Topics | How To | Related Topics
Discovery Filters for Lotus Notes iDataAgents
Filters allow you to exclude specified data (messages, files, folders/directories, databases, etc.) from data protection and recovery operations. This is useful for reducing the amount of data that is protected and recovered, in order to optimize performance and minimize storage space. The following types of filters can be used to achieve these goals:
Data Protection Filters
Data Protection Filters exclude data from various types of data protection operations including archive, backup, discovery, and replication. Listed below are the types of available data protection filters, with a link for more information on each.
For an in-depth discussion on various aspects of using data protection filters and for a list of frequently asked questions, refer to Content, Filters, and Regular Expressions.
Data Recovery Filters
Data Recovery Filters exclude data from restore and recovery operations, so that you only recover the data that you need. For more information, see Filter Data From Restore/Recover Operations.
Compliance archiving filters allow you to exclude data from archive operations by specifying which Journal mailbox folders or folder patterns that you don't want to be included in archive jobs. Filters can be established globally for all Exchange Compliance Archiver clients in a CommCell through the use of Global Filters, or on a per-subclient basis through compliance archiving Filters. Wildcards and regular expressions are also supported for filtering mailbox folder patterns. For information on which wildcards/regular expressions are supported for filtering data on this agent, see Wildcards. For step-by-step instructions, see Add/Edit a Data Protection Filter for a Subclient.
Backup filters provide the means to exclude data from your backups. For agents that support this feature, each subclient has a backup filter that is created automatically by the system. Initially, the filter is empty, which means that the subclient backs up all the data it contains. The system does not allow you to add filter entries that are not content of a particular subclient to that subclient's filter.
For example, when you create a ProxyHost subclient, you enter data paths (e.g., C:\Data) as subclient content. You may want to exclude files or subdirectories that are contained within a subclient path from backups. If you want to back up C:\Data, but exclude C:\Data\Tmp, add C:\Data\Tmp to the backup filter. Similarly, data could be excluded from the default subclient for a Windows File System iDataAgent.
Generally, you can access the backup filter from the Filters tab of the Subclient Properties dialog box.
For Unix File System and ProxyHost on Unix iDataAgents, the directories listed below are automatically filtered so that they are not included in backups:
Also, for Unix File System iDataAgents, a maximum of 1024 characters can be included in the filter path.
Windows and NetWare clients each contain a swap file (i.e., pagefile.sys on Windows and _swap_.mem on NetWare) that is automatically filtered out by the system. These file are large and contain no data that would be useful for restore purposes. To ensure that the system does not attempt to back them up, backup filters for Windows File System and NetWare iDataAgents are preconfigured to include their respective swap file.
When configuring NDS subclients in particular, it is important to consider the use of backup filters. Filtering data that is replicated across NetWare servers can enhance that efficiency of data protection and recovery operations.
|
Job Results files cannot be filtered from backup operations. |
The backup filters for each physical node will automatically exclude all shared disks that reside on all virtual servers. This ensures that a physical node data protection operations contain only copies of files that reside locally on that server, and that it does not contain copies of files from the virtual servers shared disk.
The backup filter for the virtual server will automatically exclude disks that reside on the physical nodes. This will ensure that data protection operations for the virtual server contain only copies of files that reside on the shared disk.
| When the virtual server is running the Exchange Database and/or Exchange Mailbox iDataAgent, as well as the File System iDataAgent, the shared disk contains the Exchange directories DSDATA and mdbdata. It is recommended that you exclude these directories from the File System backup, by creating filters for any subclients whose contents include the Microsoft Exchange Server installation directory. |
Windows System State components can be filtered out of Windows File System iDataAgent backup jobs. For more information and a list of components that can be filtered, refer to Excluding System State Files from Backup.
For AIX and Linux File System iDataAgents that support the VERITAS Cluster File System (VxCFS), you can use the sClusteredFSFilter and the sClusteredFSTypes registry keys to filter non-clustered and clustered file systems from your backups. These registry settings can be used to prevent backing up multiple copies. Without this feature enabled in a clustered environment, if "/" is backed up, all the clustered file systems under "/" are backed up from the physical machine as well as from the virtual machine. With this feature enabled in a clustered environment, if "/" is backed up from a physical machine, only non-clustered file systems are backed up; but if "/" is backed up from a virtual machine, only clustered file systems are backed up.
