Network Backup Version 10.0 Savepack Guide
Contents
|
What To Backup
The Backup Administrator defines "What To Backup" by adding one or several servers, workstations, laptops, virtual machines, directories or files to an Arkeia object called a Savepack.
A Savepack can also include databases, application and directory service data and other Savepacks as required by the backup policy.
Savepack List of Trees
Defines the list of Arkeia Clients, hostname and or data paths included in the Savepack for backups.
By default, each unique hostname (client machine) in the Savepack will process with a single Arkeia Flow (data stream).
The Scheduled and Immediate Backup jobs "Parallelism" option can be set to limit the amount of Flows (data stream) per backup job regardless of the quantity of unique hostnames (clients) included in the Savepack.
Name
Name
A name given to the Savepack by the end user. Savepack Names should be 2 to 24 characters long.
Savepack names should not include non-alphanumeric characters such as ' " } / \ | # $.
Compression / Deduplication Type
Compression / Deduplication Type
Defines the type of Compression or Data Deduplication that will be applied to the data included in the Savepack.
Arkeia includes the Lempel Ziv compression scheme to obtain better performance on large files.
The Compression and Deduplication options are:
Compression (Optimize for Speed) ( Default Setting )
The default configuration selection what uses the LZ1 compression algorithm and provides the fastest compression speeds but lower compression multiplier.
The compression of data occurs on the client machine before the data transverses the network.
Compression (Optimize for Shrink)
Uses the LZ3_3 compression algorithm that provides a higher compression multiplier but slower backup speeds.
The compression of data occurs on the client machine before the data transverses the network.
Compression (Balance speed/shrink)
When this setting is used, Arkeia determines what compression method (LZ1 or LZ3_3) to be used based on the data type.
Deduplication + Compression (at the target)
Data Deduplication and local compression of duplicated blocks are performed on the target Media Server.
The data streams from the clients machines are transferred uncompressed over the network.
Deduplication + Compression (at the source)
Data Deduplication and Compression of duplicated blocks are performed on the source client machine.
Only the unknown deduped blocks are passed to the backup server. Known blocks are discarded on the client machine.
No Compression
No compression or Data Deduplication is used
- Please refer to the Data Deduplication and Compression Guide for more details..
Encryption
Encryption Type Allowed
Defines the type of encryption that will be applied to the files and directories included in the Savepack.
The encryption of data occurs on the client machine before the data transverses the network.
The encryption options are:
AES
Advanced 256 bit Encryption Standard
No Encryption
(default) No encryption is used for backups
- Please refer to the Encryption Guide for more details and how to configure encryption.
Savepack Properties
Filter Options
Find Filter
Find Filter
Standard unix "find filter" syntax can be used to include or exclude files.
Although only files specified in the filter will be backed up, all files must pass through the filter.
EXAMPLES:
Only backup files larger than 1000 KB
Find Filter: -type d -o -size +1000
Only backup files smaller than 1000 KB
Find Filter: -type d -o -size -1000
Only backup files with rpm in the name
Find Filter: -name "*rpm*"
Only backup files with atime greater than 30 days
Find Filter: -atime +30
Only backup files with a ctime smaller than 30 days
Find Filter: -ctime -30
Inclusion Filter
Inclusion Filter
Regular Expressions can be used to include files from the backup job.
The Inclusion Filter supports file level and does not support directories.
- Please refer to the Exclusion/Inclusion additional examples page for more advise.
- Please refer to the Verify your Inclusion/Exclusion filter page to know how use the LOG_FILTER setting.
Exclusion Filter
Exclusion Filter
Regular Expressions can be used to exclude files from the backup job.
The Exclusion Filter works at the file level and is not a solution for excluding full directories from the backup job.
The "Inclusion Filter" takes precedence over the "Exclusion Filter".
If there are files that meet both filter criteria, the files will be processed as part of the "Inclusion Filter".
EXAMPLES:
Include/Exclude files containing 2005,2006 and 2007
/[^/]*2005|2006|2007[^/]*$
Include/Exclude files ending with .mp3, .rpm and .doc
\.mp3|\.rpm|\.doc$
Include/Exclude files who's name begins with User, user, Users and users
/User|user|Users|users[^/]*$
Include/Exclude files ending with .avi, vmdk, .mp3, .mpg and files name containing userclass and ntuser.
(\.avi|\.vmdk|\.mp3|\.mpg$)|(/[^/]*userclass|ntuser[^/]*$)
Include/Exclude files names containing USERCLASS and NTUSER
(/[^/]*NTUSER|USRCLASS[^/]*$)
- Please refer to the Exclusion/Inclusion additional examples page for more advise.
