archiver.cmd - Oracle HSM archiver commands file
∕etc∕opt∕SUNWsamfs∕archiver.cmd
SUNWsamfs
Commands for controlling the archiver are read from
∕etc∕opt∕SUNWsamfs∕archiver.cmd,
which is the archiver commands file.
The
archiver.cmd
file must be free from errors, or the
archiver does not execute.
Use the
archiver -lv
command to check the
archiver.cmd
file for syntax errors. When it is free from errors, use the
samd config
command to reconfigure the daemons.
Archiving with tape drive data protection (DIV) is enabled in the
defaults.conf file by setting the div parameter.
Distributed archiving is enabled in the defaults.conf file by setting
the distio parameter.
Archive sets and associated media are defined in the archiver command file.
Archive Sets are the mechanism that the archiver uses to direct files
in a
samfs
file system to media during archiving.
All files in the file system are members of one and only one archive set. Characteristics of a file are used to determine archive set membership. All files in an archive set are copied to the media associated with the archive set. The archive set name is simply a synonym for a collection of media volumes.
Files are written to the media in an archive file, which is written in
tar
format. The combination
of the archive set and the
tar
format results in an operation that is just
like using the command
find
to select files for the
tar
command.
In addition, the meta data (directories, the indices
of segmented files, and the
removable media information), are assigned to an archive set to be copied
to media. The archive set name is the name of the file system. (See
mcf (4)).
For segmented files, the archivable unit is the segment, not the entire file, so the properties and priorities apply to the segments themselves rather than to the entire file. The index of a segmented file contains no user data and so is assigned to the meta data archive set.
Symbolic links are considered data files for archival purposes.
Each archive set may have up to four archive copies defined. The copies provide duplication of files on different media. Copies are selected by the Archive Age of a file.
The archiver command file consists of directive lines. In this man page, the archiver directives are separated into the following sections and subsections:
General Directives section
Archive Set Assignments section
Archive Copy Definitions section
Archive Set Copy Parameters section
Archive Set Copy Parameters - General
Archive Set Copy Parameters - Priority
Archive Set Copy Parameters - Scheduling
Archive Set Copy Parameters - Recycling
VSN Pool Definitions section
VSN Associations section
Each of these lines consists of one or more fields separated by white
space. Leading white space is ignored. Everything
after a # character is ignored. Lines may be continued by using a \ (back-slash)
as the last character on the line.
All parameter settings and archive set definitions apply to all file systems
(global) until a file system directive is encountered. Thereafter, the settings
and definitions apply only to the named file system (local). The directives
archmax,
bufsize,
drives,
notify,
and
ovflmin
can only be global and hence are not allowed after the first
fs=
directive.
General directives are identified by the = (equals sign) character in the second field or by the absence of additional fields.
archmax = media target_size Set the archive file maximum size for media media to target_size. Files to be archived will be placed on the media in a single archive file of length less than or equal to target_size. If a single file is greater than target_size, then this restriction does not apply.
Sizes appropriate to the media are used by default. The default size for STK Titanium, all LTO, IBM TS1120 and IBM 3592 media is 22GB. The default size for SKT 9940 media is 11GB. The default size for STK 9840 media is 4GB. The default size for linear tape is 11GB. The default size for all other tape media is 8GB. The default size for disk is 1G. The default size for optical media is 1GB.
archivemeta = state Set the meta data archiving state on or off.
state
may be on or off. Meta data archiving is off by default.
background_interval = time Set the interval between background scans to time.
The default is 24 hours. If
time
is a multiple of days, the background scan will be performed at the
background_time.
background_time = hhmm Set the time of day for the background scan to hhmm local time.
The default 0000 (midnight).
dircache_size = size Set the maximum size of the directory cache to size. The directory cache for name lookups will not exceed this size. If the file system contains very large directories, increasing this value may help performance. The minimum value is 8M and the maximum is 512M.
The default is 64M.
prefetch_window = length Set the maximum number of small files per lun to be simultaneously prefetched during archive. The minimum value is 0 which will disable prefetching, and the maximum value is 32768.
