Difference between revisions of "Ajcody-Backup-Restore-Issues"

m (Bugs And RFE's To Review)
m (Miscellaneous Bugs & RFE's)
Line 8: Line 8:
Please see [[Ajcody-Backup-Restore-Issues]]
Please see [[Ajcody-Backup-Restore-Issues]]
====Bug/RFE's I Filed Against ZCS 8.6====
* [story] Ability to search data within backup and do "item" restores or identify locations of search results
** https://bugzilla.zimbra.com/show_bug.cgi?id=97551
* admin console backup label view doesn't list accounts in the all accounts tab
** https://bugzilla.zimbra.com/show_bug.cgi?id=97555
* admin console restore - doesn't autocomplete / suggest account matches when filling out email address box
** https://bugzilla.zimbra.com/show_bug.cgi?id=97556
* document new restore functions / options with ZCS 8+ for admin console restore
** https://bugzilla.zimbra.com/show_bug.cgi?id=97558
* admin console restore - rename "Selected Servers" panel to "Restore Options"
** https://bugzilla.zimbra.com/show_bug.cgi?id=97559
*admin console restore - if only one mailstore in env. then state such in second panel of restore about "server for the restored accounts"
** https://bugzilla.zimbra.com/show_bug.cgi?id=97560
*admin console restore - expand restore To options - To full backup label, To incremental target
** https://bugzilla.zimbra.com/show_bug.cgi?id=97561
*admin console restore - "restore to the latest backup" incorrectly described / broken
** https://bugzilla.zimbra.com/show_bug.cgi?id=97562
* admin console restore - unable to restore individual accounts [sort of]
** https://bugzilla.zimbra.com/show_bug.cgi?id=97564
* admin console restore - reuse GAL/Contact Picker Window for "restore individual accounts"
** https://bugzilla.zimbra.com/show_bug.cgi?id=97565
====Miscellaneous Bugs & RFE's====
====Miscellaneous Bugs & RFE's====

Revision as of 15:47, 29 January 2015

Attention.png - This article is NOT official Zimbra documentation. It is a user contribution and may include unsupported customizations, references, suggestions, or information.


Backup & Restore Issues

Actual Backup & Restore Issues Homepage

Please see Ajcody-Backup-Restore-Issues

Bug/RFE's I Filed Against ZCS 8.6

Miscellaneous Bugs & RFE's

For those that address other specific issues described on this page:

RFE For Live DR Restore Option
Backups And Upgrades And Prior Versions

Information To Provide When Submitting A Support Case For Backup Issues

Basic Backup Information To Submit To Support

Disk Space Usage Issues
Trend Data

If there is concerns about disks/partitions getting full, this command would be helpful for trending data on your server. Send support the resulting df.tar file . Note - adjust the tail command if you want more than 20 day's worth of trending data, the -n 20 option.

[zimbra@zcs806 tmp]$ /tmp

[zimbra@zcs806 tmp]$ tar cvf /tmp/df.tar `find /opt/zimbra/zmstat -name df.cs\* | sort | tail -n 20`
tar: Removing leading `/' from member names
 [cut - Ajcody]

[zimbra@zcs806 tmp]$ ls -lah /tmp/df.tar
-rw-r----- 1 zimbra zimbra 80K Mar 29 06:44 /tmp/df.tar

[zimbra@zcs806 tmp]$ tar tvf /tmp/df.tar
-rw-r----- zimbra/zimbra  2566 2014-03-11 00:00 opt/zimbra/zmstat/2014-03-10/df.csv.gz
-rw-r----- zimbra/zimbra  2553 2014-03-12 00:00 opt/zimbra/zmstat/2014-03-11/df.csv.gz
 [cut - Ajcody]
-rw-r----- zimbra/zimbra  2513 2014-03-28 00:00 opt/zimbra/zmstat/2014-03-27/df.csv.gz
-rw-r----- zimbra/zimbra  2531 2014-03-29 00:00 opt/zimbra/zmstat/2014-03-28/df.csv.gz
-rw-r----- zimbra/zimbra  8013 2014-03-29 06:40 opt/zimbra/zmstat/df.csv
Directory Sizes In /opt/zimbra

Please see the following and provide the output to support. Note, even though this method is faster than doing a du it still can take awhile.

* Ajcody-Server-Misc-Topics#Faster_Way_To_Get_Directory_Size_On_Filesytem_-_find_vs_du
Adjusting The Disk Alert Threshold

Note - zmlocalconfig smtp_notify must return yes if you want to receive the notifications.

If you just need to adjust the disk alert threshold, then see the following:

See current values:

 zmlocalconfig | grep zmdisklog

Example adjustment:

 su - zimbra
 zmlocalconfig -e zmdisklog_critical_threshold=98
 zmlocalconfig -e zmdisklog_warn_threshold=95

To exclude a partition from the checks [example of two being excluded]:

 su - zimbra
 zmlocalconfig -e zmstat_df_excludes="/mount/point:/mount/point2"

They might be a bug on this, where you'll keep getting email until a logrotate happens [zimbra.log?].

Some things to do to confirm and share with support or in bug. As zimbra

su - zimbra

ls -la /var/log/zimbra.log

df -h
                      5.5G  3.5G  1.7G  68% /
 tmpfs                 939M     0  939M   0% /dev/shm
 /dev/sda1             485M   79M  381M  18% /boot
 /dev/sdb1              30G  6.2G   23G  22% /opt


zmlocalconfig | grep zmdisklog
 zmdisklog_critical_threshold = 80
 zmdisklog_warn_threshold = 85  

zmlocalconfig -e zmdisklog_critical_threshold=95

zmlocalconfig -e zmdisklog_warn_threshold=90

zmlocalconfig | grep zmdisklog
 zmdisklog_critical_threshold = 95
 zmdisklog_warn_threshold = 90

zmstatctl restart


ps -eaf | grep zmstat-df

ls -la /var/log/zimbra.log

date ; grep "Disk warning" /var/log/zimbra* ; zmmailbox -z -m admin@`zmhostname` s -l 100 -t message "Subject: Disk and after:yesterday"

##Note - Emails by default go out every 10 minutes - for example:

[zimbra@zcs803 ~]$ date ; grep "Disk warning" /var/log/zimbra* ; zmmailbox -z -m admin@`zmhostname` s -l 100 -t message "Subject: Disk and after:yesterday"
Thu May 22 09:40:08 PDT 2014
/var/log/zimbra.log:May 22 08:30:00 zcs803 zimbramon[18826]: 18826:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
/var/log/zimbra.log:May 22 08:40:00 zcs803 zimbramon[22970]: 22970:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
/var/log/zimbra.log:May 22 08:50:00 zcs803 zimbramon[22970]: 22970:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
/var/log/zimbra.log:May 22 09:00:00 zcs803 zimbramon[22970]: 22970:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
  ## Note - I had readjusted the variable to not warn during this time segment ##
/var/log/zimbra.log:May 22 09:20:00 zcs803 zimbramon[8322]: 8322:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
/var/log/zimbra.log:May 22 09:30:00 zcs803 zimbramon[8322]: 8322:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
/var/log/zimbra.log:May 22 09:40:00 zcs803 zimbramon[8322]: 8322:err: Disk warning: zcs803.DOMAIN.com: / on device /dev/mapper/vg_rhel664-lv_root at 82%
num: 7, more: false

     Id  Type   From                  Subject                                             Date
   ----  ----   --------------------  --------------------------------------------------  --------------
1.  328  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 09:40
2.  327  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 09:30
3.  326  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 09:20
  ## Note - I had readjusted the variable to not warn during this time segment ##
4.  325  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 09:00
5.  324  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 08:50
6.  323  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 08:40
7.  320  mess   admin                 Disk / at 82% on zcs803.DOMAIN.com:           05/22/14 08:31

Continue to monitor your zmmailbox search results for an hour.

The Basic Information Support Needs

as root:

  • cat /etc/fstab
    • Shows us what is mounted upon boot
  • cat /proc/mounts
    • Shows us what is currently mounted and its status - you can see if a mount is read-only here.
  • df -hT
    • Lists current mounts using human-readable size information and also notes the filesystem type.

as zimbra:

  • zmprov -l gs `zmhostname` | egrep 'Back|Redo'
    • Will show us a number of variables related to backup and redologs. Also tell us if your using auto-group or the default method.
  • du -sh /opt/zimbra/redolog
    • Will might notice your redolog logs aren't rolling over, causing a possible issue.
  • ls -latr /opt/zimbra/backup
    • This is the default backup target, please adjust this path here and below if you are using a different zimbraBackupTarget value.
    • zmprov gs `zmhostname` zimbraBackupTarget
    • We'll be able to confirm permissions are right.
  • ls -latr /opt/zimbra/backup/tmp
    • This will show us if you have failed backup jobs and confirm tmp is being cleaned appropriately after the backup is done.
  • ls -latr /opt/zimbra/backup/sessions
    • This will show us what backup sessions are available and confirm permissions are correct.
    • Adjust path if your zimbraBackupTarget value is not the default path.
  • Some directory sizes in the backup directory:
    • Default path first
      • du -sh `find /opt/zimbra/backup -maxdepth 2 -type d`
    • If your using a different backup target, check that directory also. Replace /opt/zimbra/backup above with your backup path.
  • zmbackupquery
    • This should match what's in the sessions directory and it will also tell us if status of each backup and how many accts were done.
  • crontab -l | grep -i back
    • This will show use when backups are support to run and with what options they are running with.
  • zmlocalconfig | grep -i back
    • This is useful to see a number of backup options not exposed in the crontab, things related to the zip options.
  • zmvolume -l
    • This is useful to see how many volumes are being used, if HSM is being used, and if compression is being done at the volume level.
Additional Log Files Support Might Need

And send the following logs:

  • /var/log/messages
    • Filesystem issues often times are noted here and also in syslog. This might explain an interruption in the backup process. Server restarts, filesystem going full, filesystem going read-only, etc.
  • /var/log/syslog
  • /opt/zimbra/log/mailbox.log
    • The backup activity is logged here.
    • And any other mailbox.log file that would cover the event

Additional Checks For Performance Specific Issues

If Your Using a SAN or NFS For Your Backup Target - Please Check Your IOWait

Ideally, you would compare iowait and performance data from the target backup host as well as the stats available on the ZCS servers. To get graphs and stats on this from ZCS, please see Ajcody-Testing-Debugging#zmstat_and_zmstat-chart . You should submit this data and iowait conclusions if you still need to submit a support case about backup performance issues.

