 13.672G md 25 1% on 4 13.674G 13.672G ma 30 m----2----d 4% on 64.351G 61.917G 1M 16 90% 70% mm 31 1% on 0 4.003G 3.988G [8363840 inodes] md 32 4% on 1 64.351G 61.917G ma 40 m----2----d 1% on 64.351G 64.333G 1M 16 90% 70% mm 41 1% on 0 4.003G 3.997G [8382784 inodes] md 42 1% on 1 64.351G 64.333G
tyThe ty field displays the Oracle HSM device type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq field displays the user-assigned Oracle HSM equipment ordinal number for the mass storage device.
statusThe status field displays an eleven-character string in which each character position represents the value of a different file-system attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 11. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | m---------- |
A lowercase m in position 1 means that the file system is mounted. |
| 1 | M---------- |
An uppercase M in position 1 means that the file system is being mounted. |
| 2 | -u--------- |
A lowercase u in position 2 means that the file system is being unmounted. |
| 3 | --A-------- |
An uppercase A in position 3 means that data is being archived (copied to nearline or offline storage). |
| 4 | ---R------- |
An uppercase R in position 4 means that previously archived data is being released from online disk storage to free space for new files. |
| 5 | ----S------ |
An uppercase S in position 5 means that previously released data is being staged (copied from archival media to online disk storage). |
| 6 | -----1----- |
The numeral 1 in position 6 means that this is a version-1 file system. |
| 6 | -----2----- |
The numeral 2 in position 6 means that this is a version-2 file system. |
| 7 | ------c---- |
A lowercase c in position 7 means that this is a shared file system. |
| 8 | -------W--- |
An uppercase W in position 8 means that this is a single-writer file system. |
| 9 | --------R-- |
An uppercase R in position 9 means that this is a multi-reader file system. |
| 10 | ---------r- |
A lowercase r in position 10 means that the file system stores data on single-allocation mr devices. |
| 11 | ----------d |
A lowercase d in position 11 means that the file system stores data on dual-allocation md devices. |
useThe use field displays the percentage of disk space in use.
stateThe state field displays the current operating state of the device:
on means that the device is available.
ro means that the device is available read-only.
off means that the device is not available.
down means that the device is available for maintenance only.
idle means that the device is completing operations that are already in progress but is not accepting new connections.
noalloc means that no more storage can be allocated on this device. An administrator has run the noalloc (Disable Storage Allocation on a Device) command. The noalloc state persists until an administrator issues an alloc (Enable Storage Allocation on a Device) command.
ordThe ord field displays the ordinal number of the device within the user-defined Oracle HSM family set (family sets represent the equipment associated with an Oracle HSM component, such as the group of disk devices associated with an Oracle HSM file system or the tape library associated with archiving).
capacityThe capacity field displays the number of 1024-byte blocks of usable space on the disk.
freeThe free field displays the number of 1024-byte blocks of disk space available.
raThe ra field displays the read-ahead size in kilobytes.
partThe part field displays the partial stage size in kilobytes.
highThe high field displays the high water mark for the file system. The high water mark is the percentage disk utilization at which Oracle HSM starts to delete files that have archive copies from disk, so that there is space for new files.
lowThe low displays the low water mark for the file system. The low water mark is the percentage disk utilization at which Oracle HSM has freed enough space for new files and stops deleting archived files.
maxallocsz (Set the Maximum Allocation Size for a Shared File System)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device and a kilobyte value, the maxallocsz command sets maximum number of kilobytes that are allocated ahead of a write to the specified shared file system. See minallocsz (Set the Minimum Allocation Size for a Shared File System).
The specified allocation size must be a multiple of 8 kilobytes in the range [16-4194304]. The default value is 8 times the device allocation unit (DAU). See the discussion of the -a option on the sammkfs man page for additional information on DAUs.
The following example sets the maximum allocation to 1024 kilobytes on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:maxallocsz 800 1024
maxpartial (Set the Maximum Partial Release Size for the File System)When issued with an Oracle HSM family-set equipment ordinal number and a number of kilobytes, the maxpartial command sets the maximum number of bytes from the start of a file that can be retained in disk cache when the file is released.
A large file that has been designated for partial release can open more quickly when restaged from archival media because the beginning of the file is already on disk when staging begins. For further information, see partial (Set the Default Partial Release Size for the File System), partial_stage (Set Partial Stage-Ahead Point for the File System), and the release man page.
The specified kilobyte value must be an integer in the range [8-2097152]. The default is 16 kilobytes.
The following example configures disk family-set device 20 to allow retention of a maximum of 512 kilobytes on disk for each file designated for partial release:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam1 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:maxpartial 20 512
meta_timeo (Set the Metadata Cache Timeout for a Shared File System)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device and a number of seconds, the meta_timeo command sets the maximum metadata cache age for the specified device to the value supplied. Hosts use cached file-system attributes and directory data for the specified number of seconds, then refresh the cache by checking with the metadata server (MDS).
The default caching interval is 3 seconds. An interval of 0 disables metadata caching.
The following example sets the cache interval on family set device 800 to 0 seconds, so that the host checks metadata consistency with the MDS prior to each use of the file system:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:meta_timeo 800 0
mh_write (Turn on Multi-Host Read/Write)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device, the mh_write command enables simultaneous reads and writes to the same file from multiple hosts.
By default, shared Oracle HSM file systems allow simultaneous reads from multiple hosts but not writes. To reverse the effects of the mh_write command, use nomh_write (Turn Off Multi-Host Read/Write).
The following example enables multi-host reads on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:mh_write 800
min_pool (Set the Minimum Number of Shared File System Threads)The number of execution threads associated with a shared file system grows and shrinks dynamically based on load. The min_pool command sets a minimum number of shared file system threads that the system must maintain.
For the number of threads, specify an integer in the range [8-2048]. For Solaris systems, the default is 64 threads. For Linux clients, the default is 8 threads.
The following example sets the minimum number of threads to 128 on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:min_pool 800 128
minallocsz (Set the Minimum Allocation Size for a Shared File System)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device and a kilobyte value, the minallocsz command sets minimum number of kilobytes that are allocated ahead of a write to the specified shared file system. See maxallocsz (Set the Maximum Allocation Size for a Shared File System).
The specified allocation size must be a multiple of 8 kilobytes in the range [16-4194304]. The default value is 8 times the device allocation unit (DAU). See the discussion of the -a option on the sammkfs man page for additional information on DAUs.
The following example sets the minimum allocation to 512 kilobytes on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:minallocsz 800 512
mm_stripe (Set the Metadata Stripe Width)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a number of disk allocation units (DAUs) per stripe, the mm_stripe command configures the corresponding file system to write the specified number of 16-kilobyte disk allocation units (DAUs) to each logical unit number (LUN) that is available for metadata storage.
The default value, 1 DAU, writes one DAU of metadata to one LUN before switching to another LUN. A 0 (zero) value disables striping and writes metadata DAUs to metadata LUNs round-robin.
See also stripe (Set the Stripe Width for a File System) and the mount_samfs man page.
The following example disables metadata striping on family set device 10:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on sam2 -----2---r- 80% 70% /sam2 /server3 mm 11 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 12 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 13 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:mm_stripe 20 0
n (Display the Staging Status Screen)The n command opens the Staging status screen. The screen displays the status of the stager and a list of outstanding stage requests.
When issued without a parameter, the n command displays staging status for all media types. When the command is issued with an Oracle HSM media-type identifier, the command displays status for the specified media type only. Media-type identifiers are listed in Appendix A and on the mcf man page.
noabr (Disable Application Based Recovery)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the noabr command does not let Oracle Real Application Clusters (RAC) use the Application Based Recovery feature of Solaris Volume Manager to handle recovery of databases stored in Oracle HSM file systems that use asynchronous I/O with Solaris Volume Manager mirrored volumes.
By default, ABR is enabled. See abr (Enable Application Based Recovery).
The following example disables ABR on family-set device 20:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr 14 on /dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:noabr 20
noalloc (Disable Storage Allocation on a Device)When issued with an Oracle HSM equipment ordinal number, the noalloc command stops storage allocation on the specified data device until an administrator issues an alloc (Enable Storage Allocation on a Device) command.