The default is 16 files per lun.
prefetch_size = size Set the maximum size of files to be prefetched during archive. Prefetching intends to improve small file archiving performance. The minimum value is 0 and the maximum is 1M.
The default is 8K.
bufsize = media buffer_size [ lock ] Set the archive buffer size for media
media
to
buffer_size *
dev_blksize,
and (optionally) lock the buffer.
For media, specify a valid media type from the list
on the mcf (4) man page.
For buffer_size, specify a number from 2 through 8192. The
default is 16. This value is multiplied by the dev_blksize
value for the media type, and the resulting buffer size
is used. The dev_blksize can be specified in
the defaults.conf file.
The lock argument indicates whether or not the archiver should
use locked buffers when making archive copies.
If lock
is specified, the archiver sets file locks on the archive buffer
in memory for the duration of the sam-arcopy (1m) operation.
This avoids paging the buffer, and it can provide a performance
improvement. The lock argument should be specified only on
large systems with large amounts of memory. If insufficient memory
is present, it can cause an out of memory condition.
The lock argument is effective only if direct I∕O is enabled
for the file being archived.
By default, lock is not
specified and the file system sets the locks on all direct I∕O buffers,
including those for archiving.
This directive can also be specified on an archive set
basis by placing
the -bufsize=buffer_size and -lock directives
between params and endparams directives.
For
more information on this, see the -bufsize=buffer_size
and -lock directives
mentioned later on this man page.
For more information on dev_blksize, see the defaults.conf
man page. For more information on enabling direct I∕O, see the setfa (1)
man page, the sam_setfa (3x) library routine man page, or
the -o forcedirectio option on the mount_samfs (1m) man page.
drives = library count Set the number of drives to use for archiving on library (the library family set name as defined in the mcf) to count. The archiver will use only count number of drives in library to create archive copies. This directive prevents the archiver from using all drives in a library and possibly interfering with staging.
The default value is the actual number of drives in the library.
The example below uses 3 drives from library gr50:
drives = gr50 3
examine = method Set the file system examination method to method. Files in a file system are examined using the method defined by this directive. method may be any one of the following:
scanScan the file system in the traditional manner. The first scan is a directory scan, all successive scans are inode scans.
scandirsAll scans are directory scans.
scaninodesAll scans are inode scans.
noscanNo periodic scans are performed. Files are examined when they change.
The default examine method is noscan.
fs = file_system Start local definitions for file system file_system. All parameter settings and archive set definitions will apply only to this file system. This directive may be followed by copy definitions to define multiple copies for the file system meta data.
The defaults are no local definitions and one archive copy for the file system data.
interval = time Set the interval between archive operations to time.
The default time is 10 minutes.
logfile = filename Set the name of the archiver log file to
filename,
specified as an absolute pathname.
The archiver log file contains a line for each file archived. The line
contains information about the file that includes the date, time,
media, volume, archive set, and the name of the file.
Note that it is possible to have a separate log file for each file system
(by placing a logfile = definition after an fs = definition).
The default is no log file.
notify = filename Set the name of the archiver event notification script file to
filename.
This file is executed by the archiver to allow the system administrator
to process various events in a site specific fashion. The script is called
with a keyword for the first argument. The keywords are:
emerg,
alert,
crit,
err,
warning,
notice,
info,
and
debug.
Additional arguments are described in the default script, ∕etc∕opt∕SUNWsamfs∕scripts∕archiver.sh.
ovflmin = media minimum_size Set the minimum size of a file which will require more than one volume for media media to minimum_size. Files to be archived that are smaller than this size will be placed on only a single volume of the media. Files that are larger than this size will be allowed to be written to multiple volumes.
If not specified, volume overflow will not take place.
scanlist_squash = state Control the sam-arfind scanlist consolidation.
state
may be on or off. If files in two or more subdirectories
with the same parent directory need to be scanned by sam-arfind at a much
later time, the scan entries can be consolidated if
state
is
on.
The sam-arfind scanlist consolidation is
off
by default.
setarchdone = state Control the changing of the state of the archdone flag for a file when
the file is examined by sam-arfind.
state
may be
on
or
off.
