| Skip Navigation Links | |
| Exit Print View | |
|
man pages section 1M: System Administration Commands |
- administer flash archives
flar create -n name [-R root] [-A system_image] [-H] [-I]
[-M] [-S] [-c] [-t [-p posn] [-b blocksize]]
[-i date] [-u section]... [-m master]
[-f [filelist | -] [-F]] [-a author] [-D dataset]
[-e descr | -E descr_file] [-L archiver] [-T type]
[-U key=value]... [-x exclude]... [-y include]...
[-z filelist]... [-X filelist]... archiveflar combine [-d dir] [-u section]...
[-t [-p posn] [-b blocksize]] archiveflar split [-d dir] [-u section]... [-f] [-S section]
[-t [-p posn] [-b blocksize]] archiveflar info [-l] [-k keyword] [-t [-p posn] [-b blocksize]] archive
The flar command is used to administer flash archives. A flash archive is an easily transportable version of a reference configuration of the Solaris operating environment, plus optional other software. Such an archive is used for the rapid installation of Solaris on large numbers of machines. You can create a flash archive using either flar with the create subcommand or the flarcreate(1M) command. See flash_archive(4).
In flash terminology, a system on which an archive is created is called a master. The system image stored in the archive is deployed to systems that are called clones.
A flash archive can be created on a system that is running a UFS root file system or a ZFS root file system. A flash archive of a ZFS root pool contains the entire pool hierarchy except for the swap and dump volumes and any excluded datasets. The swap and dump volumes are created when the flash archive is installed.
There are two types of flash archives: full and differential. Both are created with the create subcommand. A full archive contains all the files that are in a system image. A differential archive contains only differences between two system images. Installation of a differential archive is faster and consumes fewer resources than installation of a full archive.
In creating a differential archive, you compare two system images. A system image can be any of:
a Live Upgrade boot environment, mounted on some directory using lumount(1M) (see live_upgrade(5))
a clone system mounted over NFS with root permissions
a full flash archive expanded into some local directory
To explain the creation of a differential flash archive, the following terminology is used:
The image prior to upgrade or other modification. This is likely the image as it was installed on clone systems.
The old image, plus possible additions or changes and minus possible deletions. This is likely the image you want to duplicate on clone systems.
The flar command compares old and new, creating a differential archive as follows:
files on new that are not in old are added to the archive;
files of the same name that are different between old and new are taken from new and added to the archive;
files that are in old and not in new are put in list of files to be deleted when the differential archive is installed on clone systems.
When creating a differential flash archive, the currently running image is, by default, the new image and a second image, specified with the -A option, is the old image. You can use the -R option to designate an image other than the currently running system as the new image. These options are described below.
Note that differential flash archives are not supported for a ZFS root file system.
You can run flarcreate in multi- or single-user mode. You can also use the command when the master system is booted from the first Solaris software CD or from a Solaris net image. Archive creation should be performed when the master system is in as stable a state as possible.
Following creation of a flash archive, you can use JumpStart to clone the archive on multiple systems. Using JumpStart to install one or more systems is the only installation method available for installing a flash archive of a ZFS root pool.
The flar command includes subcommands for creating, combining, splitting, and providing information about archives. A subcommands is the first argument in a flar command line. These subcommands are as follows:
Create a new flash archive, of a name you specify with the -n argument, based on the currently running system. Use the -A option (described below) to create a differential flash archive.
Combine the individual sections that make up an archive into the archive. If dir is specified (see -d option below), the sections will be gathered from dir; otherwise, they will be gathered from the current directory. Each section is assumed to be in a separate file, the names of which are the section names. At a minimum, the archive cookie (cookie), archive identification (identification), and archive files (archive) sections must be present. If archive is a directory, its contents are archived using files_archived_method (cpio and pax) prior to inclusion in the archive. If so specified in the identification section, the contents are compressed.
Note that no validation is performed on any of the sections. In particular, no fields in the identification section are validated or updated. See flash_archive(4) for a description of the archive sections.
Extract information on an archive. This subcommand is analogous to pkginfo.