The following example stops storage allocation on disk device 14:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam/sam2 server4 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam/sam1 server4 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr14on/dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:noalloc 14
The device state changes to noalloc:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam/sam2 server4 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam/sam1 server4 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr14 noalloc/dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0
noatime (Disable Access Time Updates for a File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the noatime command disables updates to the access time attribute on the corresponding file system. It is equivalent to the command atime -1. See atime (Configure Access Time Updates for a File System).
Never use the noatime command with an archiving, Oracle HSM file system.
nodio_szero (Do Not Zero Uninitialized Areas of Sparse Files with Direct I/O)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the nodio_szero command makes direct I/O to the specified device skip the zeroing of uninitialized areas. This improves performance when writing sparse files directly.
The default is nodio_szero.
For further information, see dio_szero (Zero Uninitialized Areas of Sparse Files During Direct I/O).
The following example disables the zeroing of sparse files on family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:nodio_szero 100
nodmr (Disable Directed Mirror Reads of Software Mirrors)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the dmr command does not let Oracle Real Application Clusters (RAC) use the Directed Mirror Reads feature of Solaris Volume Manager when databases are stored on Oracle HSM file systems that use asynchronous I/O with Solaris Volume Manager mirrored volumes. See dmr (Enable Directed Mirror Reads of Software Mirrors).
By default, DMR is enabled.
The following example disables DMR on family-set device 20:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0 ma 10 on sam1 m----2---r- 80% 70% /sam1 mm 11 on /dev/dsk/c0t600A0B8000118ACA000009594E1565E8d0s0 mr 12 on /dev/dsk/c0t600A0B8000114D5400000CA64E156793d0s0 mr 13 on /dev/dsk/c0t600A0B8000114D5400000CAA4E156897d0s0 mr 14 on /dev/dsk/c0t600A0B8000118ACA0000095B4E1568C8d0s0 mr 15 on /dev/dsk/c0t600A0B8000118ACA0000095D4E156948d0s0:command:nodmr 20
noforcedirectio (Do Not Use Direct I/O by Default)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the noforcedirectio command restores the usual default I/O method, Direct I/O, on the specified device.
For more information, see "Understanding I/O Performance Tuning Objectives and Options", the forcedirectio (Use Direct I/O by Default) command, and the directio, setfa, sam_setfa, and sam_advise man pages.
The following example restores the default, paged I/O, on family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:noforcedirectio 100
noforce_nfs_async (Do Not Force Asynchronous NFS)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the noforce_nfs_async command causes the specified device to write NFS (Network File System) data through to disk unless NFS has requested that the data be cached.
The following example sets noforce_nfs_async on family set 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:noforce_nfs_async 100
nohwm_archive (Do Not Start Archiving at the High Water Mark)When issued with an Oracle HSM family-set equipment ordinal number, the nohwm_archive command configures the corresponding file system so that it does not start archiving automatically when the total size of the stored files passes the high water mark, the maximum space-utilization threshold specified for the file system. See maxpartial (Set the Maximum Partial Release Size for the File System).
The following example configures disk family-set device 20 so that it does not automatically start archiving when the corresponding file-system is 80% full:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:nohwm_archive 20
nomh_write (Turn Off Multi-Host Read/Write)When issued with the Oracle HSM equipment ordinal number of a shared disk family-set device, the nomh_write command disables simultaneous reads and writes to the same file from multiple hosts and restores the default setting for the specified device.
By default, shared Oracle HSM file systems allow simultaneous reads from multiple hosts but not writes. Use nomh_write to restore the default after running mh_write (Turn on Multi-Host Read/Write).
The following example disables multi-host reads on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:mh_write 800
norefresh_at_eof (Do Not Refresh at EOF in a Multi-Reader File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device that is mounted read-only in a single-writer/multi-reader file system, the norefresh_at_eof command disables automatic updates of the file size attribute when the read buffer exceeds the end of file. See refresh_at_eof (Refresh at EOF in a Multi-Reader File System).
The following example disables automatic refresh of the file size attribute on family set device 900:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 900 on qfs1 m----2--R-d 90% 70% /qfs1 mm 901 on /dev/dsk/c5t10d0s0 md 902 on /dev/dsk/c5t10d0s1 md 903 on /dev/dsk/c5t11d0s0 md 904 on /dev/dsk/c5t11d0s1:command:norefresh_at_eof 900
noqwrite (Disable Simultaneous Reads/Writes from Different Threads)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the noqwrite command disables simultaneous read and write operations to a single file from different threads. One writer has exclusive access to the file and other writers and readers must wait for the file to be released, per the UNIX standard.
While simultaneous multi-thread reads and writes to a single file can be useful in database applications (see qwrite (Enable Simultaneous Reads/Writes to a File from Different Threads), this capability is not generally useful or desirable. So noqwrite is the default setting.
The following example disables simultaneous read and write operations to a single file on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2--R-d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:noqwrite 100
nosuid (Mount a File System with setuid Execution Disabled)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the nosuid command mounts the file system with the setuid file attribute disabled. The setuid attribute (set user ID upon execution) lets file system users run executables with the temporarily elevated permissions that may be required when using the files.
See suid (Mount a File System with setuid Execution Enabled) and the mount man page for further information.
The default is suid, mount the file system with the setuid attribute enabled.
The following example mounts the file system on family set device 100 with the setuid attribute disabled:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2--R-d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:nosuid 100
nosw_raid (Turn Off Software RAID Mode)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the nosw_raid command disables software RAID mode on the specified device and restores the Oracle HSM default.
The following example disables software-RAID mode for family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:nosw_raid 100
notrace (Disable File System Tracing)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the notrace command disables tracing for the corresponding file system.
The default is trace (enabled).
For more information about file system tracing, see "trace (Enable File System Tracing)" and the defaults.conf man page.
The following example disables tracing for the file system on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2--R-d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:notrace 100
o (Display the Optical Disk Status Screen)The o command opens the Optical disk status screen. The screen displays the status of all optical disk drives configured within the environment.
Ctrl-K (^k)Select (manual, automated library, both, priority)
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
The following example shows the optical disc drive status display.
Optical disk status samuversiontimedatety eq status act use state vsn mo 35 --l---wo-r 1 29% ready OPD004
The following table shows the field descriptions for this display.
tyThe ty field displays the Oracle HSM media type. Media-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq field displays the user-assigned Oracle HSM equipment ordinal number of the optical disc drive.
statusThe status field displays a ten-character string in which each character position represents the value of a different removable-media device attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 10. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | s--------- |
A lowercase s in position 1 means that media is being scanned. |
| 1 | m--------- |
A lowercase m in position 1 means that the automated library is operational. |
| 1 | M--------- |
An uppercase M in position 1 means that the device is in maintenance mode. |
| 2 | -E-------- |
An uppercase E in position 2 means that the device encountered an unrecoverable error during scanning. |
| 2 | -a-------- |
A lowercase a in position 2 means that the device is in audit mode. |
| 3 | --l------- |
A lowercase l in position 3 means that media has a label. |
| 3 | --L------- |
An uppercase L in position 3 means that the media is being labeled |
| 3 | --N------- |
An uppercase N in position 3 means that the media is foreign (not part of the Oracle HSM archive). |
| 4 | ---I------ |
An uppercase I in position 4 means that the device is idle and not available for new connections. But operations already in progress are continuing until completion. |
| 4 | ---A------ |
An uppercase A in position 4 means that the device needs operator attention. |
| 5 | ----C----- |
An uppercase C in position 5 means that the device needs cleaning. |
| 5 | ----U----- |
An uppercase U in position 5 means that an unload has been requested. |
| 6 | -----R---- |
An uppercase R in position 6 means that the device is reserved. |
| 7 | ------w--- |
A lowercase w in position 7 means that a process is writing to the media. |
| 8 | -------o-- |
A lowercase o in position 8 means that the device is open. |
| 9 | --------P- |
An uppercase P in position 9 means that the device is positioning a tape volume. |
| 9 | --------F- |
For a tape library, an uppercase F in position 9 means that all storage slots are occupied.
For a drive, an uppercase |
| 10 | ---------R |
An uppercase R in position 10 means that the device is ready and the media is read-only. |
| 10 | ---------r |
A lowercase r in position 10 means that the device is spun up and ready. |
| 10 | ---------p |
A lowercase p in position 10 means that the device is present. |
| 10 | ---------W |
An uppercase W in position 10 means that the device is write protected. |
actThe act field displays the activity count.
useThe use field displays the percentage of cartridge space used.
stateThe state field displays the current operating state of the optical disc drive:
ready means that the device is on and media are loaded.
notrdy means that the device is on, but media are not loaded.
idle means that the device is completing operations that are already in progress but is not accepting new connections.
off The device is not available.
down means that the device is available for maintenance only.
vsnThe vsn field displays either the volume serial number assigned to the media volume or the keyword nolabel if the volume is not labeled.
off (Turn Off A Device)The off command sets the device specified by an Oracle HSM equipment ordinal number offline. All operations stop.