When all archive copies for a file have been made, the
archdone
flag is set for that file to indicate that no further archive action is
required. The
archdone
flag is used by the archiver only during an inodes scan to avoid looking
up the path name for the inode. Setting
archdone
for files that will never be archived can be a time consuming operation
during directory scans impacting performance when large directories are
scanned. Therefore, this will no longer be done by default. To get the
previous behavior, set the
state
to
on.
The default value of
state
is
off
for
examine = scandirs
and
examine = noscan.
This option does not affect setting the state of
archdone
when archive copies are made.
waitThe archiver will not begin archiving until it receives a start command from archiver, samu, or samcmd. This is a mechanism to allow other activities to be performed before archiving begins. The wait may be applied globally or to one or more file systems.
The default is no waiting. However, if archiver.cmd does not exist then the default is to wait.
timeout = [ operation | media ] time External events may cause the archiving I∕O operations to stop for indefinite periods of time. This will hamper timely archiving of other files that are not affected by the external delays. Timeouts are provided for the operations that may get stopped. The timeout values for the write operation may also be specified for individual media. The operation may be any one of the following:
readReading the file from the disk. The default is 1 minute. This timeout will be set to the same value as the write timeout (default 15 minutes) when offline_copy = direct.
requestRequesting the archive media. The default is 15 minutes.
stageStaging the file to be archived. The default is 0 (no timeout).
writeWriting to the archive media. The default is 15 minutes for removable archive media. The default is 0 (no timeout) for disk archive media.
rearch_offline_behind = state This direcive specifies the method to be used for scheduling rearchive files that are offline at the time of the rearchive. If there are offline and online files in tar file, the archive I∕O operation needs to wait for the stage to complete, thus archiving online files are blocked. This directive separates offline files from online files and schedule the offline I∕O operation at the lowest priority. The operation may be either of the following:
onSeparates offline files from online files and schedule the offline I∕O operation at the lowest priority. An internal archive set copy is used for scheduling the rearchive offline operations. See ARCHIVE SET COPY PARAMETERS SECTION.
offDo not separate offline files from online files.
The default value is off.
archive set assignments are made by describing the characteristics of the
files that should belong to the set. The statements that do this are
patterned after the
find(1)
command. The archive set name is the first field,
followed by the path relative to the Oracle HSM
file system mount point.
The path may be enclosed in quotation mark
characters, for instance,
"project∕gifs".
Within the quoted string, the usual character escapes are
allowed, including octal character value.
The remaining fields are either the file characteristics for membership in the set, or controls for the set.
It is possible that the choice of file characteristics for several archive sets will result in ambiguous set membership. These situations are resolved in the following manner:
The archive set with the earliest definition in the command file is chosen.
Local definitions for the file system are chosen before the global definitions.
These rules imply that more restrictive archive set definitions should be closer to the beginning of the command file.
It is also possible to use the same archive set name for several different file characteristics. An example would assign files that are owned by several users into a single archive set.
Assigning files to a special archive set called no_archive prevents
files from being archived. This can be useful for temporary files.
The no_archive archive set assignment definition must be a
local definition to be effective.
The archive set assignments may be followed by archive copy definitions.
You can specify one or more of the following file characteristics:
-user uname Include files belonging to user uname.
-group gname Include files belonging to group gname.
-minsize size Include files greater than or equal to
size,
where size is a number paired with one of the following units: b (bytes), k (kilobytes), M (megabytes), G (gigabytes), or T (terabytes).
-maxsize size Include files less than size.
-name regular_expression Include files with full paths that match regular_expression. The regular expression is limited to 255 characters.
-access age Include files whose access time is older than
age,
where age is a string of numbers paired with units of the form s (seconds), m (minutes), h (hours), d (days), w (weeks), or y (years).
-nftvPerform no file time validation.
By default, the access and modification times of files are validated to assure that these times are greater than or equal to the file creation time, and less than or equal to the time at which the file is examined. This is to provide proper archiving and unarchiving.
For files that have been migrated into a directory, this may not be the
desired behavior. The
-nftv
(no file time validation) parameter may be used to prevent the validation
of file access and modification times for files that are in the archive
set defined by these definitions.
