

General Considerations
General Considerations
You should know the following considerations before using Arcserve UDP:
Upgrade Considerations
Arcserve UDP Agent (Linux) Considerations
Arcserve UDP Agent (Linux) supports only those kernels that are officially published by Linux distributors. Customized kernels are not supported.
Backup Considerations
- When integrating Arcserve UDP Agent (Windows) with Arcserve Replication and High Availability and replicating data, you may discover that a greater quantity of data is replicated from the Master server to the Replica server than the actual file size on the Master server. This behavior is known to occur when using the non-compressed and non-encrypted backup option. You can minimize the effects of this behavior by configuring the scenarios to replicate periodically.
- Starting with Microsoft Exchange Server 2013, all public folders are stored under specially designed public folder mailboxes in the Mailbox database. Therefore, to back up public folders, back up the corresponding Mailbox database.
- When you back up a Microsoft SQL instance from machine A and restore the Microsoft SQL instance to machine B, if the Microsoft SQL instance on either machine A or machine B is in windows authentication, and the database 'master' is included in the restore list, the Microsoft SQL instance on machine B will fail to start after the restore.
- To avoid possible failures (mainly for large backup jobs), the buffering I/O settings have been turned off by default. However, with buffering turned off, there can be a corresponding drop or slowdown in backup speed (I/O throughput). If you are performing small backup jobs and you want to increase job performance, you can manually turn on buffering as follows:
- Locate and open the afstor.ini file (C:\Program Files\Arcserve\Unified Data Protection\Engine\Configuration).
- Change the three buffering settings from 0 (off) to 1 (on):
- BufferingForLocal=0
- BufferingForUSB=0
- BufferingForRemoteFolder=0
Note: This is valid for non-data store destinations with the backup format set to Standard.
- The actions or configurations below may affect the size of the block-level backups because Arcserve UDP Agent (Windows) or host-based VM backup will continue to incrementally back up all changed blocks. This means that blocks that shifted due to those actions or configurations will also be included in the backup, even if no data has changed in the files. As a result, the backup size may increase. This is expected behavior. If you do not want the increased backup size, you can stop any schedules for defragmentation or exclude affected volumes for Arcserve UDP Agent (Windows) (volume exclusion does not apply to host-based VM backup).
- Volume defragmentation
- Volume is NTFS deduplication enabled
- System activities, such as page file or hibernation file IO (Applicable for host-based VM backup only)
- When using multiple backup destinations with the Arcserve UDP Recovery Point View on a single Arcserve UDP Agent (Windows) node, if you change one destination back to the normal Windows view, it will force dismount of all volumes mounted by Explorer Extension. To avoid this problem, you should delay changing views until you have used all destinations that are shown in the Arcserve UDP Recovery Point View.
Note: This is only valid for non-data store destinations.
- When a backup job is running, do not try to shrink the volume being backed up. If you do shrink the volume during the backup, data from this session may not be used for data recovery and a Verify backup is necessary afterwards to ensure the data integrity.
- If you are backing up volumes that were replicated using Arcserve Replication and High Availability, you should verify that the spool has been created on a separate volume and configure your backup settings to exclude the spool volume. This helps to avoid the backing up of unnecessary temp spool data.
- After a Microsoft Exchange 2007 and Microsoft Exchange 2010 database backup, if its path is moved to a different location, restore from the session which is backed up before moving the path will fail.
- If both compression and encryption are disabled, then Arcserve UDP Agent (Windows) can only back up the files in .VHD format; it cannot back up the files in .VHDX format.
Note: This is valid for non-data store destinations with backup format as standard.
- When Arcserve UDP Agent (Windows) backs up a Hyper-V server, the VMs might proceed to the Saved state for a short time when the backup job starts. This may lead the VM in a temporary offline status, and will be back online within a few seconds. To prevent this condition, the VM should meet certain conditions.
Note: For more information, see http://msdn.microsoft.com/en-us/library/dd405549(v=vs.85).aspx.
- Random data corruptions have been observed when backups are written to the following device and if a network connection is lost during backups or merges.
- Iomega StorCenter ix4-200d
- The considerations below are applicable if you are using Arcserve UDP Version 5.0 Update 2 and you upgrade to Arcserve UDP Version 6.0. In Arcserve UDP Version 5.0 Update 2, when an agentless backup is performed on a Hyper-V VM that does not support an online backup, the VM is placed in an offline "saved state" while the snapshot is taken for the VM. After upgrading to Arcserve UDP Version 6.0, the default behavior of the VM while being in this saved state is changed. As a result, consider the following items for Hyper-V agentless backups:
Bare Metal Recovery Considerations
The BMR process cannot create storage spaces.