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 on/dev/rmt/2cbn 100 libraryA hy 104 on historian 104:command:off 103
The device state changes to off:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 off/dev/rmt/2cbn 100 libraryA hy 104 on historian 104
on (Turn On A Device)When issued with the Oracle HSM equipment ordinal number of a device, the on command brings the device specified by an Oracle HSM equipment ordinal number online.
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 off/dev/rmt/2cbn 100 libraryA hy 104 on historian 104:command:on 103
The device state changes to on:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 on/dev/rmt/2cbn 100 libraryA hy 104 on historian 104
open (Open a Disk Storage Device for Reading)When issued with the Oracle HSM equipment ordinal number of a disk device, the open command opens the device for the read (Read Sectors On a Disk Storage Device) command.
The following example opens device 23:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam/sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:open 23
The screen switches to displaying disk sectors, starting at sector 0:
Sector: 00000000 (0) raw samuversiontimedate00000000 00000000 00000000 00000000 00000000 ................ 00000010 00000000 00000000 00000000 00000000 ................ 00000020 00000000 00000000 00000000 00000000 ................ 00000030 00000000 00000000 00000000 00000000 ................ 00000040 00000000 00000000 00000000 00000000 ................ 00000050 00000000 00000000 00000000 00000000 ................ 00000060 00000000 00000000 00000000 00000000 ................ 00000070 00000000 00000000 00000000 00000000 ................ ...
p (Display the Removable Media Load Requests Screen)The p command opens the Removable media load requests screen. The screen displays information about pending load requests for removable media. Four different screen formats are available:
Both manual and automated library requests sorted by user
Both manual and automated library requests sorted by priority.
Manual requests only
Automated library requests only.
When issued without a parameter, the p command displays load requests for all media. When the command is issued with an Oracle HSM media-type identifier, the command displays load requests for the specified media type only. Media-type identifiers are listed in Appendix A and on the mcf man page.
You can use the following control keys on this screen:
Ctrl-K (^k)Toggle between the different display formats.
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
The following example shows the removable media load requests display.
Removable media load requests all both samuversiontimedatecount: 1 index type pid user rb flags wait count vsn 0 li 15533 root 150 W--f--- 0:00 VOL002
indexThe index field displays the corresponding request's position in the preview table.
typeThe type field displays the media type code assigned to the removable media.
pidThe pid field displays the UNIX process identifier. A process identifier of 1 indicates NFS access.
userThe user field displays the name assigned to the user requesting the load.
priorityThe priority field displays the degree to which the request may take precedence over other requests.
rbThe rb field displays the user-assigned equipment ordinal number of the automated library in which the requested VSN resides.
flagsThe flags field displays a seven-character strings in which each character position represents the value of a property of the request. Character positions are numbered starting from the left, position 1, and ending at the right, position 7. The following table defines the possible flag codes for each position in the status string.
| Position | Flag | Meaning |
|---|---|---|
| 1 | W------ |
An uppercase W in position 1 means that write access has been requested. |
| 2 | -b----- |
A lowercase b in position 2 means that the request is busy. |
| 3 | --C---- |
An uppercase C in position 3 means that the request for the VSN has been cleared. |
| 4 | ---f--- |
A lowercase f in position 4 means that file-system access has been requested. |
| 5 | ----N-- |
An uppercase N in position 5 means that the requested media is foreign to the Oracle HSM file system. |
| 6 | -----S- |
An uppercase S in position 6 means that the flip side of a two-sided optical-media volume is mounted. |
| 7 | ------s |
A lowercase s in position 7 means that a file is to be staged from removable media to the file system. |
waitThe wait field displays the elapsed time since the mount request was received.
countThe count field displays the number of requests for this VSN, if a file is being staged.
vsnThe vsn field displays the volume serial name of the removable media volume.
partial (Set the Default Partial Release Size for the File System)When issued with an Oracle HSM family-set equipment ordinal number and a number of kilobytes, the partial command sets the default number of bytes from the start of a file that can be retained in disk cache when the file is released.
A large file that has been designated for partial release can open more quickly when restaged from archival media because the beginning of the file is already on disk when staging begins. Users can override the default on a file-by-file basis by specifying a size when marking a file for partial release. For further information, see maxpartial (Set the Maximum Partial Release Size for the File System), partial_stage (Set Partial Stage-Ahead Point for the File System), and the release man page.
The specified kilobyte value must be an integer in the range [8-maximumpartial], where maximumpartial is the value set by the maxpartial command. The default is 16 kilobytes.
The following example configures disk family-set device 20 to retain 256 kilobytes on disk for each file designated for partial release:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:partial 20 256
partial_stage (Set Partial Stage-Ahead Point for the File System)When issued with an Oracle HSM family-set equipment ordinal number and a number of kilobytes, the partial_stage command defines a file offset beyond which a file designated for partial release is staged to disk in its entirety rather than in part.
Partially staging a large file can give users access to data more quickly because the file can open before it is fully staged from archival media to the disk cache. But, once enough data is in cache to start opening the file, there is no further advantage to partial staging. For further information, see maxpartial (Set the Maximum Partial Release Size for the File System), partial (Set the Default Partial Release Size for the File System), and the release man page.
The specified kilobyte value must be an integer in the range [8-maximumpartial], where maximumpartial is the value set by the maxpartial command. The default is the value set by the partial command.
The following example sets the partial stage-ahead point for disk family-set device 20 to 5124 kilobytes (5 megabytes):
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% /sam2 mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:partial_stage 20 5124
priority (Set Priority in Load-Request Preview Queue)When issued with the process identifier (pid) of a load request and an integer value, the priority command moves the load request to the specified position in the list of waiting jobs. Specify the pid for one of the load requests listed by the p (Display the Removable Media Load Requests Screen) command. Specify the desired priority as an integer, with 0 (zero) being first in line (the highest priority).
The following example gives the load request with pid 17708 (for volume VOL202) a priority of 0:
Removable media load requests all both samuversiontimedatecount: 4 index type pid user rb flags wait count vsn 0 li 13850 root 900 W--f--- 0:42 VOL002 1 li 15533 root 900 ---f--- 0:28 VOL0012li17708root 900 ---f--- 0:28VOL2023 li 19937 root 900 ---f--- 0:05 VOL015:command:priority 17708 0
With its new priority, pid 17708 moves from third in the queue (index 2) to first (index 0):
index type pid user rb flags wait count vsn
0 li 17708 root 900 ---f--- 0:28 VOL202
1 li 15533 root 900 ---f--- 0:28 VOL001
2 li 15533 root 900 ---f--- 0:28 VOL001
3 li 19937 root 900 ---f--- 0:05 VOL015
qwrite (Enable Simultaneous Reads/Writes to a File from Different Threads)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the qwrite command enables simultaneous read and write operations to a single file from different threads.
In databases and other applications that handle multiple simultaneous transactions to the same file, simultaneous reads and writes improve I/O performance by queuing multiple requests at the drive level.
The default is noqwrite (Disable Simultaneous Reads/Writes from Different Threads).
The following example enables simultaneous read and write operations to a single file on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2----d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:qwrite 100
r (Display the Removable Media Status ScreenThe r command opens the Removable media status screen. The screen displays the activity on removable media devices such as tape drives. You can monitor either a specific type of device or a family of devices such as all tape devices.
When issued without a parameter, the r command displays the status for all removable media devices. When the command is issued with an Oracle HSM media type identifier, the command displays the status of the specified media type only. Media-type identifiers are listed in Appendix A and on the mcf man page.
The following example shows the removable media status display.