Split an archive into one file for each section of the archive. Each section is copied into a separate file in dir, if dir is specified (see -d option below), or the current directory if it is not. The files resulting from the split are named after the sections. The archive cookie is stored in a file named cookie. If section is specified (see -u option below), only the named section is copied.
The create subcommand requires root privileges.
The options for each subcommand are described below.
The options for the create subcommand are below. Many of these options supply values for keywords in the identification section of a file containing a flash archive. See flash_archive(4) for a description of these keywords.
author is used to provide an author name for the archive identification section of the new flash archive. If you do not specify -a, no author name is included in the identification section.
Create a differential flash archive by comparing a new system image (see DESCRIPTION) with the image specified by the system_image argument. By default, the new system image is the currently running system. You can change the default with the -R option, described below. system_image is a directory containing an image. It can be accessible through UFS, NFS, or lumount(1M).
The rules for inclusion and exclusion of files in a differential archive are described in DESCRIPTION. You can modify the effect of these rules with the use of the -x, -X, -y, and -z options, described below.
This option is not supported when creating a flash archive of a ZFS root pool.
Compress the archive using compress(1)
Exclude specified datasets from the flash archive. This option can be used multiple times to exclude multiple datasets.
Use the contents of filelist as a list of files to include in the archive. The files are included in addition to the normal file list, unless -F is specified (see below). If filelist is -, the list is taken from standard input. filelist can include directories. If a directory is listed, all files and subdirectories under that directory are included.
This option is not supported when creating a flash archive of a ZFS root pool.
The description to be included in the archive as the value of the content_description archive identification key. This option is incompatible with -E.
The description to be used as the value of the archive identification content_description key is retrieved from the file descr_file. This option is incompatible with -e.
Include only files in the list specified by -f. This option makes -f filelist an absolute list, rather than a list that is appended to the normal file list.
This option is not supported when creating a flash archive of a ZFS root pool.
Do not generate hash identifier.
Ignore integrity check. To prevent you from excluding important system files from an archive, flar runs an integrity check. This check examines all files registered in a system package database and stops archive creation if any of them are excluded. Use this option to override this integrity check.
This option is not supported when creating a flash archive of a ZFS root pool.
By default, the value for the creation_date field in the identification section is generated automatically, based on the current system time and date. If you specify the -i option, date is used instead. date is in the format YYYYMMDDhhmmss, so that, for example:
20060905031000
...stands for 3:10 AM on September 5, 2006.
By default, the value for the files_archived_method field in the identification section is pax(1). If you specify -L, the archiver (cpio(1) and pax) is used instead.
By default, the value for the creation_master field in the identification section is the name of the system on which you run flarcreate, as reported by uname -n. If you specify -m, master is used instead.
Used only when you are creating a differential flash archive. When creating a differential archive, flar creates a long list of the files in the system that remain the same, are changed, and are to be deleted on clone systems. This list is stored in the manifest section of the archive (see flash_archive(4)). When the differential archive is deployed, the flash software uses this list to perform a file-by-file check, ensuring the integrity of the clone system. Use of this option to avoids such a check and saves the space used by the manifest section in a differential archive. However, you must weigh the savings in time and disk space against the loss of an integrity check upon deployment. Because of this loss, use of this option is not recommended.
This option is not supported when creating a flash archive of a ZFS root pool.
name is supplied as the value of the content_name keyword. See flash_archive(4).
The -n name option, with name argument, is required for the create subcommand.
Create the archive from the file system tree rooted at root. If you do not specify this option, flar creates an archive from a file system rooted at /. When creating a differential flash archive, the system image specified by -R replaces the currently running system as the new image. See DESCRIPTION.
Note - The root file system of any non-global zones must not be referenced with the -R option. Doing so might damage the global zone's file system, might compromise the security of the global zone, and might damage the non-global zone's file system. See zones(5).
Skip the disk space check and do not write archive size data to the archive. Without -S, flar builds a compressed archive in memory before writing the archive to disk, to ensure you have sufficient disk space. Use -S to skip this step. The result of the use of -S is a significant decrease in the time it takes to create an archive.