- Please refer to the Verify your Inclusion/Exclusion filter page to know how use the LOG_FILTER setting.
Pre and Post Command Options
Command Before and After Savepack
Command Before Savepack
The execution of commands and scripts on the client before the backup of the Savepack.
CAUTION: Command Before Savepack
Arkeia will not perform the backup if the command entered in the Command before Savepack field fails and returns a standard error code. For instance, if a tar command should be performed on some files before the backup, and tar returns a standard error code (an error code larger than 0 -- Success, then Arkeia will not perform the backup.
If the command did not perform as expected and returned an error code, but the backup should be performed nonetheless, click on the check box Backup Savepack if command fails below the entry field.
A command can be any standard UNIX command, a shell program, a batch file (on Microsoft® Windows® computers) or any other program that can be launched from a command line. The following commands are valid entries for a Command before Savepack:
EXAMPLES:
Perl script on a Linux machine:
linux.domain.com!file:/home/john/scripts/dump.pl Batch script on a Windows machine:
windows.domain.com!file:c:/Program Files/Arkeia/Arkeia/myscript.bat
Command After Savepack
In the same way that a command can be entered to be executed before the backup of the Savepack, another command can be entered to be executed after the backup itself (see above for more details). To allow the execution of this command even if the backup itself has failed, click on the Execute if Savepack backup fails check box below the entry field.
Comment
Comment
Enter optional comments about this Savepack.
Platform Specific Options
Windows 2003 and Later
Enable VSS Snapshot Based Backups
Enable VSS Snapshot Based Backups
Choose Yes if you want the backup job to initiate Microsoft Windows Volume Shadow Copy based backups.
The default is YES
The Volume Shadow Copy Service provides the backup infrastructure for Microsoft Windows Server 2003 operating systems, as well as a mechanism for creating consistent point-in-time copies of data known as shadow copies. Previous to the Volume Shadow Copy Service, there was no standard way to produce clean (uncorrupted) snapshots of a volume. Snapshots often contained corruptions due to torn writes that required the use of utilities such as Chkdsk.exe to repair. Torn writes occur when an unplanned event (such as a power failure) prevents the system from completely writing a block of data to disk. The Volume Shadow Copy Service APIs prevent torn writes by enabling applications to flush partially committed data from memory.
Arkeia uses this API to manage VSS snapshots (on a per volume basis) and perform the backup of Windows file systems through a snapshot, instead of backing up live file systems. If a backup is started with the this option set to YES, the Arkeia Filesystem Agent (arkfs) requests the creation of a VSS snapshot for each volume to backup. The backup is performed through the snapshot instead of the live file system.
The snapshot object itself is not backed up. The contents that reside in the snapshot is backed up.
The files that reside on the file system are backed up. Since they are accessed through a snapshot, they are guaranteed to be closed and in a consistent state so there is no more "open file" problem. When the backup has finished, the snapshots are destroyed and not saved to the local filesystem. Arkeia creates one snapshot per volume during a backup which can stay on the system for up to five minutes after the end of a backup.
Unix and Linux
Backup POSIX ACLs
Backup POSIX ACLs
Choose Yes if you want the file ACLs (Access Control Lists) in this tree to be saved on backup. The value No means that Arkeia will ignore any ACLs when backing up this Savepack. Note that the setting of Backup POSIX ACLs in Tree advanced options can override the setting you enter here. Setting POSIX ACL backup to Yes increases the processing overhead for every file, whether or not the file uses ACLs.
The default is NO.
Linux Extended Attributes
Backup Linux Extended Attributes
Choose Yes if you want the Linux extended attributes in this tree to be saved on backup. The value No means that Arkeia will ignore any Linux extended attributes when backing up this Savepack. Note that the setting of Linux extended attributes in Tree advanced options can override the setting you enter here.
The default is NO.
Linux extended file attributes is a file system feature that enables users to associate computer files with metadata not interpreted by the file system, whereas regular attributes have a purpose strictly defined by the file system (such as permissions or records of creation and modification times). In Linux, the ext2, ext3, ext4, JFS, ReiserFS and XFS file systems support extended attributes (abbreviated xattr) if the libattr feature is enabled in the kernel configuration.
Arkeia supports the backup of NSS information (Metadata and trustees)on OES2 with this feature. Novell has implemented a way to push the extended attributes into the standard Linux xattr framework so that all of this specialized metadata can be preserved by standard Linux xattr APIs.