Removable media status: all samuversiontimedatety eq status act use state vsn li 150 --l------r 0 63% ready VOL004
The following table shows the field descriptions for this display.
tyThe ty field displays Oracle HSM media type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq field displays the user-assigned Oracle HSM equipment ordinal number of the mass storage device.
statusThe status field displays a ten-character string in which each character position represents the value of a different removable-media device attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 10. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | s--------- |
A lowercase s in position 1 means that media is being scanned. |
| 1 | m--------- |
A lowercase m in position 1 means that the automated library is operational. |
| 1 | M--------- |
An uppercase M in position 1 means that the device is in maintenance mode. |
| 2 | -E-------- |
An uppercase E in position 2 means that the device encountered an unrecoverable error during scanning. |
| 2 | -a-------- |
A lowercase a in position 2 means that the device is in audit mode. |
| 3 | --l------- |
A lowercase l in position 3 means that media has a label. |
| 3 | --L------- |
An uppercase L in position 3 means that the media is being labeled |
| 3 | --N------- |
An uppercase N in position 3 means that the media is foreign (not part of the Oracle HSM archive). |
| 4 | ---I------ |
An uppercase I in position 4 means that the device is idle and not available for new connections. But operations already in progress are continuing until completion. |
| 4 | ---A------ |
An uppercase A in position 4 means that the device needs operator attention. |
| 5 | ----C----- |
An uppercase C in position 5 means that the device needs cleaning. |
| 5 | ----U----- |
An uppercase U in position 5 means that an unload has been requested. |
| 6 | -----R---- |
An uppercase R in position 6 means that the device is reserved. |
| 7 | ------w--- |
A lowercase w in position 7 means that a process is writing to the media. |
| 8 | -------o-- |
A lowercase o in position 8 means that the device is open. |
| 9 | --------P- |
An uppercase P in position 9 means that the device is positioning a tape volume. |
| 9 | --------F- |
For a tape library, an uppercase F in position 9 means that all storage slots are occupied.
For a drive, an uppercase |
| 10 | ---------R |
An uppercase R in position 10 means that the device is ready and the media is read-only. |
| 10 | ---------r |
A lowercase r in position 10 means that the device is spun up and ready. |
| 10 | ---------p |
A lowercase p in position 10 means that the device is present. |
| 10 | ---------W |
An uppercase W in position 10 means that the device is write protected. |
rdlease (Set the Read Lease Time for a Shared File System)Before reading a shared file, a host must request and obtain an read lease that grants it exclusive access to the file. The rdlease command defines the number of seconds during which this lease remains valid on the specified device. If the specified time runs out before the read operation is complete, the host must reacquire the lease before proceeding.
The specified number of seconds must fall in the range [15-600]. The default is 30 seconds.
For related information, see aplease (Set Append Lease Expiration Time for a Shared File System), wrlease (Set the Write Lease Time for a Shared File System), and lease_timeo (Set the Unused Lease Timeout for a Shared File System).
The following example sets the read lease time to 2 minutes (120 seconds) on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1 mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2 mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:rdlease 800 120
read (Read Sectors On a Disk Storage Device)When issued with a hexadecimal sector address on a previously opened disk device, the read command displays disk sectors starting at the specified address.
See also open (Open a Disk Storage Device for Reading).
The following example reads device 23 starting at sector 1f4 (500):
Sector: 00000000 (0) raw samuversiontimedate00000000 00000000 00000000 00000000 00000000 ................ 00000010 00000000 00000000 00000000 00000000 ................ 00000020 00000000 00000000 00000000 00000000 ................ 00000030 00000000 00000000 00000000 00000000 ................ 00000040 00000000 00000000 00000000 00000000 ................ 00000050 00000000 00000000 00000000 00000000 ................ 00000060 00000000 00000000 00000000 00000000 ................ 00000070 00000000 00000000 00000000 00000000 ................ ...:command:read 1f4
The screen displaying disk sectors starting at sector 1f4:
Sector: 000001f4 (500) raw samuversiontimedate00000000 0a01c1c1 e39ce073 8e00e319 39e80195 ..AAc.`s..c.9h.. 00000010 a40cd199 bfd5285f ba1509f9 41c03820 $.Q.?U(_:..yA@8 00000020 90780583 7f78825b 9fa0a58b 4a519c5d .x...x.[. %.JQ.] 00000030 d2774afa bb256497 569bd2fb b775ddf2 RwJz;%d.V.R{7u]r 00000040 ba5cd1b7 2a56d93d 6e9dafab ba8b5e56 :\Q7*VY=n./+:.^V 00000050 d5686a5b 334d1f96 70060ec1 b46483c8 Uhj[3M..p..A4d.H 00000060 190339e7 3ce3e5e7 a702ff00 9394528c ..9g<ceg'.....R. 00000070 4c9f3990 6dc05c1e b907e6dd 918e14e4 L.9.m@\.9.f]...d ...
readahead (Set The Maximum Read-Ahead Parameter)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the readahead command specifies the minimum number of contiguous bytes that the device should read into memory. When correctly configured, read-ahead improves file-system performance by reducing disk I/O and making larger, contiguous blocks of information immediately available in memory.
Specify a read-ahead value in the range [1-8192] kilobytes. The default is 1024 kilobytes (1 megabyte).
You can also set a readahead directive in the samfs.cmd file. See the samfs.cmd man page for details.
The following example sets the read-ahead value for family set device 100 to 1 megabyte (1024 kilobytes):
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T128K16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:readahead 100 1024
The read-ahead value changes to 1 megabyte.
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma100m----2---r- 1% on 7.270T 7.270T1M16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T
readonly | ro (Make Device Read Only)The readonly or ro command applies the read-only attribute to the device specified by an Oracle HSM equipment ordinal number.
The following example makes device 12 read-only:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 10 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 11 1% on 0 465.253G 465.231G [975661056 inodes] mr 12 1% on 1 1.817T 1.817T mr 13 1% on 2 1.817T 1.817T mr 14 1% on 3 1.817T 1.817T mr 15 1% on 4 1.817T 1.817T:command:readonly 12
refresh (Set the Automatic Refresh Interval for the samu Display)When issued with a number of seconds, the refresh command configures samu to refresh screens after the specified number of seconds has passed.
The following example configures samu to refresh the display every 20 seconds:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 900 on qfs1 m----2--R-d 90% 70% /qfs1 mm 901 on /dev/dsk/c5t10d0s0 md 902 on /dev/dsk/c5t10d0s1 md 903 on /dev/dsk/c5t11d0s0 md 904 on /dev/dsk/c5t11d0s1:command:refresh 20
refresh_at_eof (Refresh at EOF in a Multi-Reader File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device that is mounted read-only in a single-writer/multi-reader file system, the refresh_at_eof command forces an update of the file size attribute whenever the read buffer exceeds the end of file. See norefresh_at_eof (Do Not Refresh at EOF in a Multi-Reader File System).
The following example enables automatic refresh of the file size attribute on family set device 900:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 900 on qfs1 m----2--R-d 90% 70% /qfs1 mm 901 on /dev/dsk/c5t10d0s0 md 902 on /dev/dsk/c5t10d0s1 md 903 on /dev/dsk/c5t11d0s0 md 904 on /dev/dsk/c5t11d0s1:command:refresh_at_eof 900
release (Release Archived Data Files and Disable Allocation on a Device)When issued with an Oracle HSM equipment ordinal number, the release command releases all archived files that are stored on the device, including any space associated with files otherwise designated for partial release, and turns the specified device off.
The release command sets the specified device to the noalloc (Disable Storage Allocation on a Device) state and has the sam-fsd daemon run a sub-process, sam-shrink. The sam-shrink process places all archived files that reside on the specified device in the offline state, so that the data are available from archival media only. When the process completes successfully, the specified device is left in the off state. If the process encounters any unarchived files or files that are in the process of being archived or staged back from archive to disk, the process fails and the specified equipment remains in the noalloc state.
The release process can take a significant amount of time. Monitor its progress via samu and/or via the /var/opt/SUNWsamfs/trace/sam-shrink file and the log file configured in /etc/opt/SUNWsamfs/shrink.cmd (if any).
Note that both remove (Move Data and Remove Equipment from a File System) and release shrink a file system by taking disk devices from a family set without losing data. They differ because the remove command has to consolidate files on the remaining disk devices in the set before freeing the specified device. The release command does not: in an archiving file system, all data is already duplicated on removable media. See add (Add Equipment to a Mounted File System), remove (Move Data and Remove Equipment from a File System), and the sam-shrink man page for related information.