-after date_time Include files that have been created or modified since
date_time,
where date_time is in ISO 8601 format: YYYY-MM-DD[Thh:mm:ss][Z].
If the time is not specified, it defaults to 00:00:00.
The Z specifies Coordinated Universal Time (UTC). The default is local time.
All of the following are valid date and time specifications:
2015-10-08T12:15:47
2015-10-08
2015-10-08T17:15:47Z
-release attributes Set the release attributes (see
release (1))
for all files matching the file characteristics
on this archive set definition. The
attributes
may include a (always), d (reset to default), n (never), p (partial),
or sxx (partial size xx).
-stage attributes Set the stage attributes (see
stage (1))
for all files matching the file characteristics
on this archive set definition. The
attributes
may include a (associative), d (reset to default), or n (never).
When controlling archiving for a specific file system (using the
fs =
fsname
directive), directives local to the file system level are evaluated before the
global directives. Thus, files may be assigned to a local archive set
(including the
no_archive
archive set) instead of being assigned to a
global archive set. This has implications when setting global archive
set assignments such as
no_archive.
Assume, for example, the following
archiver.cmd
segment:
no_archive. -name \\.o
fs = samfs1
allfiles .
1 10s
fs = samfs2
allfiles .
1 10s
At first look it appears that the administrator
intended not to archive any
of the
.o
files in both file systems. However, since the local archive
set assignment
allfiles
is evaluated prior to the global archive set assignment
no_archive,
the
.o
files in in both file systems are archived.
To ensure that no
.o
files are archived, the following segment would be used:
fs = samfs1
no_archive. -name \\.o
allfiles .
1 10s
fs = samfs2
no_archive. -name \\.o
allfiles .
1 10s
The archive copy definitions determine when the archive copies are made for the files matching file characteristics. These definitions consist of lines beginning with a digit. This digit is the copy number.
The first fields after the copy number are the option flags as described below:
-releaseThis causes the cache disk space for the files to be released immediately after the copy is made.
-noreleaseThis flag may be used to prevent automatic release of cache disk space until
all copies marked with this flag are made. The
-norelease
option makes the
archiver set eligible to be released after all copies have been archived, but
the files will not be released until the releaser is invoked and selects them
as release candidates. Using this flag on just one copy will have no effect on
automatic release.
The combination of
-release
and
-norelease
will cause the archiver to release the file when all the copies having
this combination are made. With this usage, the archive set is released
immediately, rather than waiting for the releaser to be invoked, as is the
case with the
-norelease
option alone.
If the
-release
option is used on a copy that does not have the
-norelease
option set, the file will get released when that copy is made, overriding the
effect of any
-norelease
usage on other copies.
The time interval between the last modification to the oldest unarchived file and the start of archiving.
Specify the archive_age as a number paired with one of the units s (seconds), m (minutes), h (hours), d (days), w (weeks), or y (years).
The default is 4 minutes (4m).
The time interval after which files should be removed removed from the archive.
The default is to never unarchive the copy.
Archive set parameters may be set after all archive sets are defined. The
beginning of this section is noted by the directive
params.
The section is ended by the end of the archiver command file or the directive
endparams.
Setting an archive set parameter requires at least three fields: the archive set copy name, the parameter name and the parameter value.
The archive set copy name is the archive set name and copy number separated by a dot (.).
Parameters may be set for all archive sets by using the pseudo archive set copy
allsets
for the directive. If the
allsets
is specified without a copy number, the parameters apply to all archive set copies. If specified with a copy number, the parameters apply to only those
archive set copies with the same copy number. All
allsets
directives must occur before those for any actual archive set copies.
The default value for all parameters is either zero (0) or
none
unless otherwise specified.
In the example below, all copies of all archive sets will be sorted by path. For all archive sets, copy 1 will use 3 drives, and copy 2 will use 2 drives.
allsets -sort path
allsets.1 -drives 3
allsets.2 -drives 2
If an archive copy of a file is being rearchived, an internal, rearchive set copy schedules the archive operation. The rearchive set copy uses the archive parameters from the archive set copy.