The following lines must be added to the nssstart.cfg file:
vi /etc/opt/novell/nss/nssstart.cfg
/ListXattrNWMetadata /CtimeIsMetadataModTime
VMware vSphere Options
Use Changed Block Tracking Where Available
Change Block Tracking (CBT) is a VMkernel feature that keeps track of the storage blocks of virtual machines as they change over time.
The VMkernel keeps track of block changes on virtual machines, which enhances the backup process for applications that have been developed with VMware’s vStorage APIs. The VMkernel’s Change Block Tracking (CBT) enhances VMware’s Storage VMotion because the process no longer consumes double the amount of resources per virtual machine.
The default is YES.
Quiesce Quest File System
VMware's vStorage API for Data Protection (VADP) supports the Quiecing of VMware Supported Windows File Systems and applications which allows to take a Shadow Copy Service Snapshot (VSS) of the Volumes and Application inside of the Virtual Machine.
The default is YES.
Transport Mode Priority List
When backing up from a Hypervisor, the need for a LAN connection between the backup machine and the hypervisor is required even if you're using a LAN-free transport mode (SAN or Hot-Add).
When backing up from a vCenter (several hypervisors seen as one), the backup client needs to connect to the vCenter through the LAN, but also needs to be able to make a direct connection to the hypervisors hosting the virtual machines to backup. If the NBD (LAN) or NBDSSL (LAN) transport modes are used, the data will not be transferred through the vCenter API which would make the transfer of data extremely slow.
This does not apply to SAN and HotAdd modes.
SAN - transfer the data over the SAN HOTADD - the agent on a VM gets data from a shared storage NBD - transfer the data over the network NBDSSL - encrypt and transfer the data over the network
File System Options
Follow Symbolic Links
Follow Symbolic Links
Defines if a backup job will follow the symbolic link and backup the source file.
The default is NO.
- Click Here for more information and examples of Following Symbolic Links.
Follow File Systems
Follow File Systems
Defines if a client process will follow the file system and backup mounted file systems within the selected backup path.
The default is YES.
File Systems Allowed
File Systems Allowed
Defines which file systems will be backed up during the backup process.
The file system allowed options are:
Local File Systems (Default)
Backup locally mounted filesystems such as ntfs, fat, fat32, ext2, ext3, ext4, reiserfs, xfs, jfs, nss, btrfs, etc..
All File Systems
Backup all files systems (including networked filesystems) and non standard file systems such as cluster, luster, etc...
All Except Networked File Systems
Backup all file systems except network mounted file systems.
Reset Access Times
Reset Access Times
Defines whether or not the access times for the files in this Savepack are changed as a result of backup or restore operations.
The default is NO
Backup Excluded Tree
Allows the backup of directories that have been Excluded from the backup jobs.
The Default is NO
By default, file system directories that have been Simply Excluded are skipped and logged has containing a .ARK_NOBACKUP file.
[1] Backing up linux.us.arkeia.com!file:/tmp/test [1] linux.us.arkeia.com: Directory `/tmp/test' not backed up : contains file .ARK_NOBACKUP
When the Savepack Advanced Option "Backup Excluded Tree" is set to YES, Arkeia will disregard the .ARK_NOBACKUP file and backup the contents of the directory and all child directories.
[1] Backing up linux.us.arkeia.com!file:/tmp/test [1] Backup of "linux.us.arkeia.com!file:/tmp/test" OK, "4" files, "48" MB, "1" seconds, "2880" MB/mn, "0" warnings
Advanced Options
Media Server
Media Server
When a backup job using this Savepack is started, the resulting backup dataset from the clients will be sent to this media server for storage.
Click the button to see a drop down list of available media servers and select the Media Server that will host the stored data.
The default Media Server is the backup server.
- Please refer to the Remote Storage (LAN/WAN) Media Server Guide for more details about the Media Server feature.
Number Of Retries
Number Of Retries
The Number of retries defines how many times Arkeia will try a connection to a client at the start of the backup and during the backup when a connection error occurs.
The default value was 3 retries every 15 minutes until the version 8.2 while it becomes ZERO retry starting from version 9.0.
Savepack Creation
Backup > What To Backup
Click "Add A Savepack"
Step 1: Create Savepack and Define Options
Enter a name for the Savepack
Select the Savepack Compression or Deduplication Type from the drop down list Default = Compression (Optimize for speed)
Select the Data Encryption Type from the drop down list if needed Default = No Encryption
Modify other Savepack features as needed by clicking on the name to expand the options.