Is HSM Running During Your Backup Window
Are You Using --zipStore

--zipStore zips the blobs vs. keeping the blobs as individual files. --zipStore does not use compression either. For most circumstances, this will give the best performance, especially with NFS. This should be the default behavior of the backups, the following RFE is when it became the default [ZCS6+] :

To see if zip's are being used for backups for example, in the backup/session directory you'll know if it is by seeing .zip files:

zimbra$ ls blobs-1.zip    blobs-2.zip    blobs-3.zip    blobs-4.zip

To see if the zip file is using compression [-Z option for unzip will indicate whether or not the archive is actually compressed] :

unzip -Z blobs-4.zip
293 files, 5982984 bytes uncompressed, 5982984 bytes compressed:  0.0%

Also, if your zmvolume has compression enabled the blobs will remain compressed within the zip also upon backup. The point being, they are uncompressed to be then put into a zip file when the backup is using --zipStore.

Restore Compatibility Between ZCS Versions

Please see the following for details. In summary, user level restores should work against older ZCS backup data.

Other related bug/rfe's:

NFS Use For Backups

Please see this RFE:

This is the proposed statement to be included in the release notes following the RFE:

Zimbra will support customers that store backups (e.g. /opt/zimbra/backup) on an NFS-mounted partition. Please note that this does not relieve the customer from the responsibility of providing a storage system with a performance level appropriate to their desired backup and restore times. In our experience, network-based storage access is more likely to encounter latency or disconnects than is equivalent storage attached by fiber channel or direct SCSI.
Zimbra continues to view NFS storage of other parts of the system as unsupported. Our testing has shown poor read and write performance for small files over NFS implementations, and as such we view it unlikely that this policy will change for the index and database stores. We will continue to evaluate support for NFS for the message store as customer demand warrants.
When working with Zimbra Support on related issues, the customer must please disclose that the backup storage used is NFS.

Things To Check

  • Check the /var/log/messages on both the zimbra server and the nfs server for nfs related errors during the time frame of your backup.
  • Check /opt/zimbra/log/mailbox.log for error messages about folders/files not being able to be written or missing directory errors.
  • Is root_squash configured on the nfs server? If it's changed to no_root_squash , does the behavior of the backup change?
  • Is the */backup directory owned by zimbra:zimbra with at least 750 or 755 permissions?
    • This parent directory as given in:
      • zmprov gs `zmhostname` zimbraBackupTarget
  • Does zimbraBackupTarget have at least the subdirectories of : sessions and tmp  : and are owned by zimbra:zimbra with 750 or 755 permissions?
    • If not, try manually creating them and then running a test backup.
    • If your backup session directory shows something like : drwxrwx---+ 2 zimbra zimbra 4096 Sep 14 00:00 TO_DELETE-full-XXXXXX , that + sign indicated extended acls are in use.

Debugging Example

Steps I wrote for one customer, where saving out the information as you walk through all the commands would give enough information [hopefully] to submit a good rfe/bug:

 1. make a test partition on nfs server - /nfs-test

 2. mount on zimbra server
 2A. mkdir /nfs-test
 2B. chmod 755 /nfs-test
 2C. mount nfs-server:/nfs-test /nfs-test
 2D. ls -la /nfs-test
 2E. mkdir /nfs-test/backup
 2F. chown zimbra:zimbra /nfs-test/backup
 2G. chmod 755 /nfs-test/backup
 2H. su - zimbra ; touch /nfs-test/backup/testfile 
 2I. ls -laR /nfs-test/
 2J. rm /nfs-test/backup/testfile

 3. Set zimbraBackupTarget
 3A. zmprov ms `zmhostname` zimbraBackupTarget /nfs-test/backup

 4. Run a full backup against one account
 4A. ex. zmbackup -f -a user@domain.com

 5. ls -laR /nfs-test/

 6. If you again, run into the same problem. You could also repeat the backup after increasing the backup 
    logging variable for the account your trying to backup. If you didn't run into the same problem, it might 
    had to do with the initial setup of the nfs mount and permissions being used during the directory creation.
 6A. zmprov aal user@domain.com zimbra.backup debug
 6B. logging will show up in /opt/zimbra/log/mailbox.log
 6C. Remove account logging when your done.  zmprov ral user@domain.com zimbra.backup

 8. Change zimbraBackupTarget back to your production path.

Setup A Fast Test NOT Using NFS

A way to "avoid" the NFS issues for testing purposes would be to setup a new zimbraBackupTarget to try doing a full backup of a couple of user accounts. I DON'T recommend this if your using auto-group for zimbraBackupMode [ zmprov gs `zmhostname` zimbraBackupMode ] , only if your using Standard mode.

[as root]
** adjust your new "backup" directory path as needed - mine is just an example**
mkdir /mnt/usb1/backup-test
chown zimbra:zimbra /mnt/usb1/backup-test
chmod 750 /mnt/usb1/backup-test
su - zimbra
**confirm backup mode as standard**
zmprov gs `zmhostname` zimbraBackupMode
**if not stand, please stop**
zmprov ms `zmhostname` zimbraBackupTarget /mnt/usb1/backup-test
**Find a couple of test accounts to do a full for. Make sure you'll have space to do the backup regards to need free space.**
zmbackup -f -a user1@domain.com user2@domain.com user3@domain.com
**Watch and confirm status of backup you just started.**
**Confirm files were backed up in right location**
ls /mnt/usb1/backup-test/sessions/
**Failed backups would most likely results in left over directory in tmp directory**
ls /mnt/usb1/backup-test/tmp/

Restore For Disaster Recovery

For Full Single ZCS Server DR Restores

Please see Network_Edition_Disaster_Recovery

Some additional notes I have on it - Ajcody-Disaster-Recovery-Specific-Notes

For Multi-Server DR Restore Specifics

Along with the above references, please see Ajcody-Notes-Multi-Server-Restore-DR

To Restore Just The LDAP Date

Let's say your ldap data was 'lost/destoyed' but everything else was intact, you should look at the zmrestoreldap command.

This section should have more precaution and background information to handle this section.

The basics:

  • To find the LDAP session labels type -lbs.
    • zmrestoreldap -lbs
  • Restore the complete LDAP directory server
    • zmrestoreldap -lb full20061130135236
  • Restore LDAP data for specific accounts
    • zmrestoreldap -lb full20061130135236 -a tac@DOMAIN.com jane@DOMAIN.com

To Restore Just The Mysql DB

Option available from the following RFE work:

  • "Allow backup of only primary message volume"
    • https://bugzilla.zimbra.com/show_bug.cgi?id=35278
      • Options to exclude types of data as described in the 6.0.6 Admin Guide:
        • Search index : If you do not restore the search index data, the mailbox will have to be reindexed after the restore.
          • zmrestore <all or account> --exclude-search-index
        • Blobs : This is a useful option when all blobs for the mailbox being restored already exists.
          • zmrestore <all or account>|--exclude-blobs
        • HSM-blobs : This is useful when all HSM blobs for the mailbox being restored already exists.
          • zmrestore <all or account> --exclude-hsm-blobs

Let's say your mysql data was 'lost/destoyed' but everything else was intact. This might be the solution for the situation:

The steps below are to REPRODUCE A DR situation to test.

  1. zmcontrol stop
  2. Caution - step to reproduce DR situation to test against
    1. mv ~/db/data ~/db/data.OLD
  3. Getting old mysql passwords
    • zmlocalconfig -s mysql_root_password
    • zmlocalconfig -s zimbra_mysql_password
  4. [as zimbra] /opt/zimbra/libexec/zmmyinit
  5. ldap start
  6. zmconvertctl start
  7. mysql is already running per the zmmyinit - mysql.server status - to check.
  8. zmrestoreoffline -a all -br --systemData --excludeSearchIndex --excludeBlobs --excludeHsmBlobs
    • Note - you most likely will want to use the -br or -rf option for this situation.
      • -br,--backedupRedologsOnly : Replays the redo logs in backup only, which excludes archived and current redo logs of the system
      • -rf,--restoreFullBackupOnly : Restores to last full backup only, which excludes incremental backups.
    • -sys,--systemData : Restores global tables and local config.
    • Note the use of -a all , you could also do this for one account first to confirm operation is successful for your circumstances.
  9. If you used -rf or -br with your zmestoreoffline, you might also need to use zmplayredo to finish it up to get the most complete restore.