You can set the archive set parameters using the archive set copy name or the pseudo archive set copy name allsets.copy-number followed by the character R. The rearchive set copy lets
you differentiate between new and rearchive operations and use different
parameters for each operation.
In the example below, all new copies are written using up to 3 drives. Rearchive
copies are limited to 1 drive and have a lower priority.
archset.2 -drives 3
archset.2R -drives 1 -priority -1000
If rearch_offline_behind = on and an archive copy of an offline file is
being rearchived, an internal rearchive offline set copy schedules
the rearchive offline operation.
The rearchive offline set copy uses the
archive parameters from the archive set copy or from a rearchive set copy
if defined. You can set the archive set parameters using the
archive set copy name or the pseudo archive set copy name allsets.copy-number followed by the character r. The
rearchive offline set copy lets you differentiate between new
and rearchive offline operations and use different parameters for
each operation.
In the example below, all new archive copies are written using up to 3 drives. Rearchive online copies are limited to 2 drives. Rearchive offline copies are limited to 1 drive and have a lower priority than new and rearchive online copies.
archset.2 -drives 3
archset.2R -drives 2
archset.2r -drives 1
The general archive set copy parameters are as follows:
-archmax target_size Set the archive file maximum size for this archive set to target_size. Files to be archived will be placed on the media in a single archive file of length less than or equal to target_size. If a single file is greater than target_size, then this restriction does not apply.
If not specified, the
archmax
value for the media is used.
-bufsize = buffer_size Set the archive buffer size to
buffer_size *
dev_blksize.
The default buffer_size is 16.
Valid values are 2 through 8192.
If not specified, the default
buffer size value for the media is used.
This directive can also be specified as a global directive.
For more information on specifying an archive buffer size, see
the bufsize = media buffer_size [lock] directive
described on this man page in the GENERAL DIRECTIVES section.
-directio state Set the file reading method for archival.
state
may be on or off.
The reading performance of files for archival can be changed by using
this parameter. If users are not reading files at the same time that
they are being archived, then selecting
on
allows the archiver to read the file without using the system buffer
cache and using pages that users might need. In the event that users are
reading files while they are being archived, then
off
may be a better choice because the system buffer cache will provide data
to the user and the archiver. The default is
on.
-distio state Set whether distributed tape IO can be used for this archive set.
state
may be on or off.
This parameter can be used to limit which archive sets will use the
distributed IO feature, for example only archive sets containing
large files. The default is
on.
-disk_archive diskvol (Obsolete) Defines a disk archive set. This parameter is obsolete. Disk archive sets should be defined in the VSN associations or VSN pool definitions section. For more information on disk archiving, see the Oracle HSM documentation library.
All of the other archive set parameters work with disk archiving except:
-fillvsns,
-ovflmin
minimum_size,
-reserve
method,
-tapenonstop,
-distio.
None of these cause an error if applied to an archive set that is assigned to disk archiving.
-drivemax max_size Set the multiple drives maximum size for this archive set to
max_size.
When the
-drives
parameter is selected, the amount of data selected to be archived to each
drive will be limited to
max_size.
Using this parameter can result in better drive utilization, because
drives can take different amounts of time to archive files.
Bydefault, this parameter is not set.
-drivemin min_size Set the multiple drives minimum size for this archive set to
min_size.
When the
-drives
parameter is selected, multiple drives will be used only if more than
min_size
data is to be archived at once. The number of drives to be used in parallel
will be the lesser of
total_size ∕ min_size
and the number of drives specified by
-drives.
The default value is
archmax.
-drives number Set the maximum number of drives to use when writing the archive images for this archive set copy to removable media.
Segments are striped across the specified number of drives. The segments are separated into number archive files.
The default is 1 (one drive).
The example below lets the archiver use up to 3 drives for archiving files in the
archive set named
set_name.3:
set_name.3 -drives 3
-fillvsns [ minfill ]When a group of files is archived together, separate them into smaller groups that will fill volumes more efficiently. The optional minfill parameter specifies the smallest amount of free space that that the archiver will try to fill.
By default, when archiving files as a group, the archiver searchs through the volumes associated with the corresponding archive set and stores the files on the first volume that contains enough free space to store them together. As a result, volumes may not always fill to capacity.
