By default, BackupEDGE will use passive FTP. If you require active FTP, append ,p to the
machine field, such as ftp.server.com,p or ftp.server.com:2000,p. Many FTP connections
that go through a firewall will require this mode. If [Test URL] appears to hang, it may be
necessary to kill the edgemenu process, append ,p, and try again.
Connection Timeouts
Many firewalls terminate inactive connections after 15 minutes (the default, but usually
selectable). While a data transfer connection is almost always transmitting something, the FTP control connection remains open and quiescent, and can time out, causing the transfer to close.
BackupEDGE 02.03.01 build 2 and later implement keep-alive packets on the control
connection to prevent this.
Gateway Anti-Virus FTP Inhibition
BackupEDGE utilizes the FTP “REST” command to perform Instant File Restore. This command allows an archive segment to be opened at the exact block where a file begins.
By default, some Firewall / UTMs (Unified Threat Management Systems) block the FTP “REST”
command. This must be enabled. Examples...
+ Edit Backup Schedule ---+ | Schedule Name: [midday_backup] | | Time: [12:01 ] (14:28:24) Enabled: [X] | | Sequence: [Change] web2v.microlite.com:esequence/onsite | | Backup Domain: [Change] system | | Primary Resource: [Change] web2v.microlite.com:url!url0 | | | | +---+ January 2015 | | | Every Sunday of the week (None) | Su Mo Tu We Th Fr Sa | | | Every Monday of the week Differ | D D 3 | | | Every Tuesday of the week Differ | 4 D D D D D 10 | | | Every Wednesday of the week Differ | 11 D D D D D 17 | | | Every Thursday of the week Differ | 18 D D D D D 24 | | | Every Friday of the week Differ | 25 D D D D D 31 | | | Every Saturday of the week (None) | | | +---+ | | | | Notify / Advanced: [Change] [Reset Dates] | | Mail Summary To: root Print Summary To: NONE | | Mail Failures To: NONE Print Failures To: NONE | |[Next] [Back To Select] [Cancel]| +---+ |+Local Machine: web2v.microlite.com Administering: web2v.microlite.com ----+|
• On modern Sonicwall firmware, go into the Security Services, Gateway Anti-Virus
menu, and click on Configure Gateway AV Settings. Make sure Enable FTP ‘REST’ requests with Gateway AV is checked, and click Ok.
• On legacy Sonicwall firmware, you must log in to the Sonicwall, then manually change the URL from the home page URL (usually main.html) to "diag.html". Find the 'Enable FTP 'REST' requests with Gateway AV' checkbox and enable it, then save the settings.
8.8 - RecoverEDGE Reminder
After adding a new Resource to BackupEDGE and creating at least one successful backup, always remember to re-create your RecoverEDGE media and / or images so that they will understand how to use the Resource. Test them to make sure you can read from the Resource if necessary.
s
9 - Configuring NFS Backups
9.1 - General Concepts
NFS Backups are similar to D2D Backups, except that the mount command used mounts a
directory on an NFS server instead of a filesystem on a local hard drive. Two Resources combine to make NFS Backups function:
• FSP Resource, or FileSystem Partition Resource, defines and controls the directory on remote filesystem where archives are stored. No other files may be in this directory except those created by BackupEDGE.
• AF Resource, or Attached FileSystem Resource, defines the commands BackupEDGE must use to mount and unmount the remote filesystem containing the FSP Resource. No other user or process should mount and unmount the remote filesystem.
9.2 - Multiple Archives Per Medium
BackupEDGE supports performing multiple backups using NFS Backups. The quota for NFS Backups is defined by the user during setup. This is the general behaviour:
9.3 - Compatibility Matrix
9.4 - NFS Backup Notes
• Quota (maximum capacity) must be entered in the FSP Resource definition.
• Multiple archives per medium utilizes archive expiration times and lazy reclamation to maximize the number of archives stored for maximum safety.
• Compression and optional encryption are supported.
• Full file checksumming, for maximum data integrity, is supported.