Performance Issues And Time To Complete

Please see Ajcody-Notes-BackupPlans#What_About_Backups.3F_I_Need_A_Plan

Also created an RFE for increase backup performance:

Understanding Option Flags For zmbackup & zmrestore

First, they don't make sense if your just reading from the help output - I will not argue this point at all.

The biggest problem with the options I point out below is that you can often include them in the command and they do nothing or you include them for a particular situation and they don't apply. Why is this a problem? Because they are silent and give no output telling you that it's not necessary, it's redundant, or it will actually cause your intended results to fail simply because you included the option.

zmrestore Options

Problems mostly revolve around these options.

To Times In The Past (If -lb Isn't Used, Implies Your Using Times/Incr/RedoSeq AFTER Last Full)
  • -restoreToIncrLabel
    • <arg> Replay redo logs up to and including this incremental backup
      • Requires: --label or -lb
  • -restoreToTime
    • <arg> Replay rodo logs until this time
      • Requires: --label or -lb
  • -restoreToRedoSeq
    • <arg> Replay up to and including this redo log sequence
Redolog Variables
  • --backedupRedologs or -br
    • Replays the redo logs in backup only, which excludes archived and current redo logs of the system
      • Only useful when restoring against latest full backup (NOT using the -lb option).
      • Will restore using incremental backup data after the last full backup as well but NOT including any redolog activity.
  • --restorefullBackup or -rf
    • Restores to the full backup only, not any incremental backups since that backup.
      • The default behavior of zmrestore in general is to always play from a "full" and "incrementals" that are associated with it UNLESS you state otherwise.
        • If you do:
          zmrestore -a user@domain.com -lb full6monthsago
        • It will playback the incremental data associated with that full from 6 months ago.
        • If you do:
          zmrestore -a user@domain.com -lb full6monthsago -rf
        • It will ONLY playback the data in the full from 6 months ago.
      • Will not progress past the data that's in the last full backup, no incremental backups after it in other words.
      • Implies NO redolog play, so there's no need to use -br.
Targets And Labels
  • --label or -lb
    • <arg> The label of the full backup to restore. Restores to the latest full backup if this is omitted.
  • --target or -t
    • <arg> Specifies the backup target location. The default is <zimbra_home>/backup.
Impact Of AutoGroup Option Being Used

Place for notes about how autogroup backup option might impact or limit command options.

zmbackup Options

Problems mostly revolve around these options:

  • --target or -t
    • <arg> Specifies the target backup location. The default is <zimbra_home>/backup.
  • --zip or -z
    • Zips email blobs in backup - using compression
  • --zipStore
    • Zips email blobs in backup - does NOT use compression
Deleting Old Backups -del

Caution You want to delete from the oldest label to newest. The -del option will automatically purge all older sessions prior to the label you used. To find out the label names, use zmbackupquery.

Format example:

zmbackup -del <oldest_backup_label>
Impact Of AutoGroup Option Being Used

Place for notes about how autogroup backup option might impact or limit command options.

Changing Default Backup Target

To find out what the current backup target is, do:

 zmprov gacf | grep zimbraBackupTarget
 zimbraBackupTarget: /opt/zimbra/backup

This is also configurable at the "server" level:

 zmprov gs `zmhostname` | grep zimbraBackupTarget
 zimbraBackupTarget: /opt/zimbra/backup

For example:

 zmprov ms `zmhostname` zimbraBackupTarget /san/mount/backup

Issues of changing the default path in regards to the admin web console. Please see:

Another way to change the backup path is described at Changing_Backup_directory_and_General_Information. I recommend reviewing it as well.

See Change Location For Backup Or Restore Source Data for non-default adjustments to CLI commands in regards to a non-default path for backups.

RFE & Bugs Concerning Option Flags For Restore And Backup

Setting Account Status After Restore Is Done

I filed an RFE for this:

Use DL COS Status To Generate User List Of Accounts

I filed an RFE for this:

An Overview Of Some Backup/Restore Items

Shared Blobs (messages)

I believe we are a little light on describing the shared blobs situation. Shared blobs can cause different corrections to a problem as compared to a normal message issue that isn't shared. I'll start some notes on this here.

From Backup and Restore

  • "When backing up shared messages, the backup process looks to see whether a Binary Large Object file (BLOB) representing a message already exists in the backup. If it does, it simply flags this object as such and does not copy its content again."
  • "Keeping the same backup target saves disk space, because shared binary large object files (BLOB) and other files do not have to be duplicated every time the backup process runs.


Remote copies of backup data for DR use

You will want to copy over /opt/zimbra/redolog/archive/* and /opt/zimbra/redolog/redo.log frequently in order to stay current. The redo.log file being open is not a problem since the crash recovery step can work with redo.log file in any state.

The redolog/archive/ contains logs that have not yet been backed up by an incremental (or by a full in auto-grouped mode)

The redo.log rolls over when it reaches zimbraRedoLogRolloverFileSizeKB (by default 100MB prior to ZCS 5.0.11 and 1GB after). When ZCS restarts after a crash, it seems to work through the current redo.log ok regardless of its state, if the current log really must be copied."

My Initial Thoughts On This

Start of process:

  • Weekend full on prod
  • rsync full-xxx on prod > remote sessions/
  • rsync redolog/* > remote redolog/
    • through non-full and incremental times every x about of minutes
  • weekday nights 10pm incre-xxx on prod
  • rsync incr-xxx on prod > remote sessions/
  • rsync redlog/* > remote redolog/

Created three separate rsync cron rules.

  • Full - once a week
    • Confirms full is done and then looks for latest full-xxxx and rsyncs that specific directory
  • Incre - once a night except for full schedule night
    • Confirms incre is done and then looks for latest incr-xxx and rsyncs that specific directory
  • Redolog/ - every x amount of minutes (outside of full and incr backups sessions)
    • Soes full rsync of redolog/ - probably want delete/remove option?
    • lsof will report /opt/zimbra/redolog/redo.log is open.

Somewhere we need to account for accounts.xml in this process. And also confirm what else might be missing. Also, steps on the actually restore process depending on when/where the DR event took place.

What's Needed For Later Restores

In regards to what is moved and what is needed for later restores, you must remember this "flow" of the backups:

  • Full backup files that contains all the information needed to restore mailboxes (to that point in time)
  • Incremental backup files that contains the LDAP directory server files and all the redo log transactions written since the last backup (to that point in time)
  • Redo logs that contains current and archived transactions processed by the Zimbra server since the last incremental backup (to that point in time - more about this topic is above)

Variables to be aware of in regards to backup/restore [ viewed with - zmprov gs [server-name] | grep Backup ]:

  • This is that gui path option - note, it doesn't change default to redo file path [see below]


  • Variables to be aware of in regards to redo files [ viewed with - zmprov gs [server-name] | grep Redo ]:

Mysql Table That References Most Recent Backup Session Of Users (AutoGroup Backup Mode)

During the backup we update the zimbra.mailbox table for each mailbox to record the most recent backup time. This is in the "last_backup_at" column within Mysql.

This data is used by auto-grouped backup to figure out which mailboxes to backup.

Creating A List Of Last Backup Of Users

Remove the , | head , below to get a full listing of all your accounts. Note, this reports on the users that exist on the mailstore your running the command.

[zimbra@zcs806 tmp]$ mysql zimbra -NBe 'select from_unixtime(last_backup_at), comment from mailbox' | sort | head
2014-06-05 01:00:24     dluser558@zcs806.DOMAIN.com
2014-06-05 01:00:26     dluser557@zcs806.DOMAIN.com
2014-06-05 01:00:28     dluser556@zcs806.DOMAIN.com
2014-06-05 01:00:30     dluser555@zcs806.DOMAIN.com
2014-06-05 01:00:31     dluser554@zcs806.DOMAIN.com
2014-06-05 01:00:32     dluser551@zcs806.DOMAIN.com
2014-06-05 01:00:32     dluser552@zcs806.DOMAIN.com
2014-06-05 01:00:32     dluser553@zcs806.DOMAIN.com
2014-06-05 01:00:33     dluser550@zcs806.DOMAIN.com
2014-06-05 01:00:34     dluser549@zcs806.DOMAIN.com

I include also below the results from the zmbackupquery for the first user:

[zimbra@zcs806 tmp]$ date ; zmbackupquery -a dluser558@zcs806.DOMAIN.com
Wed Jun 11 14:29:48 PDT 2014
Account: dluser558@zcs806.DOMAIN.com

    Label:   full-20140605.080023.296
    Type:    full
    Started: Thu, 2014/06/05 01:00:23.296 PDT
    Ended:   Thu, 2014/06/05 01:02:37.739 PDT
    Acct ID: 85d3c8f4-eea8-4cf2-8a84-8f1fcc87d555

    Label:   full-20140529.080016.546
    Type:    full
    Started: Thu, 2014/05/29 01:00:16.546 PDT
    Ended:   Thu, 2014/05/29 01:02:51.556 PDT
    Acct ID: 85d3c8f4-eea8-4cf2-8a84-8f1fcc87d555

    Label:   full-20140522.080015.818
    Type:    full
    Started: Thu, 2014/05/22 01:00:15.818 PDT
    Ended:   Thu, 2014/05/22 01:02:42.126 PDT
    Acct ID: 85d3c8f4-eea8-4cf2-8a84-8f1fcc87d555

    Label:   full-20140515.080016.160
    Type:    full
    Started: Thu, 2014/05/15 01:00:16.160 PDT
    Ended:   Thu, 2014/05/15 01:02:34.204 PDT
    Acct ID: 85d3c8f4-eea8-4cf2-8a84-8f1fcc87d555
Possible Issue That A Failed Or Interrupted Backup Causes

An interrupted backup can cause an issue because the table currently gets updated right off the bat rather than waiting for backup to be successfully completed.