The example below fills each volume until less than 1 gigabyte of free space remains:
-fillvsns 1G
-lockLock the archive copy buffer for the duration of
the sam-arcopy (1m) operation.
The -lock directive is effective only if direct I∕O is enabled for
the file being archived.
If not specified, the file system controls the locks on the archive
copy buffer. By default, this directive is disabled.
This directive can also be specified as a global directive.
For more information on controlling the archive buffer locks, see
the bufsize = media buffer_size [lock] directive
described on this man page in the GENERAL DIRECTIVES section.
-offline_copy method Specify the method used for archiving files that are offline when the archiver runs. The offline archiving method can have one of four possible values:
noneFiles are staged as needed for each archive tar file before copying to the archive volume.
directDirect copy.
Copy files directly from the offline volume to the archive volume without using
the cache. Source volume and destination volume are different and two drives
are available. For best performance in this mode, increase
the value of the file system mount parameter stage_n_window (the default value is 256k).
stageaheadStage the next archive tar file while the current archive tar file is written to the destination. With this method, one archive tar file is created on one tape drive (or disk archive) while the offline files needed to create the next archive tar file are being staged from another tape drive (or disk archive). Two drives are available and room is available on cache for all files in one archive tar file.
stageallStage all files before archiving. Use only one drive. The disk cache has enough free space to hold all files.
-ovflmin minimum_size Set the minimum size of a file that will require more than one volume in this archive set to minimum_size. Files to be archived that are smaller than this size will be placed on only a single volume of the media. Files that are this size or larger will be allowed to overflow one volume to at least one additional volume.
If not specified, the
ovflmin
value for the media will be used.
-rearch_stage_copy copy_number Use copy_number for staging an offline copy when rearchiving the copy defined by the archive set. By default, the file will be staged from the copy being rearchived. This option can be used if the copy being rearchived is not available or copy_number is located on a faster media.
-reserve [ set | dir | user | group | fs ]Reserve volumes for this archive set, so that media are not shared with other archive sets.
A ReserveName identifies the reserved volumes by archive set, file system, and∕or owner. The keyword set reserves media by archive set. The keyword fs reserves media by file system. One of the mutually exclusive keywords dir, user, and group reserves media based on the ownership of the files being archived:
The dir keyword reserves volumes using the first 31 characters of the directory path that immediately follows the path
specification in the archive set description.
The user keyword reserves volumes using the user name associated with the file.
The group keyword reserves volumes using the group name associated with the file.
-rsort method | -sort method Sort the files in the archive set using the specified
method
prior to archiving, so that files that share the property specified by
method
are kept together.
If you do not specify a method, the archiver sorts the files by path. If you specify
-rsort
, the archiver sorts the files by reversing the order specified by
method.
The sorting method can be one of the following:
ageSort each archive file by ascending modification time. The oldest files are archived first.
noneNo sorting of the archive file is performed. Files are archived in the order encountered on the file system.
pathSort each archive file by the full pathname of the file. This method will keep files in the same directories together on the archive media.
prioritySort each archive file by descending archive priority. The higher priority files are archived first.
sizeSort each archive file by ascending file size. The smallest files are archived first. The largest files are archived last.
-tapenonstop When files are archived to tape, the default writing mechanism closes the removable media tape file in between each archive file. This action causes the tape subsystem to write a TapeMark followed by an EOF1 label and two TapeMarks. Before another archive file can be written, the tape must be positioned backwards over the EOF1 label.
tapenonstop
parameter causes the archiver to not close the removable media tape file
between each archive file, and write a Tape Mark to separate the archive files.
This speeds writing archive files to tape. The tape cannot be unloaded in
between archive files.
The following parameters allow you to configure a priority system for archiving files. In the following priority parameters, the values are floating-point numbers such that
-3.400000000E+38 ≤ value ≤ 3.402823466E+38.
-priority age value Set the "Archive Age" property multiplier for files in this archive set to value.
-priority archive_immediate value Set the "Archive immediate" property multiplier for files in this archive set to value.