- Filters Options
Find, Inclusion and Exclusion Filters No filters are set by default
- Pre and Post Commands Options
* Command Before Savepack * Backup Savepack if Command Fails * Command After Savepack * Execute if Savepack Backup Fails No commands are set by default
- Platform specific options
Windows 2003 and Later * Enable VSS snapshot-based backups Default = YES Unix/Linux Options * Backup POSIX ACLs Default = NO * Backup Linux extended attributes Default = NO * VMware vSphere Options * Use Changed Block Tracking where available Default = YES * Quiesce guest file system Default = YES
- File system options
* Follow File Systems Default = NO * File Systems Allowed Default = Local File Systems * Reset access times Default = NO * Backup excluded Tree Default = NO
- Advanced options
* Media Server Default = backup server * Number of Retries Default = 3
Click "Next"
Step 2: Add Trees to Your Savepack
The Savepack "Add Tree" Function is split into five tabbed sections to support Files and Folders, Savepacks, Objects, Volume Images and to View and Delete Selected trees.
Add Files and Folders
The name of computer, or hostname (to backup an entire computer), or a directory name to backup one specific directory and all the files and sub-directories it contains, or a single file.Files and Folders are also used to backup Agent Options such as MySQL, PostgreSQL, Lotus, vMware, etc...
"Add File and Folders" is the default Savepack Tab.
Add Savepacks
A Parent Savepack can include Child Savepacks that contains trees which will be called at the time of backup.
Add Objects
An Object will back up and restore the result of a command.
This command should be entered in the Object backup command field. Please note that the file created by the Object command will not reside on the Arkeia client, but an entry will be created in the Arkeia database to restore the data backed up.
Add Volume Images
A Volume Image is either an entire disk or an entire partition used to accomplish a raw disk backup.
If a Image tree is to be used, the Browser should be used to select the disk or the partition present in the /dev directory of the UNIX computer, or the drive designation in Windows of the partition or HDD.
A Volume Image disk backup is reported as a single file equivalent to the partition size.
Trees Selected
Displays the list and paths of all trees currently selected in the Savepack.
Trees can be unselected/deleted from this screen.
Select the entire hostname name or expand the hostname to select specific data structures which are to be included for backups in the Savepack. As trees are selected for backup, the "tree selected" value in the upper right hand corner will increase
Click on "Trees Selected" to view, verify and delete selected trees.
Click on theto delete a specific tree as needed or exit the "Trees Selected" screen.
Click "Next"
Step 3: Configure Trees Options for Savepack
A Tree is a directory, file or folder on a client computer's HDD that's included in a Savepack.
Trees can have the same options as Savepacks. The defaults for these common options the same as the parent Savepack.
A change made to an option in a Tree will override the same option in the Savepack but only for that tree.
This section will focus on the unique Tree options that are not part of the parent Savepack options.
Select a tree to view it's options.
Click on "Trees Management Options" to expand and view the options
Modify the options as needed such as assigning a Multiflow Value
Click "Apply" to save the changes to the tree
Click on "Configuration Details" to review the recent changes
Un-Select the tree to get out of edit mode
Select another tree to modify if needed and repeat the editing process
Click "Finish" to create the Savepack and save all modifications
Review the Savepack created
Savepack Management
Add Trees and Clients
Clients and File System paths can be added to the Savepack at anytime.
Select a Savepack to Modify
Click "Modify" <Savepack Name>
Click "Next"
Add trees or clients as needed As trees are selected for backup, the "trees selected" value in the upper right hand corner will increase
Click on "trees selected" to view the client trees selected in this savepack Trees can also be deleted from this screen
Click "Apply and Exit" to save the changes or "Next" to modify specific Tree Options.
Verify and View the changes
Delete Trees and Clients
Clients and File System paths can be deleted to the Savepack at anytime.
Select a Savepack to Modify
Click "Modify" <Savepack Name>
Click "Next"
Click on "trees selected" to view the client trees selected in this savepack Trees must be deleted using this screen
Click theto delete select trees
Click "Apply and Exit" to save the changes or "Next" to modify specific Tree Options.
Verify and View the changes
Adding A Savepack To A Savepack
It is possible to incorporate another Savepack into an existing Savepack.
This allows complex configurations to be backed up in a simple configuration. Since each Savepack have its own configuration and parameters, modifying the parameters of one does not mean modifying the every Savepack in the backup environment.
Backup > What To Backup
Click "Add A Savepack"
Enter a Name for the Savepack
Click "Next"
Click the "Add A Savepack Tab"
Select Savepacks as needed As Savepacks are selected for backup, the "tree selected" value in the upper right hand corner will increase
Click "trees selected" to view Savepacks that are selected. Savepack Trees can also be deleted from this screen.