Possible RFE: To update zimbra.mailbox.last_backup_at column for successfully backed-up mailboxes to the very end of the backup process, to either just before or just after renaming the /opt/zimbra/backup/tmp/<backup label> directory to /opt/zimbra/backup/sessions/<backup label>.

Setting To Null To Cause A New Backup For User

To undo what was done by an interrupted backup for example, you need to clear this column (set it to null) for the affected mailboxes. By clearing the column, you're forcing the next AG backup to choose these mailboxes because they look like they have never been backed up. If you don't clear this column, you have to wait until the next cycle. (7 days)

Example syntax to view:

mysql zimbra -e "select last_backup_at from mailbox where id=27"

Example syntax to change data of the last_backup_at to NULL"

mysql zimbra -e "UPDATE mailbox SET last_backup_at = NULL WHERE id = 27"
Related Bugs RFE's

Restore Requires accounts.xml File

Also, see:

Change "Location" For Backup Or Restore Source Data

Remember that zmbackup and zmrestore can take flags as well in regards to the location of items.

  • zmrestore & zmbackup can both take : -t,--target (default <zimbra_home/backup)

You can't state a different location with redo logs though. There's a command called zmplayredo [for newer versions of ZCS] and it has a variable to point to the redologs to play from [ --logfiles ]. It will replay into the default redolog directory or redolog file. The mailbox has to be stop to run zmplayredo . This is a command to manual kick off a replay of a redo log. This is normally done with the zmrestore when options about to a specific time aren't included.

Manual Removal Of Older Backup Sessions

General Situation:

  • Keep in mind every restore requires starting with data from a full backup.
  • For each account on the server, there must be at least 1 full backup after the deletion is complete.
  • You should also make sure all incremental backups made after the oldest of the remaining backups are retained.
    • This basically gets reduced to deleting only those backups that are old enough, based on your full/incremental schedule.

More Specific Issues:

  • Does the accounts.xml dependency cause issues with this?
    • No. Just don't delete it.
  • What about the contents of the backup/tmp/ data or shared blobs type references?
    • Don't touch this directory either. It is used during backup and restore. You don't want to change its content while an operation is going on, so best to leave it alone.
  • What if a zimbra server is running some type of restore of backup command while the manual removal is running on the nfs server?
    • You shouldn't remove the backups that are being used in a restore currently underway. You are responsible for avoiding the race condition.
      • Please understand you are responsible for avoiding the race condition. Make sure no backup or restore is happening at the moment, then rename the directories that will be deleted, preferably move them to another subdirectory, e.g. /opt/zimbra/backup/sessions_to_delete. Then delete.

LDAP Backup Related Items

Backup Schedule On LDAP Only Server [non-Mailstore]

If you look at the code in /opt/zimbra/bin/zmschedulebackup  :

if ($BACKUP_MODE eq 'Standard') {
    # default schedule: full backup 1am every sunday, incr backup 1am every weekday
    # deletes backups older than a month at 12am everyday
  if (isLdapOnly()) {
      @default = ("0 1 * * 6 /opt/zimbra/bin/zmbackup $target\n",
            "0 0 * * * /opt/zimbra/bin/zmbackup -del 1m\n"); 
  } else { 
      @default = ("0 1 * * 6 /opt/zimbra/bin/zmbackup -f $account $target $compress \n",
          "0 1 * * 0-5 /opt/zimbra/bin/zmbackup -i $compress\n",
            "0 0 * * * /opt/zimbra/bin/zmbackup -del 1m\n");
    %commandmap = ("f", "/opt/zimbra/bin/zmbackup -f $target $compress $account",
                   "i", "/opt/zimbra/bin/zmbackup -i $compress",
                   "d", "/opt/zimbra/bin/zmbackup -del");
} else {  # Auto-Grouped mode
    # default schedule: full backup 1am everyday, no incr backup
    # deletes backups older than a month at 12am everyday
    @default = ("0 1 * * 0-6 /opt/zimbra/bin/zmbackup -f $target $compress\n",
                "0 0 * * * /opt/zimbra/bin/zmbackup -del 1m\n");
    %commandmap = ("f", "/opt/zimbra/bin/zmbackup -f $target $compress",
                   "i", "/opt/zimbra/bin/zmbackup -i $compress",
                   "d", "/opt/zimbra/bin/zmbackup -del");

Notice the specific check and then format for ldapOnly:

  if (isLdapOnly()) {
      @default = ("0 1 * * 6 /opt/zimbra/bin/zmbackup $target\n",
            "0 0 * * * /opt/zimbra/bin/zmbackup -del 1m\n"); 

Related Bugs To LDAP Backups

Some bugs to be aware of - most are resolved/fixed:

A Way To Verify Backup Integrity

I filed an RFE for this:

Negative Seek Offset Error & RFE

Explanation of negative seek offset error during a restore attempt and manual fixes are located here:

Auto-Group Backups Rather Than Default Method Topics

General Description And Official References

Having trouble completing that entire full backup during off-hours? Enter the hybrid auto-grouped mode, which combines the concept of full and incremental backup functions - you’re completely backing up a target number of accounts daily rather than running incrementals.

Auto-grouped mode automatically pulls in the redologs since the last run so you get incremental backups of the remaining accounts; although the incremental accounts captured via the redologs are not listed specifically in the backup account list. This still allows you to do a point in time restore for any account.

Administrative manual page:


Compare the sections called "Standard Backup Method" & "Auto-Grouped Backup Method"


Configuration details are here on that page:


Good explanation:


Simply divide your total accounts by the number of groups you choose (zimbraBackupAutoGroupedNumGroups is 7 by default) and that’s how many will get a full backup session each night. Newly provisioned accounts, and accounts whose last backup is a specified number of days older are picked first. (zimbraBackupAutoGroupedInterval is defaulted to 1d)

Think of auto-grouped mode as a full backup for the scheduled group as well as an incremental (via redologs) for the all other accounts at the same time.

Bugs - RFE's To Review For Auto-Group

Please see:

Enabling Auto-Group For Backups - Schedule In Crontab

  • You'll need to make the variable changes as listed in the administrative guide and then the following for set the crontab correctly.
  • Run zmschedulebackup --help to see a list of options.
  • Run zmschedulebackup -D , which will now set a new default schedule (crontab) that uses auto-group settings now that your variable are set for it.
  • Incremental backups aren't performed as they were.
    • You'll see there's no longer the -i option in cron.
    • Incrementals are performed, in a manner, but the redologs are just copied into the full backup session

Two bugs to look at as well:

Standard Mode Default Cron Setup
  • Here is a schedule without auto-grouped backups enabled:
    • The Default Schedule For Normal Backups [ zmschedulebackup --help ]:
 f    0 1 * * 6
 i    0 1 * * 0-5
 d 1m 0 0 * * *
    • To set the backup to standard mode and use the default schedule one would:
      • zmprov mcf zimbraBackupMode Standard
      • zmschedulebackup -D
    • The crontab would look like:
 0 1 * * 6 /opt/zimbra/bin/zmbackup -f -a all   
 0 1 * * 0-5 /opt/zimbra/bin/zmbackup -i -a all
 0 0 * * * /opt/zimbra/bin/zmbackup -del 1m
Auto-Group Mode Cron Setup
  • Here is a schedule with auto-grouped enabled:
    • The Default Schedule For Auto-Group Backups [ zmschedulebackup --help ]:
f 0 1 * * 0-6
d 1m 0 0 * * *
    • To set the backup to auto-group mode and use the default schedule one would:
      • zmprov mcf zimbraBackupMode Auto-Grouped
      • zmschedulebackup -D
    • The crontab would look like:
0 1 * * 0-6 /opt/zimbra/bin/zmbackup -f
0 0 * * * /opt/zimbra/bin/zmbackup -del 1m

Some Variables For Auto-Group

The below might not be complete or the defaults, I just wanted to save this before I forget them. Try to get more complete details on these later.

zmprov gacf | grep Backup
zimbraBackupAutoGroupedInterval: 1d
zimbraBackupAutoGroupedNumGroups: 7
zimbraBackupAutoGroupedThrottled: FALSE
zimbraBackupMode: Auto-Grouped

Auto-group And Redologs

Please see Ajcody-Backup-Restore-Issues#Redologs_And_Auto-group_In_Regards_To_Backups

Problems Switching To Auto-Groups Because It Wants To Run A Full Against All Accounts

Please see the following bug/rfe made about problems switching over to Auto-group when the first backup run of it tries to backup ALL of he accounts. I have the full how-to within the bug. It basically manipulates the last_backup_at for each account.