-priority archive_overflow value Set the "Multiple archive volumes" property multiplier for files in this archive set to value.
-priority archive_loaded value Set the "Archive volume loaded" property multiplier for files in this archive set to value.
-priority copy1 value Set the "Copy 1" property multiplier for files in this archive set to value.
-priority copy2 value Set the "Copy 2" property multiplier for files in this archive set to value.
-priority copy3 value Set the "Copy 3" property multiplier for files in this archive set to value.
-priority copy4 value Set the "Copy 4" property multiplier for files in this archive set to value.
-priority copies value Set the "Copies made" property multiplier for files in this archive set to value.
-priority offline value Set the "File off line" property multiplier for files in this archive set to value.
-priority queuewait value Set the "Queue wait" property multiplier for files in this archive set to value.
-priority rearchive value Set the "Rearchive" property multiplier for files in this archive set to value.
-priority reqrelease value Set the "Required for release" property multiplier for files in this archive set to value.
-priority size value Set the "File size" property multiplier for files in this archive set to value.
-priority stage_loaded value Set the "Stage volume loaded" property multiplier for files in this archive set to value.
-priority stage_overflow value Set the "Multiple stage volumes" property multiplier for files in this Archive Set to value.
As files are identified to be archived, they are placed in a list known as an Archive Request. The archive request is scheduled for archival at the end of a file system scan. The following archive set parameters control the archiving workload and assure timely archival of files:
-queue_time_limit time Set the schedule queue time limit for the archive request to time. At the end of the time limit, a notification message will be sent once to alert monitoring entities that the ArchReq has been in the schedule queue longer than the time limit.
-startage time Set the interval between the first file to be archived in the archive request and the start of archiving to time. This allows time to accumulate archival work after the first file has been scheduled for archival. The default is set to two hours.
-startcount count Set the start archiving file count to count. When count files have been identified for archival in the archive request, the archival operation begins. The default is set to 500,000.
-startsize size Set the minimum total size of all files to be archived after the first file to be archived in the archive request to size (in bytes). This allows the accumulation of archival work to be based on the total size of the files that have been scheduled for archival. The default is set to 90% of the -archmax value.
If more than one of -startage, -startcount,
or -startsize are specified, the first condition
encountered starts the archival operation.
If neither -startage, -startcount, nor -startsize
are specified, the archive request is scheduled based
on the examine=method directive, as follows:
If examine = scan | scaninodes | scandirs, the archive request
is scheduled for archiving after the file system scan. Note
that examine = noscan is the default.
If examine = noscan, the default values are as follows:
startage 2 hours
startcount 500,000
startsize 90% of archmax
The -startage, -startcount, and -startsize directives
optimize archive timeliness versus archive work done.
These values override the examine=method specification, if any.
Example 1. If it takes an hour to create files for an Archive Set
that uses -sort path, then you can specify -startage 1h
ensure that all files are created before scheduling the archive request.
Example 2. You can specify -startsize 150G to direct the
archiver to wait until 150 gigabytes of data
are ready to be archived in an archive set.
Example 3. If you know that 3000 files will be generated for
archival, then specify -startcount 3000 to ensure that
the files get archived together.
The following archive set parameters control recycling by archive set. If none of the following parameters are set for an archive set and the name of the archive set is not specified on the recycler's command line, the archive set will not be recycled. Volumes which comprise that archive set (unless also assigned to other archive sets) could be recycled as part of recycling the library which contains them.
-recycle_dataquantity size This option sets a limit of
size
bytes on the amount of data
the recycler will schedule for rearchiving so as to clear volumes of
useful data. Note that the actual number of volumes selected for
recycling may also be dependant on the
-recycle_vsncount
parameter.
The default is 1 gigabyte (1G).
-recycle_hwm percent This option sets the high water mark (hwm) for the archive set. The hwm is expressed as a percentage of the total capacity of the volumes associated with the archive set. When the utilization of those volumes exceeds percent, the recycler will begin to recycle the archive set. The default is 95%. This option is ignored for disk media recycling.
-recycle_ignore This option inhibits the recycler from recycling this archive set. All recycling processing occurs as usual, except any media selected to recycle are not marked "recycle". This allows the recycler's choice of media to recycle to be observed, without actually recycling any media.