The following example, releases all files on disk slice /dev/dsk/c5t8d0s8 and shrinks family set samfs4:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on samfs4 -----2---r- 80% 70% mm 21 on /dev/dsk/c5t8d0s6 md 22 on /dev/dsk/c5t8d0s7 md23 on/dev/dsk/c5t8d0s8:command:release 23
The device state changes to noalloc:
md 23 noalloc /dev/dsk/c5t8d0s8
Finally, when the process completes successfully, the device state changes to off:
md 23 off /dev/dsk/c5t8d0s8
remove (Move Data and Remove Equipment from a File System)When issued with an Oracle HSM equipment ordinal number, the remove command moves data from the specified device to the remaining devices in the file system and then turns the specified device off.
The remove command sets the specified device to the noalloc (Disable Storage Allocation on a Device) state and has the sam-fsd daemon run a sub-process, sam-shrink, to carry out the actual work. The sam-shrink process moves data from the specified equipment to the file system's other, on storage devices (if the specified equipment is a stripe group, there has to be another stripe group with the same number of devices). When the process completes successfully, the specified device is left in the off state. If data could not be moved for any reason, the equipment is not removed and remains in the noalloc state.
The removal process can take a significant amount of time. Monitor its progress via samu and/or via the /var/opt/SUNWsamfs/trace/sam-shrink file and the log file configured in /etc/opt/SUNWsamfs/shrink.cmd (if any).
Note that both release (Release Archived Data Files and Disable Allocation on a Device) and remove shrink a file system by taking disk devices from a family set without losing data. They differ because the remove command has to consolidate files on the remaining disk devices in the set before freeing the specified device. The release command does not: in an archiving file system, all data is already duplicated on removable media. See add (Add Equipment to a Mounted File System), and the sam-shrink man page for related information.
The following example moves all files from disk slice /dev/dsk/c5t8d0s3 and shrinks family set qfs1:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on qfs1 -----2---r- 80% 70% mm 11 on /dev/dsk/c5t8d0s0 md 12 on /dev/dsk/c5t8d0s1 md13 on/dev/dsk/c5t8d0s3:command:remove 13
The device state changes to noalloc:
md 13 noalloc /dev/dsk/c5t8d0s3
Finally, when the process completes successfully, the device state changes to off:
md 13 off /dev/dsk/c5t8d0s3
s (Display Device Status Screen)The s command opens the Device status screen. The screen displays the status for all devices configured within the environment.
You can use the following control keys on this screen:
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
The following example shows the device status display.
Device status samuversiontimedatety eq state device_name fs status sk 100 on /etc/opt/SUNWsamfs/dcstkconf 100 m--------r stk_dismount / (2275) 0, volser VOL073 sg 120 on /dev/rmt/2cbn 100 ---------p sg 130 on /dev/rmt/5cbn 100 --l----o-r Ready for data transfer sg 140 on /dev/rmt/6cbn 100 ---------p sg 150 on /dev/rmt/4cbn 100 ---------p hy 151 on historian 151 ----------
tyThe ty field displays Oracle HSM media type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq displays the user-assigned Oracle HSM equipment ordinal number of the mass storage device.
stateThe state field displays the current operating state of the device.
device_nameThe device_name field displays the path to the device. For file system devices, this is the file system name.
fsThe fs field displays the user-assigned Oracle HSM equipment number of the family set to which the device belongs.
statusThe status field displays a ten-character string in which each character position represents the value of a different removable-media device attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 10. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | s--------- |
A lowercase s in position 1 means that media is being scanned. |
| 1 | m--------- |
A lowercase m in position 1 means that the automated library is operational. |
| 1 | M--------- |
An uppercase M in position 1 means that the device is in maintenance mode. |
| 2 | -E-------- |
An uppercase E in position 2 means that the device encountered an unrecoverable error during scanning. |
| 2 | -a-------- |
A lowercase a in position 2 means that the device is in audit mode. |
| 3 | --l------- |
A lowercase l in position 3 means that media has a label. |
| 3 | --L------- |
An uppercase L in position 3 means that the media is being labeled |
| 3 | --N------- |
An uppercase N in position 3 means that the media is foreign (not part of the Oracle HSM archive). |
| 4 | ---I------ |
An uppercase I in position 4 means that the device is idle and not available for new connections. But operations already in progress are continuing until completion. |
| 4 | ---A------ |
An uppercase A in position 4 means that the device needs operator attention. |
| 5 | ----C----- |
An uppercase C in position 5 means that the device needs cleaning. |
| 5 | ----U----- |
An uppercase U in position 5 means that an unload has been requested. |
| 6 | -----R---- |
An uppercase R in position 6 means that the device is reserved. |
| 7 | ------w--- |
A lowercase w in position 7 means that a process is writing to the media. |
| 8 | -------o-- |
A lowercase o in position 8 means that the device is open. |
| 9 | --------P- |
An uppercase P in position 9 means that the device is positioning a tape volume. |
| 9 | --------F- |
For a tape library, an uppercase F in position 9 means that all storage slots are occupied.
For a drive, an uppercase |
| 10 | ---------R |
An uppercase R in position 10 means that the device is ready and the media is read-only. |
| 10 | ---------r |
A lowercase r in position 10 means that the device is spun up and ready. |
| 10 | ---------p |
A lowercase p in position 10 means that the device is present. |
| 10 | ---------W |
An uppercase W in position 10 means that the device is write protected. |
snap (Save a Snapshot of the Current Screen to a File)When issued with a file path and name, the snap command saves the current screen as a text file with the specified name.
The following example saves a snap shot of the File systems screen to the file /export/home/samadmin/filsysscrn.txt:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:snap /export/home/samadmin/filsysscrn.txt
stage_flush_behind (Set the Stage Flush Behind Size)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a number of kilobytes, the stage_flush_behind command configures the corresponding file system to asynchronously write the specified number of kilobytes from dirty (modified) stage memory pages to disk.
As staged files are copied from archival media into memory, flush-behind insures that the staged data are quickly saved to the file system's non-volatile primary disk cache. Unsaved, dirty cache pages are kept to a minimum during staging, which helps to maintain the integrity of the file data.
On the other hand, flush-behind increases overhead and can reduce I/O performance. The process of flushing a data block may delay subsequent re-writes of that block. So flush-behind is disabled by default.
Set the specified value in the range [0-8192] kilobytes, where 0—the default—disables the flush-behind capability.
The following example sets the stage_flush_behind for disk family-set device 20 to 5124 kilobytes (5 megabytes):
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:stage_flush_behind 20 5124
stage_n_window (Size the Buffer for Reading Directly from Archival Media)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a number of kilobytes, the stage_n_window command sets the buffer size that Oracle HSM uses when reading files from the corresponding file system that are configured to never automatically stage to disk. Oracle HSM reads such files directly from the archival media to memory in chunks of the specified number of kilobytes.
Specify a number of kilobytes in the range [64-2097152]. The default is 8192 kilobytes.
The following example sets the stage_n_window for disk family-set device 20 to 16384 kilobytes (16 megabytes):
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 20 on sam2 -----2---r- 80% 70% mm 21 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 22 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 23 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:stage_n_window 20 16384
stclear (Clear a Stage Request)When entered with an Oracle HSM media-type identifier and the volume serial number (VSN) of a removable media cartridge, the stclear command removes the corresponding staging request from the staging queue. Media-type identifiers are listed in Appendix A and on the mcf man page.
In the example, stclear clears the staging request for the LTO (li) volume with VSN VOL001 from the staging queue:
Staging queue by media type: all samuversiontimedatety length fseq ino position offset vsnli 451.611k 20 1030 207cc 473 VOL001li 341.676k 20 1031 207cc 7fc VOL002 li 419.861k 20 1032 207cc aa9 VOL003 li 384.760k 20 1033 207cc df2 VOL004 li 263.475k 20 1034 207cc 10f5 VOL005 li 452.901k 20 1035 207cc 1305 VOL006 li 404.598k 20 1036 207cc 1690 VOL007 : command:stclear li.VOL001
The staging request for the LTO (li) volume VOL001 has been cleared:
Staging queue by media type: all samuversiontimedatety length fseq ino position offset vsn li 341.676k 20 1031 207cc 7fc VOL002 li 419.861k 20 1032 207cc aa9 VOL003 li 384.760k 20 1033 207cc df2 VOL004 li 263.475k 20 1034 207cc 10f5 VOL005 li 452.901k 20 1035 207cc 1305 VOL006 li 404.598k 20 1036 207cc 1690 VOL007
stidle (Stop Staging as Soon as Possible)The stidle command idles the stager as soon as current staging operations are complete.