Backup And Deletion Schedule - zmschedulebackup

How To Adjust The Deletion Schedule

For my example below, I first set the backup to the "default" schedule. And I then adjust that "default" to have backups delete with a 14 day interval rather than 1 month.

[zimbra@mail3 ~]$ zmschedulebackup -D
Default schedule set

Current Schedule:

        f 0 1 * * 6 -a all
        i 0 1 * * 0-5
        d 1m 0 0 * * *

[zimbra@mail3 ~]$ zmschedulebackup -q
Current Schedule:

        f 0 1 * * 6 -a all
        i 0 1 * * 0-5
        d 1m 0 0 * * *

[zimbra@mail3 ~]$ zmschedulebackup -R f "0 1 * * 6" i "0 1 * * 0-5" d 14d "0 0 * * *"
Schedule replaced

Current Schedule:

        f 0 1 * * 6 -a all
        i 0 1 * * 0-5
        d 14d 0 0 * * *

[zimbra@mail3 ~]$ zmschedulebackup -q
Current Schedule:

        f 0 1 * * 6 -a all
        i 0 1 * * 0-5
        d 14d 0 0 * * *

[zimbra@mail3 ~]$ crontab -l | grep backup
0 1 * * 6 /opt/zimbra/bin/zmbackup -f   -a all
0 1 * * 0-5 /opt/zimbra/bin/zmbackup -i  
0 0 * * * /opt/zimbra/bin/zmbackup -del 14d

The Zip - Compression Option For Backups

Using the zip option will compress all those thousands of single files that exist under a user's backup, decreasing performance issues that arise from writing out thousands of small files as compared to large ones. This is often seen when one is :

  • Using nfs for the backup directory
  • Copying/rsyncing backups to a remote server
  • Are using some third party backup software (to tape) to archive/backup the zimbra backup sessions.

Optional Tweaks To The Zip Options

Please see this comment and those underneath it within this RFE:

  • "Use zip files for shared blobs of a full backup made with --zip option"
    • http://bugzilla.zimbra.com/show_bug.cgi?id=26624#c6
      • Use zmlocalconfig. To set:
        • $ zmlocalconfig -e key=val
      • To unset:
        • $ zmlocalconfig -u key
      • Once you set the key you will be able to view it.
        • $ zmlocalconfig
      • backup_zip_copier_private_blob_zips
        • How many zip files to distribute a mailbox's private (unshared) blobs over; default 4 (blobs-1.zip through blobs-4.zip); range 1 to 10,000
      • backup_zip_copier_copy_buffer_size
        • File copy buffer size; default 16384 (16KB); range 4KB to 1MB
      • backup_zip_copier_queue_capacity
        • Each zip file gets a queue of files to add. This key sets the queue size. Default is 10. Range is 1 to 10,000.
      • backup_zip_copier_deflate_level
        • Compression level. Default is -1. (same as in java.util.zip.ZipOutputStream). -1 is same as level 6. This behavior comes from zlib library which the JVM uses to implement zip. Other than the special default value, the level can range from 0 to 9. 0 means no compression. 1 means fastest compression and 9 means best compression.
      • backup_disable_shared_blobs
        • This one isn't limited to zip backups. When this is set to true, all blobs are backed up as private backups. Default is false.
      • backup_debug_use_old_zip_format
        • If true, backup will behave like ZCS 5.0.4 and earlier. Shared blobs are never zipped, and private blobs are added to a single blobs.zip file in zip backup. Default is false.

Need To Write Fewer Files - Add The Zip Option To Your Backup Commands

RFE to make zip option the default for backups:

There is very little details in the official documentation on this option unfortunately. This does have a really good explanation though:


From the administrative manual on the Backup section:


It says,

"-zip can be added to the command line to zip the message files during backup. Zipping these can save backup storage space."

It's implied that instead of having all the individual message files in the backup that it will bunch them all together into zip files. The body of a shared blob is added once to a shared-blobs zip file, then a small pointer-only entry is added to a mailbox's zip file. Same effect as in non-zipped case. This will be useful when the number of message files is causing disk i/o issues. Maybe your trying to rsync the backup session directories off to another server or your running a third party backup on it to save to tape. The default use of -zip will use compression, if this also causes overhead that you need to avoid you can use the -zipStore option.

Note about -zipStore:

"when used with the -zip option, it allows the backup to write fewer files (-zip), but not incur the compression overhead as well"

The zip options effect backups that are in blob formats (full's). Incremental backups are bascially redologs, not the full message store of the user. In summary, the zip option will not impact the increment type backups. Auto-group backups are a mixture of both fulls and incrementals.

How To Use As A Default Option?

You'll add the options to the zimbra crontab file. This can be done with the zmschedulebackup command.

Run zmschedulebackup with help option:

zmschedulebackup --help

You'll see:

-z: compress - compress email blobs with zip

It appears that you'll need to manually add the options about -zipStore , if you want it, to the crontab file.

See bug :


What Does It Look Like When I Use Zip?

Shared blobs are zipped and blobs (messages) are zipped per root store directory.

zimbra$ ls blobs-1.zip    blobs-2.zip    blobs-3.zip    blobs-4.zip

General Backup & Restore Debugging

You'll be monitoring the /opt/zimbra/log/mailbox.log file

Include the -d / --debug option on the CLI for either zmrestore or zmbackup .

To increasing logging for backup/restore-related logs - /opt/zimbra/log/mailbox.log . Enable DEBUG log level for "zimbra.backup" logger in :

  • /opt/zimbra/conf/log4j.properties for "temporary" change - until next restart. This could take a couple of minutes before jetty "sees" the changes.
  • /opt/zimbra/conf/log4.properties.in for "permament" change that will stick after restart. A restart of jetty/mailbox would be required for this change - zmmailboxctl restart .

For incremental backups, this should log each redolog being copied to the backup and also log which ones will be deleted out of archive directory. Those not deleted are kept because they are newer than 1 hour (default). The kept logs are deleted (but not copied again) during the next incremental backup.

Redolog Files

Redologs Copied To Backup Session And When Deleted

Archived logs that are less than an hour old at the time of incremental backup are copied to the backup but aren't deleted to support post-crash waitset reinitialization mechanism. The interval is set in localconfig key backup_archived_redolog_keep_time, which is in seconds, default=3600.

An Outline Of The Step
  1. /opt/zimbra/redolog/redo.log (starts and then grows to zimbraRedoLogRolloverFileSizeKB size - default 100MB)
  2. This flushes to /opt/zimbra/redolog/archive/[file] upon hitting the zimbraRedoLogRolloverFileSizeKB.
  3. 1 & 2 keep repeating when zimbraRedoLogRolloverFileSizeKB is hit.
  4. When a backup is done, the archive/* files are copied. The redo.log file is not moved.
    1. When the backup processes archive/* logs, it first figures out the last sequence copied to backup. All newer logs are copied to the current backup. Then, all logs are deleted except those that are too new, determined by localconfig parameter backup_archived_redolog_keep_time, which defaults to 1 hour. (This is part of the waitset feature.)
      • In standard backup mode, only incremental backups move the redologs.
      • In auto-grouped mode, every backup is a hybrid of full and incremental and thus redologs are moved.
Redologs And Auto-group In Regards To Backups


Think of auto-grouped mode as a full backup for the scheduled group as well as an incremental (via redologs) for the all other accounts at the same time. Auto-grouped mode automatically pulls in the redologs since the last run, so you get incremental backups of the remaining accounts. Although the incremental accounts captured via the redologs are not listed specifically in the backup account list. This still allows you to do a point in time restore for any account.
If You Have Older Redologs Not Being Deleted

According to the code, only archived logs newer than 1 hour old (default for backup_archived_redolog_keep_time) should remain after an incremental backup. It is a bug if you are seeing older logs sticking around. If so, look at mailbox.log and see if any error was logged. If you enable DEBUG logging for "zimbra.backup" logger in log4j.properties you will see log statements for each copy and deletion.

The zimbraRedoLogDeleteOnRollover variable

zimbraRedoLogDeleteOnRollover shouldn't have an effect on "If you have older redologs not being deleted". By default it's FALSE and affects whether or not stuff makes it into /opt/zimbra/redolog/archive at all. With it set to TRUE there's just /opt/zimbra/redolog/redo.log and it's deleted/not rolled over into archive. As discussed above old redologs are deleted after the incremental; thus if you don't take incremental backups you should set this value to TRUE or periodically script manual deletion of /opt/zimbra/redolog/archive. (And with zimbraRedoLogEnabled FALSE there's no redo.log at all.)

If You Don't Run Incremental Backups Or Don't Need Archive Redologs

You would set zimbraRedoLogDeleteOnRollover to TRUE.

(Auto-Grouped backups you can still leave this to the default of FALSE.)