Click "Next"
Modify individual Savepack (Savepack Tree) Options as needed
Click "Finish"
Verify and View the changes Selecting the Savepack at the Main Savepack Screen will display the list of Savepacks in the Savepack. Normal Savepacks list will have client hostnames and tree paths.
Delete A Savepack
Deleting a Savepack from the Arkeia Software does not:
- Remove the client machine from the index
- Recycle any backup data
- Delete or purge the Arkeia Index
Select a Savepack to delete
Click "Delete" <Savepack Name>
Click "OK" to confirm the deletion of the Savepack
Savepack Tree Options
A Tree is a directory, file or folder on a client computer's HDD that's included in a Savepack.
A Tree can also be another Savepack inside of an existing Savepack.
Trees can have the same options as Savepacks. The defaults for these common options are as for the parent Savepack.
A change made to an option in a Tree will override the same option in the Savepack but only for that tree.
This section will focus on the unique Tree options that are not part of the parent Savepack option.
Options
Deduplication and Compression
The Deduplication and Compression options of the Savepack Tree are inherited from the Parent Savepack by default.
The Savepack Tree options can be modified to be independent of the Savepack itself.
- Please refer to the Compression and Deduplication Section for more details about this feature.
Encryption
The Encryption options of the Savepack Tree is inherited from the Parent Savepack by default.
The Savepack Tree options can be modified to be independent of the Savepack.
- Please refer to the Encryption Section for more details about this feature.
Tree Management Options
Multiflow
Multiflow
The Multiflow value determines if Arkeia should launch a backup tree with several concurrent data flows.
By default, the value of Multiflow is "0" for all trees, meaning that Arkeia will use only one data flow between the server and the client and will backup the trees in a sequential order unless every tree contains a different hostname such as backing up many client machines.
In the example below, each client will be processed by a dedicated data stream (Flow) in parallel.
Same Hostname
To create more than one data flow (and, therefore, launch parallel backups) when many trees from the same client are configured in the Savepack, a value other than 0 should be entered in the MutliFlow field.
In the example below, four trees are created in a Savepack, and each of these trees has been assigned a different flow number. The backup job will process with four different flows to the computer, one for each of the trees, and backup all the trees in parallel. If two of these trees are assigned the same Multiflow number, they will be backed up sequentially instead of in parallel.
EXAMPLE: Multiflow 1 backups.us.arkeia.com!file:/bin Multiflow 2 backups.us.arkeia.com!file:/boot Multiflow 3 backups.us.arkeia.com!file:/etc Multiflow 4 backups.us.arkeia.com!file:/home
Different Hostname
The default Multiflow rule is "1 flow per hostname" in the Savepack regardless of the Multiflow default setting of "0".
In the example below, all client machines will be processed using a dedicated data steam in parallel.
Priority
Priority
Defines the order in which trees should be processed The possible values for Priority range from 1 (highest priority) to 100 (lowest).
Trees that have a higher Priority value will be backed up before the trees that have a lower priority. The default value for all trees is 50.
The priority level is only taken into account when trees have the same Multiflow number. (default = 0 for all trees)
In the example below, the trees will process sequentially in the order of Priority.
Click "Configuration Details" to access the tree details screen.
- Please refer to How To: Configure Priorities and Chains for more details about this feature.
Chaining
Chain
The “Chain” field is used to make dependencies between trees from different or same client machines.
A chain is created between two trees when you put the same value in the chain field.
You also have to set the “priority” level for each chained tree to control the backup order
The default setting for all trees is “0” (no chaining).
In the example below:
- The bin directory will backup first because it's Priority #1 and Chain #1
- The boot directory will backup second because it's Priority #2 and Chain #1
- The etc directory will backup third because it's Priority #3 and Chain #2
- The home directory will backup fourth because it's Priority #4 and Chain #2
- Please efer to How To: Configure Priorities and Chains for more details about this feature.
Filters Options
The Tree Filter Options support the same features as the Savepack.
Pre and Post Commands Options
The Pre and Post Commands Options support the same features as the Savepack.
Platform Specific Options
The Platform Specific Options support the same features as the Savepack except for the Linux Disaster Recovery option under "Unix/Linux.
Check this box if you want to include the data required to perform a Disaster Recovery included in the Linux Clients !sysinfo: plugin.
A separate License is required for this feature.
File System Options
The File System Options support the same features as the Savepack.
Advanced Options
The Advanced Options support the same features as the Savepack.