In the example, stidle idles staging:
Staging queue by media type: all samuversiontimedatety length fseq ino position offset vsn li 451.611k 20 1030 207cc 473 VOL001 li 341.676k 20 1031 207cc 7fc VOL002 li 419.861k 20 1032 207cc aa9 VOL003 li 384.760k 20 1033 207cc df2 VOL004 li 263.475k 20 1034 207cc 10f5 VOL005 li 452.901k 20 1035 207cc 1305 VOL006 li 404.598k 20 1036 207cc 1690 VOL007 : command:stidle
stripe (Set the Stripe Width for a File System)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a number of disk allocation units (DAUs), the stripe command configures the corresponding file system to write the specified number of DAUs to each logical unit number (LUN) that is available for data storage.
Specify a number of DAUs in the range [0-255], where a 0 (zero) value disables striping and writes DAUs to LUNs round-robin.
For file systems of the ms equipment type and file systems of the ma type that do not include striped group equipment (gXXX), the default value is 1 DAU with a minimum DAU size of 128 kilobytes. For shared file systems and for ma file systems that include striped groups, the default is 0.
The system also sets stripe width to 0 automatically if a file system contains mismatched striped groups.
For more information about file system types, see the mcf man page and mm_stripe (Set the Metadata Stripe Width).
The following example changes data striping on family set device 10 to two DAUs:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 10 on sam2 -----2---r- 80% 70% /sam/sam2 mm 11 on /dev/dsk/c0t600A0B800026B6F000001F0F4CCFF833d0s0 mr 12 on /dev/dsk/c0t600A0B800026E584000021864C4943AAd0s0 mr 13 on /dev/dsk/c0t600A0B800026E5840000218B4CBD9C4Ed0s0:command:stripe 20 2
strun (Start Staging)The strun command starts the stager.
In the example, strun starts staging:
Staging queue by media type: all samuversiontimedatety length fseq ino position offset vsn li 451.611k 20 1030 207cc 473 VOL001 li 341.676k 20 1031 207cc 7fc VOL002 li 419.861k 20 1032 207cc aa9 VOL003 li 384.760k 20 1033 207cc df2 VOL004 li 263.475k 20 1034 207cc 10f5 VOL005 li 452.901k 20 1035 207cc 1305 VOL006 li 404.598k 20 1036 207cc 1690 VOL007 : command:strun
suid (Mount a File System with setuid Execution Enabled)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the suid command mounts the file system with the setuid attribute enabled. The setuid attribute (set user ID upon execution) lets file system users run executables with the temporarily elevated permissions that may be required when using the files.
See nosuid (Mount a File System with setuid Execution Disabled) and the mount man page for further information.
The default is suid, mount the file system with the setuid attribute enabled.
The following example mounts the file system on family set device 100 with the setuid attribute enabled:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma100on qfs1 m----2----d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:suid 100
sw_raid (Turn On Software RAID Mode)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the sw_raid command enables software RAID mode on the specified device.
The following example enables software-RAID mode for family set device 100:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:sw_raid 100
sync_meta (Synchronize File System Metadata)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and a Boolean value (0 for no or 1 for yes), the sync_meta command determines whether metadata is written to disk every time it changes. A 0 (no) value configures the file system to cache metadata changes in a buffer prior to writing. This reduces the amount of physical I/O, improves I/O efficiency, and thus improves overall performance. A 1 (yes) value configures Oracle HSM to write metadata changes to disk immediately. This sacrifices performance in order to maximize data consistency.
For Oracle HSM file systems that are neither shared nor mounted multi-reader, the default is 0. For shared and multi-reader file systems, the default is 1. If a shared file system is configured for failover, the value must be 1.
The following example configures the file system on family set device 100 to synchronize metadata for maximal data consistency:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2----d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on / dev/dsk/c5t11d0s1:command:sync_meta 100 1
The second example configures the file system on family set device 200 to cache metadata changes for best performance:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 200 on qfs2 m----2----d 90% 70% /qfs2 mm 201 on /dev/dsk/c5t10d0s3 md 202 on /dev/dsk/c5t10d0s4 md 203 on /dev/dsk/c5t11d0s3 md 204 on /dev/dsk/c5t11d0s4:command:sync_meta 200 0
t (Display Tape Drive Status Screen)The t command opens the Tape drive status screen. The screen displays status information for all tape drives configured within the environment.
You can use the following control keys on this screen:
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
The following example shows the tape drive status display.
Tape drive status samuversiontimedatety eq status act use state vsn sg 120 ---------p 0 0% notrdy sg 130 ---------p 0 0% notrdy sg 140 ---------p 0 0% down drive set down due to ACS reported state sg 150 --l------r 0 41% ready VOL014
tyThe ty field displays Oracle HSM media type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq displays the user-assigned Oracle HSM equipment ordinal number of the mass storage device.
statusThe status field displays a ten-character string in which each character position represents the value of a different removable-media device attribute. Character positions are numbered starting from the left, position 1, and ending at the right, position 10. The following table defines the possible state codes for each position in the status string.
| Position | Value | Meaning |
|---|---|---|
| 1 | s--------- |
A lowercase s in position 1 means that media is being scanned. |
| 1 | m--------- |
A lowercase m in position 1 means that the automated library is operational. |
| 1 | M--------- |
An uppercase M in position 1 means that the device is in maintenance mode. |
| 2 | -E-------- |
An uppercase E in position 2 means that the device encountered an unrecoverable error during scanning. |
| 2 | -a-------- |
A lowercase a in position 2 means that the device is in audit mode. |
| 3 | --l------- |
A lowercase l in position 3 means that media has a label. |
| 3 | --L------- |
An uppercase L in position 3 means that the media is being labeled |
| 3 | --N------- |
An uppercase N in position 3 means that the media is foreign (not part of the Oracle HSM archive). |
| 4 | ---I------ |
An uppercase I in position 4 means that the device is idle and not available for new connections. But operations already in progress are continuing until completion. |
| 4 | ---A------ |
An uppercase A in position 4 means that the device needs operator attention. |
| 5 | ----C----- |
An uppercase C in position 5 means that the device needs cleaning. |
| 5 | ----U----- |
An uppercase U in position 5 means that an unload has been requested. |
| 6 | -----R---- |
An uppercase R in position 6 means that the device is reserved. |
| 7 | ------w--- |
A lowercase w in position 7 means that a process is writing to the media. |
| 8 | -------o-- |
A lowercase o in position 8 means that the device is open. |
| 9 | --------P- |
An uppercase P in position 9 means that the device is positioning a tape volume. |
| 9 | --------F- |
For a tape library, an uppercase F in position 9 means that all storage slots are occupied.
For a drive, an uppercase |
| 10 | ---------R |
An uppercase R in position 10 means that the device is ready and the media is read-only. |
| 10 | ---------r |
A lowercase r in position 10 means that the device is spun up and ready. |
| 10 | ---------p |
A lowercase p in position 10 means that the device is present. |
| 10 | ---------W |
An uppercase W in position 10 means that the device is write protected. |
actThe status field displays the activity count.
useThe use field displays the percentage of cartridge space used.
stateThe state field displays the current operating state of the device. Valid device states include:
on means that the device is available.
ro means that the device is available read-only.
off means that the device is not available.
down means that the device is available for maintenance only.
idle means that the device is completing operations that are already in progress but is not accepting new connections.
noalloc means that no more storage can be allocated on this device. An administrator has run the noalloc eq command, where eq is the equipment ordinal number of the device. The noalloc state persists until an administrator issues an alloc eq command.
vsnThe vsn field displays the volume serial number assigned to a media volume, the keyword nolabel if a volume is not labeled, or a blank if a volume is not loaded or if the device is off.
thresh (Set High and Low Watermarks)When issued with the Oracle HSM equipment ordinal number of a disk family-set device and two percentages, the thresh (threshold) command defines the high and low water marks for the corresponding file system.
The high water mark is the percentage disk utilization where the Oracle HSM releaser process starts to delete archived (backed-up) files from disk in order to make space for new files. The low-water mark is the percentage of disk utilization where the Oracle HSM releaser has freed enough space for new files and stops deleting archived files from the disk. When properly configured, these two parameters insure that the largest possible number of files are kept available on disk while maintaining sufficient space for new and enlarged files, thus minimizing excessive staging activity and maximizing performance and capacity.