Redolog Sequence And The Backup Session

Redologs will exist in the incremental backup sessions. The zmbackupquery command will reference the redologs associated with the backup. For example"

[zimbra@mail3 ~]$ ls /opt/zimbra/backup/sessions/incr-20080925.224528.230/redologs/
redo-20080925.213136.165-seq53.log  redo-20080925.220726.521-seq55.log  redo-20080925.224209.287-seq57.log
redo-20080925.215516.450-seq54.log  redo-20080925.221749.133-seq56.log

[zimbra@mail3 ~]$ zmbackupquery -lb incr-20080925.224528.230
Label:   incr-20080925.224528.230
Type:    incremental
Status:  completed
Started: Thu, 2008/09/25 18:45:28.230 EDT
Ended:   Thu, 2008/09/25 18:45:39.099 EDT
Redo log sequence range: 53 .. 57
Number of accounts: 2

In the above example, we see the sequence range of "53 .. 57" is referring to the files in the backup session directory called redologs.

RedoLog Variables

Changing Redolog File Size And Location

The /opt/zimbra/redolog/redo.log rolls over when it reaches zimbraRedoLogRolloverFileSizeKB (by default 100mb).

The "roll overs" then goto /opt/zimbra/redolog/archive/

zmprov gacf | grep Redo
zimbraRedoLogArchiveDir: redolog/archive
zimbraRedoLogDeleteOnRollover: FALSE
zimbraRedoLogEnabled: TRUE
zimbraRedoLogFsyncIntervalMS: 10
zimbraRedoLogLogPath: redolog/redo.log
zimbraRedoLogRolloverFileSizeKB: 102400
Need To Move Redologs Because Partition Getting Full

Let's say you have a partition getting full and you need to move the redolog to another partition or nfs mount temporary to deal with the potential crisis that will happen when the partition becomes full. You'll need to reallocate the complete redolog/ directory and the archive subdirectory to the same partition because the roll over from redo.log to the archive directory happens with a rename function within the java code. This will require downtime since you'll need to move the actual redo.log file and zimbra can't be running while you do this. You can use a symlink to your new partition path. For example:

su - zimbra
zmcontrol stop
su -
 ** as root
mkdir /data/redolog
chown zimbra:zimbra /data/redolog
mount /dev/sdb1 /data/redolog
mv /opt/zimbra/redolog/* /data/redolog/
 ** or use rsync
rmdir /opt/zimbra/redolog
ln -s /opt/zimbra/redolog /data/redolog
ls -laR /data/redolog
** confirm ownership is with zimbra and double check zimbra can write in this directory
su - zimbra
touch /data/redolog/testfile
rm /data/redolog/testfile
zmcontrol start
Automatic Deleting Of Redo Logs On Rollover

This variable (zimbraRedoLogDeleteOnRollover) is set TRUE or FALSE.

zmprov gacf | grep zimbraRedoLogDeleteOnRollover

To modify it

zmprov mcf zimbraRedoLogDeleteOnRollover TRUE

Want To See What's In Redolog Files

This is for older versions of ZCS - newer versions should use zmredodump if it's available.

If you suspect there's too much redolog activity during a time window or have another need to inspect the contents of the redolog, dump it and examine it:

$ zmjava com.zimbra.cs.redolog.util.RedoLogVerify /opt/zimbra/redolog/redo.log > out.file

Pick the right redolog file, either redo.log or one of the files under archive/, based on timestamp.

zmplayredo And zmredodump

zmplayredo - Replaying Content From Any Redolog File

zmplayredo is a newer command, first introduced in 5.0.5 I believe. The mailbox has to be stop to run zmplayredo.

The help output from 6.0.8:

$ zmplayredo --help
usage: zmplayredo <options>
    --fromSeq <arg>         Replay from this redolog sequence (inclusive)
    --fromTime <arg>        Replay from this time (inclusive)
 -h,--help                  Show help (this output)
    --logfiles <arg>        Replay these logfiles, in order
    --mailboxId <arg>       Replay for this mailbox only
    --queueCapacity <arg>   Queue capacity per player thread; default=100
    --stopOnError           Stop replay on any error
    --threads <arg>         Number of parallel redo threads; default=50
    --toSeq <arg>           Replay to this redolog sequence (inclusive)
    --toTime <arg>          Replay to this time (inclusive)

Specify date/time in one of these formats:

    2010/11/19 13:55:08
    2010/11/19 13:55:08 802
    2010/11/19 13:55:08.802

Specify year, month, date, hour, minute, second, and optionally millisecond.
Month/date/hour/minute/second are 0-padded to 2 digits, millisecond to 3 digits.
Hour must be specified in 24-hour format, and time is in local time zone.
zmredodump - Replaying Content From Any Redolog File

zmredodump is a newer command and very useful. It does not require mailboxd to be stopped like zmplayredo does.

The help output from 6.0.8:

$ zmredodump --help
usage: zmredodump [options] <redolog file/directory> [...]
where [options] are:

 -h,--help        show this output
    --m <arg>     one or more mailbox ids separated by comma or white
                  space.  The entire list must be quoted if using space as separator.  If
                  this option is given, only redo ops for the specified mailboxes are
                  dumped.  Omit this option to dump redo ops for all mailboxes.
    --no-offset   don't show file offsets and size for each redo op
 -q,--quiet       quiet mode.  Only print the log filename and any errors.
                  This option can be used to verify the integrity of redologs with minimal
    --show-blob   show blob content.  Item's blob is printed, surrounded
                  by <START OF BLOB> and <END OF BLOB> markers.  The last newline before end
                  marker is not part of the blob.

Multiple log files/directories can be specified.  For each directory, all
redolog files directly under it are processed, sorted in ascending redolog
sequence order.
Using zmredodump To Get Message Blobs To Inject With zmlmtpinject - RFE

Please see:

Expand Ability To Parse Redologs - RFE

Please see:

Getting A Sequence or Time Variable For Restore Or Replay

You can see the changes within the redo logs with the command below. You can point it to any redolog.

zmjava com.zimbra.cs.redolog.util.RedoLogVerify /opt/zimbra/redolog/redo.log > out.file

You'll get output like this:

VERIFYING: redo.log
sequence: 59
open: 1
filesize: 512
serverId: d5c5d6a7-b82f-4c29-b0cd-91818057196b
firstOpTstamp: 1222385426273
lastOpTstamp:  1222385426273
version: 1.22

txn 1222383600.1 [PurgeOldMessages] ver=1.22, tstamp=1222385426273, change=20200, mailbox=1
txn 1222383600.1 [CommitTxn] ver=1.22, tstamp=1222385426329, mailbox=1, txnType=PurgeOldMessages
txn 1222383600.2 [PurgeOldMessages] ver=1.22, tstamp=1222385486337, change=13500, mailbox=3
txn 1222383600.2 [CommitTxn] ver=1.22, tstamp=1222385486351, mailbox=3, txnType=PurgeOldMessages
txn 1222383600.3 [PurgeOldMessages] ver=1.22, tstamp=1222385546357, change=20201, mailbox=1
txn 1222383600.3 [CommitTxn] ver=1.22, tstamp=1222385546383, mailbox=1, txnType=PurgeOldMessages
txn 1222383600.4 [PurgeOldMessages] ver=1.22, tstamp=1222385606391, change=13501, mailbox=3
txn 1222383600.4 [CommitTxn] ver=1.22, tstamp=1222385606404, mailbox=3, txnType=PurgeOldMessages
txn 1222383600.5 [PurgeOldMessages] ver=1.22, tstamp=1222385666416, change=20202, mailbox=1
txn 1222383600.5 [CommitTxn] ver=1.22, tstamp=1222385666428, mailbox=1, txnType=PurgeOldMessages
txn 1222383600.6 [PurgeOldMessages] ver=1.22, tstamp=1222385726435, change=13502, mailbox=3
txn 1222383600.6 [CommitTxn] ver=1.22, tstamp=1222385726459, mailbox=3, txnType=PurgeOldMessages
txn 1222383600.7 [PurgeOldMessages] ver=1.22, tstamp=1222385786476, change=20203, mailbox=1
txn 1222383600.7 [CommitTxn] ver=1.22, tstamp=1222385786486, mailbox=1, txnType=PurgeOldMessages
txn 1222383600.8 [PurgeOldMessages] ver=1.22, tstamp=1222385846493, change=13503, mailbox=3
txn 1222383600.8 [CommitTxn] ver=1.22, tstamp=1222385846506, mailbox=3, txnType=PurgeOldMessages
txn 1222383600.9 [PurgeOldMessages] ver=1.22, tstamp=1222385906739, change=20204, mailbox=1
txn 1222383600.9 [CommitTxn] ver=1.22, tstamp=1222385906775, mailbox=1, txnType=PurgeOldMessages
txn 1222383600.10 [PurgeOldMessages] ver=1.22, tstamp=1222385966944, change=13504, mailbox=3
txn 1222383600.10 [CommitTxn] ver=1.22, tstamp=1222385966963, mailbox=3, txnType=PurgeOldMessages
txn 1222383600.11 [PurgeOldMessages] ver=1.22, tstamp=1222386026972, change=20205, mailbox=1
txn 1222383600.11 [CommitTxn] ver=1.22, tstamp=1222386026990, mailbox=1, txnType=PurgeOldMessages

How Do I Figure Out Which Sequence or Time Variable To Use For Restore Or Replay

In 5.0.10+ we'll have a CLI wrapper (zmredodump) with a slightly different command line syntax, but the below long syntax works in earlier versions.