The tmpfs, ctfs, proc, devfs and lofs (loopback) file systems are automatically skipped during backups by the Solaris File System iDataAgent. The filters for these file systems and others can be altered by using the ignoreFStype registry key. The addition of the lofs file system as a skipped file system has an impact on the Solaris Global Zone with regard to filesystem submount points that might previously had been expected to be backed up (e.g., if the root of a zone is an lofs file system and it has the ufs, zfs, or vxfs submount file system). These submount file systems will be skipped during backup unless they are added to the subclient content.
For subclients of NAS iDataAgents, note the following backup filter behavior:
Filters can be applied to the content (URLs of site collections) of a subclient for database backup sets.
URLs of site collections can be excluded from data protection operations of a subclient. For example, specifying http://sptest/sites/* will exclude all site collections starting with http://sptest/sites. The use of wildcard expressions when adding entries to this exclusion list are supported. See Wildcards for more information.
Exception filters can also be specified. However, wildcards are not supported in Exception filters.
Instead of a backup filter, the default subclient for a Lotus Notes client has a discovery filter option. Discovery filters allow you to prevent database files that match specific criteria from being discovered when the default subclient is backed up. Each default subclient has a discovery filter that is created automatically by the system. Initially, the filter is empty, which means that the default subclient discovers all new databases at backup time. These filters do not affect manual discovery operations. If you manually discover databases, you see an unfiltered list of all of the databases on the server.
The discovery filters are flexible and allow you to exclude paths and file names that match certain patterns. These paths are always relative to the Domino data path for each partition and always start with a backslash (e.g., \mail\user.nsf). Note that when you change a discovery filter, the change is effective the next time the subclient is backed up. Lotus Notes transaction log and user-defined subclients do not have backup or discovery filters.
Each Migration Archiver Agent subclient has a migration archiving filter. These filters allow you to specify the data (messages, files and folders/directories) you want to exclude from a migration archiving operation. The filters are flexible and allow you to exclude mailbox names and file or folder/directory names that match certain patterns.
The filter is initially empty for the Exchange Mailbox/Public Folder Archiver Agents, which means that the subclient archives all the data (e.g., mailboxes/public folders) that is specified in the subclient content.
The File Archiver Agents, by default, exclude from migration archiving certain file types that are essential to the normal functioning of file systems. For Windows and NAS configurations, these include *.dll, *.bat, *.exe, *.cur, *.ico, *.lnk. For Unix, these include *.a, *.ksh, *.csh, *.sh, *.lib, *.so. For NetWare, these include *.nlm, *.bat, *.exe, *.cur, *.ico, *.lnk. Although you can remove these entries from the exclusion filter, it is not recommended. In addition, files in the following folders/directories cannot be archived and are filtered out automatically by the system:
|
For agents that support Data Classification, the subclient properties in the CommCell Console does not include a Filters tab. Nonetheless, some data protection filter-like capabilities are available as archiving rules from the Rules tab for these agents. See Migration Archiving - File Archiver Agents for an overview. CommCell Application files/folders/directories, disk mount path folders/directories, and the CommServe database should not be archived by the File Archiver Agents and must be manually filtered out. |
See Also:
Exclusion Filters are supported at the Replication Set level. An Exclusion Filter allows you to define files and folders to be excluded from replication activities, for all Replication Pairs in the Replication Set. For example, you can "exclude" the C:\Temp directory. Filtering is configured in the Filters tab of the Replication Set Properties.
| The filter string cannot contain more than 255 characters, if a single filter string exceeds 255 characters then this filter will not be honored during replication/restores. |
For CDR on Windows, note the following when adding, deleting, or editing Filters for a Replication Set:
For CDR on UNIX, note the following when adding, deleting, or editing Filters for a Replication Set:
If the newly configured filtering excludes files or folders that had previously been replicated, these files and folders on the destination(s) will be handled according to your Orphan Files settings.
For instructions on how to Abort and Restart replication, see Start/Suspend/Resume/Abort Data Replication Activity. The use of wildcards or regular expressions is supported in defining Exclusion Filters. For a list of supported wildcards, see Wildcards - File System.
The following reports provide an option to display the exclusion and exception filters that were used in client data protection operations on the CommCell:
This information is useful in determining which client data was not backed up or archived, and can assist in performance tuning.