For each watermark, specify an integer value in the range [0-100]. A highs of 80% and a low of 70% is a typical setting.
The following example sets the high and low water marks to 80% and 70% percent for the sam1 file system on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on sam1 m----2----d 90% 60% /sam1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:thresh 100 80 70
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on sam1 m----2----d 80% 70% /sam1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1
trace (Enable File System Tracing)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the trace command enables tracing for the corresponding file system.
The default is trace (enabled).
For more information about file system tracing, see notrace (Disable File System Tracing) and the defaults.conf man page.
The following example enables tracing for the file system on family set device 100:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 100 on qfs1 m----2----d 90% 70% /qfs1 mm 101 on /dev/dsk/c5t10d0s0 md 102 on /dev/dsk/c5t10d0s1 md 103 on /dev/dsk/c5t11d0s0 md 104 on /dev/dsk/c5t11d0s1:command:trace 100
u (Display Staging Queue Screen)The u command opens the Staging queue screen. The screen lists all files in the staging queue.
When issued without a parameter, the u command displays the staging queue for all media. When the command is issued with an Oracle HSM media-type identifier, the command displays the staging queue for the specified media type only. Media-type identifiers are listed in Appendix A and on the mcf man page.
You can use the following control keys on this screen:
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
Ctrl-K (^k)Display the path on the second line of each entry
The following example shows the staging queue display.
Staging queue by media type: all samuversiontimedatety length fseq ino position offset vsn li 451.611k 20 1030 207cc 473 VOL001 li 341.676k 20 1031 207cc 7fc VOL002 li 419.861k 20 1032 207cc aa9 VOL003 li 384.760k 20 1033 207cc df2 VOL004 li 263.475k 20 1034 207cc 10f5 VOL005 li 452.901k 20 1035 207cc 1305 VOL006 li 404.598k 20 1036 207cc 1690 VOL007
tyThe ty field displays the Oracle HSM device type. Media-type identifiers are listed in Appendix A and on the mcf man page.
lengthThe length field displays the length of the file in kilobytes.
fseqThe fseq field displays the file system's user-assigned equipment ordinal number
inoThe ino field displays the inode number of the file that is being staged.
positionThe position field displays the location of the archive file on the specified media volume.
offsetThe offset field displays the position of the file that is being staged relative to the start of the archive file.
vsnThe vsn field displays the volume serial number assigned to the media volume, the keyword nolabel if the volume is not labeled, or a blank if a volume is not loaded or if the volume is loaded into a device that is off.
unavail (Make a Device Unavailable to Oracle HSM)The unavail command keeps Oracle HSM from using the device specified by a given Oracle HSM equipment ordinal number.
The command is useful when you want to reserve a drive for other uses, such as loading media for file-system restores during disaster recovery.
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 on/dev/rmt/2cbn 100 libraryA hy 104 on historian 104:command:unavail 103
The device state changes to unavail:
Device configuration: samuversiontimedatety eq state device_name fs family_set sk 100 on /etc/opt/SUNWsamfs/SL8500_T10K_parms 100 libraryA ti 101 on /dev/rmt/0cbn 100 libraryA ti 102 on /dev/rmt/1cbn 100 libraryA ti103 unavail/dev/rmt/2cbn 100 libraryA hy 104 on historian 104
unload (Unload a Drive)The unload command unloads media mounted on the device specified by an Oracle HSM equipment ordinal number. If the device has magazines, the command unloads the mounted cartridge and ejects the magazine. See load (Load a Cartridge into a Drive).
Device status samuversiontimedatety eq state device_name fs status sn 100 on /etc/opt/SUNWsamfs/dcstkconf 100 m--------r empty li 101 on /dev/rmt/2cbn 100 ---------p empty li 102 on /dev/rmt/5cbn 100 ---------p empty li 103 on /dev/rmt/6cbn 100 --l------r idle li 104 on /dev/rmt/4cbn 100 ---------p empty hy 151 on historian 151 ----------:command:unload 103
v (Display Robot VSN Catalog Screen)The v command opens the Robot VSN catalog screen. The screen lists catalog information for all media volumes stored in a given automated library.
When issued without a parameter, the v command displays the catalogs for all libraries. When the command is issued with a user-assigned Oracle HSM equipment ordinal number, the command displays the volume catalog for the specified device only.
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down to the next catalog.
Ctrl-U (^u)Move up to the previous catalog.
Ctrl-I (^i)Display detailed information, two lines per entry. Press Ctrl-I once to display times and barcodes. Press Ctrl-I a second time to display volume reservations on the second line.
1Sort entries by library slot.
2Sort entries by count.
3Sort entries by percent used.
4Sort entries by Volume Serial Number (VSN).
5Sort entries by access time.
6Sort entries by barcode.
7Sort entries by label time.
Ctrl-K (^k)Toggle between the sorting methods enumerated above.
/Search for a specified Volume Serial Number (VSN) and, if found, display the corresponding record at the top of the list.
%Search for a specified barcode and, if found, display the corresponding record at the top of the list.
$Search for a specified slot number and, if found, display the corresponding record at the top of the list.
The following example shows the automated library catalog display sorted by slot:
Robot VSN catalog by slot : eq 900 samuversiontimedatecount 32 slot access time count use flags ty vsn 0 2013/07/02 12:53 3 96% -il-o-b----f li VOL001 1 2013/07/02 12:53 2 98% -il-o-b----f li VOL002 2 2013/07/02 12:52 1 86% -il-o-b----f li VOL003 3 2013/07/02 12:52 1 95% -il-o-b----f li VOL004 4 2013/07/02 12:51 1 79% -il-o-b----- li VOL005 5 2013/07/02 12:45 0 14% -ilEo-b----f li VOL006 MEDIA ERROR 6 2013/07/02 12:46 0 0% -il-o-b----- li VOL007 7 2013/07/02 12:46 0 0% -il-o-b----- li VOL008 8 2013/07/02 12:47 0 0% -il-o-b----- li VOL009 9 2013/07/02 12:47 0 0% -il-o-b----- li VOL010 10 2013/07/02 12:48 0 0% -il-o-b----- li VOL011 11 2013/07/02 12:48 0 0% -il-o-b----- li VOL012 12 2013/07/02 12:48 0 0% -il-o-b----- li VOL013 13 2013/07/02 12:49 0 0% -il-o-b----- li VOL014 14 2013/07/02 12:49 0 0% -il-o-b----- li VOL015 15 2013/07/02 12:50 0 0% -il-o-b----- li VOL016 16 2013/07/02 12:50 0 0% -il-o-b----- li VOL017 17 2013/07/02 12:51 0 0% -il-o-b----- li VOL018 18 2013/07/02 12:51 0 0% -il-o-b----- li VOL019 19 none 50 0% -il-oCb----- li CLN020
The next example shows the same screen sorted by the percentage of each volume that has been used:
Robot VSN catalog by % used : eq 900 samuversiontimedatecount 32 slot access time count use flags ty vsn 1 2013/07/02 12:53 2 98% -il-o-b----f li VOL002 0 2013/07/02 12:53 3 96% -il-o-b----f li VOL001 3 2013/07/02 12:52 1 95% -il-o-b----f li VOL004 2 2013/07/02 12:52 1 86% -il-o-b----f li VOL003 4 2013/07/02 12:51 1 79% -il-o-b----- li VOL005 5 2013/07/02 12:45 0 14% -ilEo-b----f li VOL006 MEDIA ERROR 12 2013/07/02 12:48 0 0% -il-o-b----- li VOL013 13 2013/07/02 12:49 0 0% -il-o-b----- li VOL014 14 2013/07/02 12:49 0 0% -il-o-b----- li VOL015 15 2013/07/02 12:50 0 0% -il-o-b----- li VOL016 16 2013/07/02 12:50 0 0% -il-o-b----- li VOL017 17 2013/07/02 12:51 0 0% -il-o-b----- li VOL018 19 none 50 0% -il-oCb----- li CLN020 6 2013/07/02 12:46 0 0% -il-o-b----- li VOL007 7 2013/07/02 12:46 0 0% -il-o-b----- li VOL008 8 2013/07/02 12:47 0 0% -il-o-b----- li VOL009 9 2013/07/02 12:47 0 0% -il-o-b----- li VOL010 10 2013/07/02 12:48 0 0% -il-o-b----- li VOL011 11 2013/07/02 12:48 0 0% -il-o-b----- li VOL012 18 2013/07/02 12:51 0 0% -il-o-b----- li VOL019
tyThe ty field displays Oracle HSM media type. Device-type identifiers are listed in Appendix A and on the mcf man page.
eqThe eq displays the user-assigned Oracle HSM equipment ordinal number of the mass storage device.
countThe count field displays the number of accesses to this volume since the last audit operation.
slotThe slot field displays the slot number that identifies a storage cell in a robotic library.
access timeThe access time field displays the date and time when the volume was last accessed.
useThe use field displays the percentage of cartridge space used.
flagsThe flags field displays a twelve-character string in which each character position represents an attribute of the catalog entry. Character positions are numbered starting from the left, position 1, and ending at the right, position 12. In some cases, more than one flag can occur in a field, and one flag overrides the other.