To locate the correct restore-to time, you have to start with an approximate time the message was added/deleted. Look at the redolog files. The filename contains the GMT time when the file was rolled over, which is roughly the tstamp of the last operation in the file. If your time data is accurate you can find the specific file. Or you have a range of files to examine.

Use the redolog verify tool to dump the contents into text form, the -m / --message option to show message body data:

zmjava com.zimbra.cs.redolog.util.RedoLogVerify -m <filename or directory> ... > out.file

If the message was deleted and you don't know the id, you must go by some other clue such as the subject. Search the file to locate your message. You can cut/paste the message and lmtp-inject it to recover the message. No need to go through with a restore if this is all you needed.

Are You Messages Really Gone - Things To Check If zmplayredo Isn't Doing What You Expect

Here's something I found out testing zmplayredo for a customer case. Testing on a ZCS 6.0.8 single ZCS server.

Created a test account and sent it one message that is in the Inbox. I delete the msg in zwc but don't purge the Trash - msg is in Trash now.

Log events of above action:

2010-10-27 15:07:13,375 INFO [btpool0-3://] 
- FF3.0 (Linux)/6.0.8_GA_2661;] mailop - Moving VirtualConversation (id=-257) to 
Folder Trash (id=3). Affected message ids: 257.

Stop mailboxd so I can use zmplayredo and then start mailboxd back up after zmplayredo is finished :

 zmmailboxdctl stop
 zmplayredo --mailboxId 17 --logfiles /opt/zimbra/redolog/archive/* --fromTime 20101027090000 --toTime 20101027140000
 zmmailboxdctl start

Log event for above:

2010-10-27 15:07:50,383 INFO [main] [] mailop - Adding Message: id=257, Message-ID=
<1604937630.920784.1288201208382.JavaMail.root@corp.zimbra.com>, parentId=-1, folderId=2, folderName=Inbox.
2010-10-27 15:07:50,404 INFO [main] [] RedoableOp - Message 257 is already in mailbox 17

Log into ZWC with the test account. The msg is not in the Inbox, but it's still in Trash folder. I purge it from Trash.

Log event for deletion.

2010-10-27 15:09:38,761 INFO [btpool0-2://]
- FF3.0 (Linux)/6.0.8_GA_2661;] mailop - Deleting Message (id=257).

Then I redo the stop/start of mailboxd and zmplayredo again.

 zmmailboxdctl stop 
 zmplayredo --mailboxId 17 --logfiles /opt/zimbra/redolog/archive/* --fromTime 20101027090000 --toTime 20101027140000 
 zmmailboxdctl start 

Log event for the above:

2010-10-27 15:10:29,192 INFO [main] [] mailop - Adding Message: id=257, Message-ID=
<1604937630.920784.1288201208382.JavaMail.root@corp.zimbra.com>, parentId=-1, 
folderId=2, folderName=Inbox.

Log back into ZWC with test account and now I can confirm that the msg is not in Trash and it is now showing in Inbox.

Gap In Redo Log

The error message from either a backup or restore command:

"Error occurred: Found gap in redo log sequence; missing 5965 through 6149;
To avoid future restore problems, discard all existing backups and take a
full backup of all accounts; If this error occurred during restore,
try the --ignoreRedoErrors option"

The output is pretty accurate in how to handle the situation.

If you get the error during a backup, the recommendation is to move your old backups out. The directories in /opt/zimbra/backup/sessions/* . You'll want to keep them around just in case and then proceed to do a full backup.

If you get the error during a restore, you would add the flag --ignoreRedoErrors to your restore command.

Another possible related issue is if your /tmp or /opt/zimbra/redolog/ is filling up.

Error Executing redoOp

Errors with restores that involve the message 'error executing redoOp' will not show up in the admin console but will when you attempt the restore from CLI. This can also be the cause when you use the RestoreToTime option from the admin console and it doesn't seem to work correctly - the restore stopping prematurely from the specified date/time.

I created the following RFE in regards to the admin console issue:

  • "Include --ignoreRedoErrors option and error feedback in Admin Console for restore"
    • http://bugzilla.zimbra.com/show_bug.cgi?id=52358
      • This could explain why your restore to time isn't working in the Admin Console but does from CLI when you see an error about redologs and then reattempt restore with the --ignoreRedoErrors and it works.

Another RFE that was made but marked as 'WONTFIX' that gives a background story to the issue is:

When you hit this error with your backup data during restore attempts, there's basically only a couple of options to handle this that are recommended by support:

  1. Try to get full backups of your accounts or the accounts in question and then test against them and the preceding incrementals after the full.
  2. Attempt restores via the CLI using the additional option of : --ignoreRedoErrors
  3. Do your restore against the latest full backup of the account in question and then use the zmplayredo command against the redologs in the incrementals and/or the /opt/zimbra/redologs/* directory . This will give you more control to walk the restored account up to the point in time you want it at. One should really read through the whole section above, Ajcody-Backup-Restore-Issues#Redolog_Files , to understand the whole concept of redologs and then the use of zmplayredo.

Generally, "fixing" the redolog itself is not an option.

Why Do My Fulls Not Report All Accounts?

Are you sure it was a full backup that was ran or just a full session that was generated from your incremental backup job? When an incremental is ran, it will create a "full" session for any new accounts it discovers after the last actual full backup job.

For example, here's a full session that was created by an incremental backup job:

Label: full-20081010.060126.559
Type: full
Status: completed
Started: Fri, 2008/10/10 01:01:26.559 CDT
Ended: Fri, 2008/10/10 01:01:28.988 CDT
Redo log sequence range: 705 .. 705
Number of accounts: 1

Label: incr-20081010.060009.420
Type: incremental
Status: completed
Started: Fri, 2008/10/10 01:00:09.420 CDT
Ended: Fri, 2008/10/10 01:01:26.413 CDT
Redo log sequence range: 700 .. 704
Number of accounts: 392

[zimbra@servername ~]$ du -sh /opt/zimbra/backup/sessions/full-20081010.060126.559
1.2M /opt/zimbra/backup/sessions/full-20081010.060126.559
[zimbra@servername ~]$ du -sh /opt/zimbra/backup/sessions/incr-20081003.060010.622
452M /opt/zimbra/backup/sessions/incr-20081003.060010.622

Notice the Start and End times, this will show that the full is related to the incremental job.

You'll want to run zmbackupquery against your full labels to see your "main" full backup session - assuming you can't simply guess based upon the cron entry for it [ su - zimbra ; crontab -l | grep backup ]. For example, to see all your fulls from today's date back to October 01, 2008 and the accounts within each session - you would do:

zmbackupquery -v --from 20081001.000000 --type full

The -v flag outputs the accounts, the --from uses YearMonthDay.HourMinuteSecond , and the --type can be full or incremental. To just see one particular sessions date, you would use the lb [label] flag:

zmbackupquery -v -lb [your full label, ex. full-20081001.000000]

Issues After /opt/zimbra/backup Became Full

Bugs/RFE's on this issue:

See alsoAjcody-Backup-Restore-Issues#Possible_Issue_That_A_Failed_Or_Interrupted_Backup_Causes and the Bugs & RFE's under that section.

You can run into numerous issues if you allow your backup directory to become full.

Confirm your /opt/zimbra/backup/accounts.xml is being updated after a backup. You might see that the newer account.xml* file is accounts.xml.new . This is a sign of problems.

Confirm that the files in /opt/zimbra/backup/tmp/* don't have 0 byte lengths. There might be files like 1.xml and 3.xml in there. If they show 0 bytes, you need to remove them. The backup/restore commands if the file exist and they are empty. Your errors might look like this:

[zimbra@mailb ~]$ zmrestore -a USER@DOMAIN -ca -pre restore_
Error occurred: system failure: Unable to parse XML file /opt/zimbra/backup/tmp/restore/shared_blobs/1.xml

If you tried doing restores (redirected -ca -pre) before to clear up the above issues, you might find you can't do a successful restore AND you can't delete the account afterwards.

If you get errors like :

zmprov da restore3_tester3@XXXXXX.edu
ERROR: service.FAILURE (system failure: writing new mailbox row for account 89e7d9f4-013e-4cf1-a352-7b2f0a00d5af)

zmprov gmi restored_USER
ERROR: service.FAILURE (system failure: writing new mailbox row for account 56a7f654-f85b-45cc-931a-81d9bb9076bf)

You'll need to delete the account via a ldapdelete command.

Query And Stopping A Backup or Restore In Progress

Please see the url below for Backup In Progress - topic name "Aborting Full Backup In Progress":

  • A matter of using zmbackupquery & zmbackupabort .