Data Store Considerations
File Copy Considerations
- File copy does not support copying files from Volume Mount Points and Nameless volumes (volumes which do not have a drive letter). The Arcserve UDP Agent (Windows) only performs file copy on mounted volumes that have an assigned drive letter.
- File Copy to Cloud does not support machines where the name contains Japanese characters.
- The File Copy job and File Restore job cannot be run at the same time. Running both at the same time results in a warning message: Another job is running.
- During the File Copy default configuration, the Windows System protected files/folders are skipped.
- If the File Copy is running from a replicated session and the Replicate source has multiple backup sessions, then the File Copy job runs for each session separately. For example, if there are five backup sessions in the backup destination and you add a Replicate task, then the Replicate task replicates all the sessions in a single job. Now, if you add a File Copy task and the File Copy source is the Replicate destination, then five File Copy job runs to replicate each session.
Gateway or Site Considerations
- When you use the Arcserve UDP Console to manage multiple sites, the email alerts will work only when the servers in all sites and the Console are able to communicate with the same email server.
- Only the resources tab on the Console has site-awareness. The other tabs on the Console display consolidated data for all the sites.
GUI Considerations (Agent)
- If you change the structure of the database in Microsoft Exchange or Microsoft SQL Server (add, delete, or move the database), the GUI is not updated dynamically. The Arcserve UDP Agent (Windows) gets the latest information after the service is restarted.
- If you search a file or folder to restore using Internet Explorer, and there are many files or folders in the search result, Internet Explorer may display the message, A script on this page is causing Internet Explorer to run slowly. If it continues to run, your computer may become unresponsive. Do you want to abort the script? Some scripts can take an excessive amount of time to run, so Internet Explorer prompts you to decide whether you would like to continue running the slow script. To resolve this issue, perform one of the following tasks:
- Modify the registry to change the script time-out value in Internet Explorer for specific client machines. For more information, see http://support.microsoft.com/kb/175500.
- Use another browser and try again.
- The Arcserve UDP Recovery Point View (Agent Explorer Extension) does not support browsing and restoring a volume without a drive letter.
- When using Internet Explorer 10, you may see some misplaced borders or distorted text on the GUI. This has no functional effect on use of the product, but if you would like a better viewing experience, try using Google Chrome or Mozilla Firefox web browsers.
- You can log in to the Linux Backup Server from the Arcserve UDP Console only when you perform a restore.
High Availability Considerations
- After adding Arcserve Replication and High Availability control service, the Arcserve UDP user interface could manage the scenarios, process Replication and High Availability engine remote deployment, register the licenses and generate the reports for that control service.
- Only full system scenarios could be created using the scenario creation wizard in the Arcserve UDP user interface. Use the Replication and High Availability manager user interface to create other types of scenarios.
- Only full system scenarios root directory could be edited in the Arcserve UDP user interface. Use the Replication and High Availability manager user interface to edit other types of scenarios.
- Arcserve HA integration in Arcserve UDP 6.0 supports the scenario operation for Arcserve Replication and High Availability r16.5 SP3 and later.
- In the Arcserve UDP user interface, the Replication and High Availability templates cannot be created, edited, and used to create scenarios.
- The following scenario operations are not supported in the Arcserve UDP user interface:
- Host maintenance
- Restore & rewind
- Offline sync
- Auto configuration
- Configure HTTP Proxy for Control Service
- In the Arcserve UDP user interface, Amazon EC2 is not supported in full system scenarios.
- The feature editing the properties for running scenarios is not supported in the Arcserve UDP user interface.
- The report data on the high availability tab will update every one hour.
Installation/Uninstallation/Remote Deployment Considerations
License Considerations
If after opening the Central License Manager under the Arcserve UDP Help menu, no licenses are shown, then the product is in the 30 day free trial period.
Merge Considerations
- During the merge job:
The Data Protected column of full backup in the Protection Summary section is displayed as the sum of source data size of all to-merge recovery points. Once the merge job is complete, it will be displayed as the size of the newly merged full backup. This applies to all type of destinations.
The Space Occupied column of full backup in the Protection Summary section is displayed as the sum of the occupied disk size of all to-merge recovery points. Once the merge job is done, it will be displayed as the disk size of the newly merged full backup. This applies to only non-deduplication data stores and non-data store destinations. (local disk or network share)
- When a merge job runs, the following operations fail on the session which is in the same backup chain as the merge job.
- Mount recovery point.
- Restore browse and search, if the sessions do not have a catalog.
- The merge job might fail if the following operations are performed before it runs:
- Mount recovery point. Dismount the mounted volumes and trigger a new merge job.