The following table defines the possible flag codes for each position in the string:
| Position | Flag | Meaning |
|---|---|---|
| 1 | A----------- |
An uppercase A in position 1 means that the volume in the corresponding library slot needs an audit. |
| 2 | -i---------- |
A lowercase i in position 2 means that the corresponding library slot is in use. |
| 3 | --l--------- |
A lowercase l in position 3 means that the volume in the corresponding library slot is labeled. |
| 3 | --N--------- |
An uppercase N in position 3 means that the volume in the corresponding library slot is unlabeled or foreign to the Oracle HSM file system. |
| 4 | ---E-------- |
An uppercase E in position 4 means that there was a media error when accessing the corresponding volume. |
| 5 | ----o------- |
A lowercase o in position 5 means that the corresponding library slot is occupied by a volume. |
| 6 | -----C------ |
An uppercase C in position 6 means that the corresponding library slot is occupied by a cleaning tape. |
| 6 | -----p------ |
A lowercase p in position 6 means that the corresponding library slot is occupied by a priority volume. |
| 7 | ------b----- |
A lowercase b in position 7 means that the volume in the corresponding library slot has a barcode. |
| 8 | -------W---- |
An uppercase W in position 8 means that the volume in the corresponding library slot is write-protected by a physical switch on the cartridge. |
| 9 | --------R--- |
An uppercase R in position 9 means that the volume in the corresponding library slot is logically read-only. |
| 10 | ---------c-- |
A lowercase c in position 10 means that the volume in the corresponding library slot has been selected for recycling. |
| 11 | ----------d- |
A lowercase d in position 11 means that the volume in the corresponding library slot has a duplicate volume serial number. |
| 11 | ----------U- |
An uppercase U in position 11 means that the volume that should reside in the corresponding library slot is unavailable. |
| 12 | -----------f |
A lowercase f in position 11 means that the archiver found the volume in the corresponding library slot full. |
| 12 | -----------X |
An uppercase X in position 11 means that the corresponding library slot is an export slot. |
vsnThe vsn field displays the volume serial number assigned to the media, the keyword nolabel if the volume is not labeled, or a blank if a volume is not loaded or if the device is off.
LVTimeThe LVTime field displays the date and time when the entire tape was last verified.
LVPosThe LVPos field displays the position where the last verification was canceled.
w (Display the Pending Stage Queue Screen)The w command opens the Pending stage queue screen. The screen displays requests for files that are not currently available on disk and must be staged from removable media volumes that are not currently loaded.
When issued without a parameter, the w command displays queued stage requests for all media types. When the command is issued with an Oracle HSM media-type identifier, the command displays the pending stage requests for the specified media type only. Media-type identifiers are listed in Appendix A and on the mcf man page.
You can use the following control keys on this screen:
Ctrl-B (^b)Move back one page.
Ctrl-F (^f)Move forward one page.
Ctrl-D (^d)Move down the page.
Ctrl-U (^u)Move up the page.
Ctrl-K (^k)Display the path on the second line of each entry
The following example shows the pending stage queue screen for all media types. Thirteen requested files reside on one volume that is waiting to load:
Pending stage queue by media type: all samuversiontimedatevolumes 1 files 13 ty length fseq ino position offset vsn li 1.383M 1 42 3a786 271b VOL002 li 1.479M 1 56 3a786 5139 VOL002 li 1018.406k 1 60 3a786 6550 VOL002 li 1.VOLM 1 65 3a786 7475 VOL002 li 1.528M 1 80 3a786 99be VOL002 li 1.763M 1 92 3a786 ce57 VOL002 li 1.749M 1 123 3a786 11ece VOL002 li 556.559k 1 157 3a786 1532f VOL002 li 658.970k 1 186 3a786 17705 VOL002 li 863.380k 1 251 3a786 1dd58 VOL002 li 1.268M 1 281 3a786 1f2b7 VOL002 li 1.797M 1 324 3a786 23dfa VOL002 li 1.144M 1 401 3a786 2bb6d VOL002
tyThe ty field displays the Oracle HSM device type. Media-type identifiers are listed in Appendix A and on the mcf man page.
lengthThe length field displays the length of the file in kilobytes.
fseqThe fseq field displays the file system's user-assigned equipment ordinal number
inoThe ino field displays the inode number of the requested data file.
positionThe position field displays the location on the media where the required archive file starts.
offsetThe offset field displays the starting location of the requested data file relative to the archive file position.
vsnThe vsn field displays the volume serial number assigned to the media, the keyword nolabel if the volume is not labeled, or a blank if a volume is not loaded or if the device is off.
wr_throttle (Set the Maximum Size of Outstanding Writes)The wr_throttle command limits the amount of data that any one file can keep in memory before writing to the Oracle HSM disk family-set device specified by a supplied equipment ordinal number. Throttling keeps outstanding writes from consuming excessive memory.
The default data amount is 5% of the system memory, expressed in kilobytes.
:wr_throttle eq value
For eq, specify the user-assigned Oracle HSM equipment number for a file system.
For value, specify an integer number of kilobytes. If value=0, there is no limit.
The following example sets the wr_throttle value for family set device 100 to 838656 kilobytes (819 megabytes), which is 5% of the 16 gigabytes of memory available on the host system:
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:wr_throttle 100 838656
writebehind (Set the Write-Behind Parameter)When issued with the Oracle HSM equipment ordinal number of a disk family-set device, the writebehind command specifies the maximum number of bytes that the device should cache in memory before writing data to disk. When correctly configured, write-behind improves file-system performance by writing information in larger, contiguous blocks.
Specify a write-behind value in the range [1-8192] kilobytes. The default is 512 kilobytes.
You can also set a writebehind directive in the samfs.cmd file. See the samfs.cmd man page for details.
The following example sets the write-behind value for family set device 100 to 1024 kilobytes (1 megabyte):
Mass storage status samuversiontimedatety eq status use state ord capacity free ra part high low ma 100 m----2---r- 1% on 7.270T 7.270T 1M 16 80% 70% mm 110 1% on 0 465.253G 465.231G [975661056 inodes] mr 120 1% on 1 1.817T 1.817T mr 130 1% on 2 1.817T 1.817T mr 140 1% on 3 1.817T 1.817T mr 150 1% on 4 1.817T 1.817T:command:writebehind 100 1024
wrlease (Set the Write Lease Time for a Shared File System)Before writing data to a shared file, a host must request and obtain an write lease that grants it exclusive access to the file. The wrlease command defines the number of seconds during which this lease remains valid on the specified device. If the specified time runs out before the write operation is complete, the host must reacquire the lease before proceeding.
The specified number of seconds must fall in the range [15-600]. The default is 30 seconds.
For related information, see aplease (Set Append Lease Expiration Time for a Shared File System), rdlease (Set the Read Lease Time for a Shared File System), and lease_timeo (Set the Unused Lease Timeout for a Shared File System).
The following example sets the write lease time to 2 minutes (120 seconds) on family set device 800:
File systems samuversiontimedatety eq state device_name status high low mountpoint server ma 800 on shareqfs1 m----2c--r- 80% 70% /shareqfs1mm 801 on /dev/dsk/c6t50020F2300004655d0s0 mr 802 on /dev/dsk/c6t50020F2300004655d0s1 ma 810 on shareqfs2 m----2c--r- 80% 70% /shareqfs2mm 811 on /dev/dsk/c6t50020F2300004655d0s6 mr 812 on /dev/dsk/c6t50020F2300004655d0s7:command:wrlease 800 120