• Instant File Restore is available from any archive. It is not necessary to read through an entire archive to restore individual files and directories.
• MySQL™ hot backups are supported.
Medium Archive Behaviour NFS via
FSP/AF
The process will attempt to create a new archive. If starting a new segment would cause the quota to be reached, the oldest expired archive will be erased in its entirety (all segments). If starting a new segment would exceed the quota and no more expired archives exist, the backup will FAIL.
Operating System
Linux OSR 5.0.7 OSR 6 UW7 NFS Backup Compatibility
Automatic mount / unmount YES YES YES YES
Backup / Restore / Instant File Restore YES YES YES YES
Multiple archives per medium YES YES YES YES
Quotas and Lazy Retention YES YES YES YES
Disable Lazy Retention YES YES YES YES
9.5 - Theory of Operation
For the most part, NFS Backup Resources are very similar to more conventional ones such as tape or DVD. However, there are a few points you should be aware of before using them.
Segments
In a tape backup, BackupEDGE streams the data directly on to the media. In NFS Backups,
BackupEDGE streams the data into archive files on the remote server. This potentially subjects
the files to filesystem size limitations and ulimit or other arbitrary operating system limitations. To work around these limits, BackupEDGE automatically segments archives; that is, it divides one logically long archive into short archive files (called segments) that can be managed by the operating system using the storage device. By default, these segments are 1 gigabyte -1 block in length.
BackupEDGE can write multiple archives to NFS Backup Resources, and each archive may
contain multiple segments. BackupEDGE handles segments automatically, and provides tools for managing the segments. To maintain consistent archives, the individual segments should never be manipulated by operating system commands. This is why segments do not have names that make sense to humans.
Quotas
Each NFS Backup Resource is assigned a storage quota. BackupEDGE will not attempt to use more storage on the target medium than that assigned by the quota.
Retention Times
By default, all archives created to an NFS Backup Resource using the Scheduler have a retention time (or expiration time) of one week. They will not be erased automatically until the retention time is up, but will not necessarily be erased just because its retention time is up. An archive past its retention time is called an Expired Archive.
Space Reclamation
Archives are retained on NFS Backup Resources at least until their expiration time has passed. After that, they are deleted in one of two ways...
Lazy Reclamation Enabled (Default)
If Lazy Reclamation is enabled, archives will remain on media as long as possible, just in case they may be needed even after their retention time is up. This allows maximum space utilization on the media. For an archive to be deleted...
• The retention time must be up, i.e. it must be an Expired Archive.
• Adding a segment to a new archive would cause the defined quota (as defined by the user when defining the Resource) to be exceeded.
If both conditions are true, the oldest Expired Archive will be deleted in its entirety. This process ensures that a maximum number of older archives are available on the target media.
If the quota is reached and none of the archives has expired, the backup will prompt for additional media.
By default, each backup in a Scheduled Job has a Retention Time of 1 week. This is may be changed on a per-schedule basis in the default simple Scheduler, and on a per-backup basis in the advanced Scheduler.
Lazy Reclamation Disabled
Disabling Lazy Reclamation (un-checking the Lazy Reclamation field in the Resource
Definition) configures BackupEDGE to check for and immediately erase all expired archives any
time a new backup is started to the FSP Resource. Only unexpired archives will be retained. By default, each backup in a scheduled job has a Retention Time of 1 week. Note that usually, if you are backing up multiple machines and / or schedules to the same data store, you will create multiple Resources, one per machine/schedule combination. Each FSP Resource would use a different directory on the NFS server and have a different quota, where combined quotas should not exceed available free space. A typical schedule would look like this:
Sample NFS Backup Schedule
This Schedule will perform Monday through Friday backups. In the example, a five backup rotation will be created. Because of the one week default retention, Expired Archives (those older than one week old) will be retained on the target medium at least one week, and possibly longer based on the Lazy Reclamation flag in the Resource definition. If the quota is reached and none of the archives has expired, the backup will fail.
Changing the retention time in the Schedule to 2 weeks, three weeks, etc. allows easy creation of multiple minimal storage rotations.