-recycle_mailaddr mail-address This option specifies an email address to which informational messages should be sent when this archive set is recycled. The default is not to send any mail.
-recycle_mingain percent This option limits selection of volumes for recycling to those which would increase their free space by percent or more. Volumes not meeting the mingain parameter are not recycled. The default is 50%.
-recycle_vsncount count This option sets a limit of count on the number of volumes the recycler will schedule for rearchiving so as to clear volumes of useful data. Note that the actual number of volumes selected for recycling may also be dependant on the -recycle_dataquantity parameter. The default is 1. This option is ignored for disk media recycling.
-recycle_minobs percent This option is used to set a threshold for the recycler's rearchiving process. When the percentage of obsolete files within an archived tar file on the disk reaches this threshold, the recycler begins moving the valid files from the archive into a new tar file. Once all of the valid files have been moved, the original tar file is marked as a candidate to be removed from the disk archive. This option is ignored for removable media recycling. The default is 50%.
-unarchage time_ref "Set the Unarchive Age computation time reference for this archive set to time_ref. The age of the files will be computed for unarchiving a copy from this time reference. For selecting the desired time reference, time_ref may be:
accessThe age of files for unarchiving a copy is computed from the access time of the file.
modifyThe age of files for unarchiving a copy is computed from the modification time of the file.
The default
time_ref
is
access.
Collections of volumes may be defined in this section. The beginning
of the section is noted by the directive
vsnpools.
The section is ended by the end of the archiver command file or the directive
endvsnpools.
A VSN pool definition requires at least three fields: the pool name, the media type, and at least one VSN.
The media type is the two character mnemonic as described in the
mcf (4)
man page.
The dk or cb identifiers can be used to define a disk archive set.
For more information on disk archiving, see
the StorageTek Storage Archive Manager Configuration and Administration Guide.
regcmp(3C).
VSN associations are defined after all archive sets are defined. The beginning
of the section is noted by the directive
vsns.
The section is ended by the end of the archiver command file or the directive
endvsns.
A VSN association requires at least three fields: the archive set copy, the media type, and at least one VSN.
The archive set copy is the archive set name and copy number separated by '.'.
VSN associations may be set for all archive sets by using the pseudo archive set Copy
allsets
for the directive. If the
allsets
is specified without a copy number, the VSNs apply to all archive set copies. If specified with a copy number, the VSNs apply to only those
archive set copies with the same copy number. All
allsets
directives must occur before those for any actual archive set copies.
If an archive copy of a file is being rearchived, the rearchive set copy uses the VSN associations from the actual archive set copy. If desired, the VSN associations may be set using the archive set copy name followed by the character 'R'. The rearchive set copy allows the users to differentiate 'new' and rearchive operations, and use different VSNs for each operation.
The media type is the two character mnemonic as described in the
mcf (4)
man page.
VSNs are regular expressions as defined in
regcmp(3C).
or VSN pool denoted by the option name
-pool vsn_pool_name
Each VSN on a vsns line is used without leading or trailing spaces as input
to
regcmp(3C).
The compiled form is saved with the archive set copy
definition. When a volume is needed for an archive set copy, each VSN of
each library or manual drive that has sufficient space and is allowed to
be used for archives, is used as the "subject" argument to
regex(3C).
The archive set copy vsn expressions are used as the "re" argument to
regex(3C).
If
regex(3C)
returns with a successful match, the volume is used for the archive set copy.
Example:
set_name.3 mo optic.*
Assigns all files in set_name.3 to the mo media with VSNs beginning with optic.
VSN associations may be defined for all archive sets by using the pseudo
archive set copy
allsets
for the directive. If the
allsets
is specified without a copy number, the VSN associations apply to all archive set Copies. If specified with a copy number, the VSN associations apply to
only those archive set copies with the same copy number. All
allsets
directives must occur before those for any actual archive set copies.
archiver (1m),
archiver.sh (1m),
sam-archiverd (1m),
sam-arcopy (1m),
sam-arfind (1m),
sam-recycler (1m).
regcmp(3C).