Content type included in the archive as the value of the content_type archive identification key. If you do not specify -T, the content_type keyword is not included.
Include the user-defined keyword(s) and values in the archive identification section. See flash_archive(4).
Include the user-defined section located in the file section in the archive. section must be a blank-separated list of section names as described in flash_archive(4).
Exclude the file or directory exclude from the archive. Note that the exclude file or directory is assumed to be relative to the alternate root specified using -R. If the parent directory of the file exclude is included with the -y option (see -y include), then only the specific file or directory specified by exclude is excluded. Conversely, if the parent directory of an included file is specified for exclusion, then only the file include is included. For example, if you specify:
-x /a -y /a/b
all of /a except for /a/b is excluded. If you specify:
-y /a -x /a/b
all of /a except for /a/b is included.
This option is not supported when creating a flash archive of a ZFS root pool.
Include the file or directory include in the archive. Note that the exclude file or directory is assumed to be relative to the alternate root specified using -R. See the description of the -x option, above, for a description of the interaction of the -x and -y options.
This option is not supported when creating a flash archive of a ZFS root pool.
Use the contents of filelist as a list of files to exclude from the archive. If filelist is –, the list is taken from standard input.
This option is not supported when creating a flash archive of a ZFS root pool.
filelist is a list of files prefixed with a plus (+) or minus (-). A plus indicates that a file should be included in the archive; the minus indicates exclusion. If filelist is –, the list is taken from standard input.
This option is not supported when creating a flash archive of a ZFS root pool.
The options for flar info subcommand are as follows:
Only the value of the keyword keyword is returned.
List all files in the archive. Does not process content from any sections other than the archive section.
The following are flar info options used with tape archives:
The block size to be used when creating the archive. If not specified, a default block size of 64K is used.
Specifies the position on the tape device where the archive should be created. If not specified, the current position of the tape device is examined.
The archive to be analyzed is located on a tape device. The path to the device is specified by archive (see OPERANDS).
The options for flar split and combine (split and combine archives) subcommands are as follows:
Retrieve sections from dir, rather than from the current directory.
(Used with split only.) Extract the archive section into directory called archive, rather than placing it in a file of the same name as the section.
(Used with split only.) Extract only the section named section from the archive.
Appends section to the list of sections to be included. The default list includes the cookie, identification, and archive sections. section can be a single section name or a space-separated list of section names.
The following options are used with tape archives (with both split and combine):
The block size to be used when creating the archive. If not specified, a default block size of 64K is used.
Used only with -t. Specifies the position on the tape device where the archive should be created. If not specified, the current position of the tape device is used.
Create an archive on or read an archive from a tape device. The archive operand (see OPERANDS) is assumed to be the name of the tape device.
The following operand is supported:
Path to tape device if the -t option was used. Otherwise, the complete path name of a flash archive. By convention, a file containing a flash archive has a file extension of .flar.
Example 1 Creating a Flash Archive
The command below creates a flash archive named pogoS9 and stores it in /export/home/archives/s9fcs.flar. The currently running system is the basis for the new archive.
# flar create -n pogoS9 /export/home/archives/s9fcs.flar
Example 2 Creating Differential Flash Archives
The command below creates a differential flash archive.
# flar create -n diff_pogoS9 -A /images \ /export/home/archives/diff_s9fcs.flar
In the following example the old system image is accessed through lumount.
# lumount s9BE /test
# flar create -n diff_pogoS9 -A /test \
/export/home/archives/diff_s9fcs.flarThe following example shows the use of the -R option to specify a new system image other than the currently running system.
# flar create -n diff_pogoS9 -R /test \ -A /images /export/home/archives/diff_s9fcs.flar
Note the caveat on the use of the -R option in the description of that option, above.
The following exit values are returned for the create, split, and combine subcommands:
Successful completion.
An error occurred.
The following exit values are returned for the info subcommand:
Successful completion.
Command failed. If the -k option is used and the requested keyword is not found, flar returns 2.
See attributes(5) for descriptions of the following attributes:
|
cpio(1), pax(1), flarcreate(1M), flash_archive(4), attributes(5)
|