- Restore browse and search, if the sessions do not have a catalog. Do not perform Restore browse and search UI operations for 10 minutes, and then trigger a new merge job.
Note: When the backup destination is a shared folder, you can click the Run a merge job manually now link in the Arcserve UDP Agent (Windows) UI home page Summary section to trigger a new merge job.
Plan Considerations
- Only if your backup task destination is recovery point server, you can add a new replicate task. In other words, if the backup task destination is a local or remote share, you cannot add a new replicate task.
- Replicate tasks support replication between all kinds of data stores except:
- Replication from a data store with Encrypt Data enabled to a data store without Encrypt Data enabled.
- Replication from a data store with Deduplication Data enabled to a data store without Deduplication Data enabled.
Restore Considerations
- If the volume does not have a drive letter, restoring to the original location is not supported. The restore operation is blocked on the UI.
- If a database or storage group is renamed or removed after a backup, restoring to the original location fails. Either restore from the latest recovery point or restore to the disk.
- When restoring the Microsoft Exchange database to a relational database, select the relational database on the machine where the Microsoft Exchange database is backed up. If you want to restore the Microsoft Exchange database to the relational database on another machine, dump the database file onto that machine and attach it to the relational database.
- Restoring a Microsoft Exchange database on to a database residing on a compressed volume or directory is not recommended or supported by Microsoft. This can lead to a dismounted state of the database after restore. As a workaround, move the database path to an uncompressed volume or directory before restore and perform the restore again. For more information see http://support.microsoft.com/kb/327772.
For best practices on NTFS compression in Windows, see http://support.microsoft.com/kb/251186.
- If you change the password of the remote destination, update the password in Settings before submitting any restore jobs.
- When you restore files on an NTFS volume with the encryption attribute, to a non-original backup computer, the file is not accessible. This inaccessibility is a system limitation. Also, note the following limitations:
- For a catalog restore, the job succeeds as normal, the file is restored, however it cannot be accessed.
- For a catalog-less restore, the job fails and the encrypted file is logged as inaccessible.
- If the following conditions exist in a two node CCR cluster (cluster continuous replication), the database restore succeeds but the restore job has a status of "Failed":
- The active node is paused.
- An Arcserve UDP Agent (Windows) database restore to the original location is attempted from the active node.
- If you are restoring a Microsoft SQL Server instance, the restore may be incomplete if you get the following error:
The process cannot access the file because it is being used by another process.
Also, the SQL Server event log reports the following error with Event ID 8355:
Server-level event notifications cannot be delivered. Either Service Broker is disabled in msdb, or msdb failed to start. Event notifications in other databases could be affected as well. Bring msdb online, or enable Service Broker.
If this occurs, free up some available memory or increase the memory to 4 GB or more.
- On the restore Browse Recovery Points dialog, when you expand a directory that contains 50 or greater items, a pagination window displays. If you select that folder, and then you expand it but you do not browse all pages, only the items on the pages which have been browsed are selected. For example, the first page displays by default. If you do not browse to the other pages, only the items on the first page are selected. If you go to page 3 directly, then only the items on the first page and the third page are selected.
Support Considerations
Backup for a striped /mirrored volume is not supported if more than one extent of the volume resides on the same disk. For disk types supported, see Disks Supported by Arcserve UDP Agent (Windows) in the Arcserve UDP Agent (Windows) User Guide or online help.
Miscellaneous Considerations
- To mount a deduplication NTFS volume, ensure that you mount it on a Windows Server 2012 system and that the deduplication feature is installed.
- To browse the directory for restore destination, file copy source, copy recovery point destination, and mount recovery point location, the path limit is 255 characters.
- To browse the directory for backup destination, the path limit is 156 characters. To browse the directory for fie copy destination, the path limit is 255 characters.
- After you rename a Microsoft Exchange database, reboot the Microsoft Exchange server. Otherwise, the Microsoft Exchange VSS writer does not update the database name and Arcserve UDP Agent (Windows) still displays the old database name.
- The following tasks fail on Windows 2008 core machines:
- Mount recovery point
- Catalog-less backup browse and search
- To allow you to troubleshoot your computer problem effectively, Arcserve UDP Agent (Windows) does not work in the Windows Safe Mode.
- To avoid a possible time sync error, if you change the system time when the browser is connected to the Arcserve UDP Agent service, close the browser and log in again.
- After switching to the Arcserve UDP Recovery Point View and mounting a volume under the recovery point, if that recovery point is also included in a merge job, the merge job will fail. To avoid this problem, switch back to the Normal view for the merge job to succeed.
- The remote deployment may fail because the drivers are not signed by Microsoft. You have to install the drivers locally or import the certificate.
Copyright © 2016 .
All rights reserved.
 
|
|