Please see url below for Restore In Progress - topic name "Stopping a Restore Process" :

A -del Delete In Progress

Please see the bug / rfe's I filed

zmbackupabort syntax

From ZCS 6.0.8

$ zmbackupabort -h
usage: zmbackupabort <options>
 -d,--debug          Display diagnostics for debugging purposes.
 -h,--help           Displays this help message.
 -lb,--label <arg>   Label of full backup set to abort.
 -r,--restore        Abort the restore in progress.
 -s,--server <arg>   Mail server hostname. Default is localhost.
 -t,--target <arg>   Backup target location (default

zmbackupquery syntax

From ZCS 6.0.8

$ zmbackupquery -h
usage: zmbackupquery <options>
 -a,--account <arg>   Account email addresses seperated by white space or
                      "all" for all accounts.
 -d,--debug           Display diagnostics for debugging purposes.
    --from <arg>      List backups whose start date/time is at or after
                      this date/time.
 -h,--help            Displays this help message.
 -lb,--label <arg>    The label of full backup to query.
 -s,--server <arg>    Mail server hostname. Default is localhost.
 -t,--target <arg>    Backup target location (default
    --to <arg>        List backups whose start date/time is at or before
                      this date/time.
    --type <arg>      Backup set type to query.  "full" or "incremental";
                      both if unspecified.
 -v,--verbose         Show account list in each backup.

Specify date/time in one of these formats:

    2010/11/19 14:06:22
    2010/11/19 14:06:22 923
    2010/11/19 14:06:22.923

Specify year, month, date, hour, minute, second, and optionally millisecond.
Month/date/hour/minute/second are 0-padded to 2 digits, millisecond to 3 digits.
Hour must be specified in 24-hour format, and time is in local time zone.

Restore To Time Problems

There's seems to be some syntax issues when using this variable. Please review the following to confirm your syntax.


The gist, you MUST use the -lb full-200xxxxxx option when your trying to restore anything that ISN'T meant to include the latest information of the mailbox. The -lb argument should specify a full backup that took place prior to the time of the backup you wish to restore.


zmrestore -a USER@DOMAIN.com -restoreToIncrLabel incr-20080731.060007.644 -lb full-20080726.050017.306 -br -ca -pre restore_
zmrestore -a USER@DOMAIN.com -restoreToTime 20080801011800 -lb full-20080726.050017.306 -br -ca -pre restore_

Backup label (-lb) for fulls can be found by:

zmbackupquery | grep full

Backup labels (-restoreToIncrLabel) for incrementals can be found by:

zmbackupquery | grep incr

Important Options You Might Want Or Need To Include

--ignoreRedoErrors : If you attempt a restore and you see an error about problems related to playing the redolog, you'll want to run the restore command again and include this option.

--skipDeletes : Please see http://bugzilla.zimbra.com/show_bug.cgi?id=31824#c5 for details on this.

-t /path/to/backup_dir : If you are restoring from another backup directory besides your current default path.

Variables that are asking for TIME rather than LABELS should follow this syntax (from zmrestore --help):

Specify date/time in one of these formats:

    2008/08/06 09:55:50
    2008/08/06 09:55:50 572
    2008/08/06 09:55:50.572

Specify year, month, date, hour, minute, second, and optionally millisecond.
Month/date/hour/minute/second are 0-padded to 2 digits, millisecond to 3 digits.
Hour must be specified in 24-hour format, and time is in local time zone.

Bugs And RFE's To Review

Update - Bug/RFE's I Filed Against ZCS 8.6

Older Bug/RFE's

Restore An Individual Message

The zmrestore command is at a mailbox level.

An RFE was filed already to expand this. It is currently targeted for the Helix release.

A way around the current limitations would be to use lmtpinject. Please see the following for details on that:

The difficultly would be determining the message your trying to find within your backups that was "deleted" in prod.

User Deleted A Bunch Of Data And Notified You Hours Later Wanting It Restored

  1. To determine the time of the delete, use the zmredodump command.
    • You'll use this "time" for the restore command.
    • Example:
      • $ zmprov gmi USER@DOMAIN.com  ::: gives me mailboxid 17
      • $ zmredodump --show-blob --m 17 /opt/zimbra/redolog/ | grep Delete  ::: returns:
        • [0000f311 - 0000f350: 64 bytes; tstamp: 2010/11/19 11:21:30.852 CST] txn 1290033504.2544 [DeleteItem] ver=1.28, tstamp=1290187290852, change=3913, mailbox=17, ids=[304, 308], type=5
        • [0000f351 - 0000f382: 50 bytes; tstamp: 2010/11/19 11:21:30.869 CST] txn 1290033504.2544 [CommitTxn] ver=1.28, tstamp=1290187290869, mailbox=17, txnType=DeleteItem
  2. See Ajcody-Backup-Restore-Issues#Restore_To_Time_Problems about setting up the proper restore syntax.
    • You'll want to include the -ca -pre restore_ options since we'll first get the old data into a temporary account.
    • Remember to use zmbackupquery -a USER@DOMAIN.com --type full to find what full label will be needed for the restore.
    • Also, if you need/want the incremental label, do a zmbackupquery -a USER@DOMAIN.com --type incremental .
      • Example:
        • $ zmbackupquery -a USER@DOMAIN.com --type full  ::: shows the latest as being:
          • Label: full-20101119.070018.607
          • Type: full
          • Started: Fri, 2010/11/19 01:00:18.607 CST
          • Ended: Fri, 2010/11/19 01:00:29.751 CST
          • Acct ID: f33d6daf-8875-4496-8bee-6df345f295e7
    • Do the restore now with the proper information for the variable flags.
      • Note: You might need to include -br , -rf , or neither depending on the time frames involved.
      • Example:
        • $ zmrestore -a USER@DOMAIN.com -restoreToTime 20101119112000 -lb full-20101119.070018.607 -ca -pre restore_
      • It would be best to log into the restore_USER@DOMAIN.com account to confirm the data is as you expect it.
  3. Use zmmailbox with the getRestUrl option against the "restore_USER" account now to export the data.
    • Examples:
      • Export ALL the 'old' data from the restored account
        • $ /opt/zimbra/bin/zmmailbox -z -m restore_USER@DOMAIN.com getRestURL "//?fmt=tgz" > /var/tmp/user-export.tgz
      • You can also export SOME of the 'old' data from the restore account using other options. One option is with the before and after variables. NOTE - We have to set the query string as a variable to get around some of the shell issues.
      • For example:
        • $ query='before:11/20/2010 after:11/1/2010'
          • ZCS5 might require you to have a %20 rather than the actual space character.
        • $ /opt/zimbra/bin/zmmailbox -z -m restore_USER@DOMAIN.com getRestURL "//?fmt=tgz&query=$query" > /var/tmp/user-export.tgz
          • Note': the $query above will be the expansion of the variable you set prior in the shell, query='before:11/20/2010 after:11/1/2010 .
  4. Use zmmailbox with the postRestUrl option to IMPORT the RESTORE_USER account into the "USER" account now with appropriate options.
    • Examples:
      • $ /opt/zimbra/bin/zmmailbox -z -m USER@DOMAIN.com postRestURL “//?fmt=tgz&resolve=replace” /var/tmp/user-export.tgz
    • Note : A critical option in the above command is the &resolve=replace one. There are various ways you can handle the importing of data. Please review the following to determine what is best for you needs.

Restore Deleted Items - skipDeletes Option - ZCS6+

From the RFE comment:

        Added new option --skipDeletes to zmrestore.  If specified, skip 
        over delete operations during redo replay.  Delete ops are:

        DeleteItem (hard delete)
        MoveItem, if moving item to Trash folder

        Skipping these deletes can lead to other problems later, such as 
        conflicting paths, but it is assumed the priority is recovering as 
        much data as possible when using this option.

Restore Account Not Yet In Backups

Please see:

How To Restore To Events Within The Same Day

Work in progress and investigation.

Users Trash Items

User Ability To Recover Trash Purge

Please see RFE:

Retention Policy About Purges

Please see Mailbox_Purge

Can't Restore Or Find An Account That Was Renamed

When an account is "renamed", the old account name will no longer be "found" is your "default" type restore or backup queries. This can cause some confusions when one needs to restore to a time frame of when the account was under it's older name.

Restoring From CLI

First, identify what sessions labels hold the differing account names:

zmbackupquery -a USER@DOMAIN
zmbackupquery -a USER-RENAMED@DOMAIN

You should know have what you need to do follow the steps in the Ajcody-Backup-Restore-Issues#Restore_To_Time_Problems section.

Restoring From Admin Console GUI

I'll give a detail explanation of the situation when working around restores of renamed accounts in the admin console [web GUI]:

  1. If you in the GUI goto the "Restore" button, it first asks for an account rather than giving an option for date/time/session. I think you already stated, that "renamed" accounts don't show up in this query window. Therefore, one wouldn't really progress to the next window that would allow you to change the backup session label.
  2. They way you get around this is, you actually double click on the full session listing that you see on the backup page in the admin UI. This will bring you to another page, that is specific to that session. In there, you should see the old account name prior to the rename. You can then highlight that account listing and click on the "Restore" button. This will bring up the restore dialog, which will now have the date/time/session label auto-filled out.

Quota Is Stopping A Redirected Restore


Reasoning for need, maybe the msg files coming from the restore are no longer "shared message blobs" and therefore increase the mailbox to a size it wasn't in the past. Changes to HSM maybe?

I think I'll need to create a RFE about adding an option to the zmrestore command to also include an option to set the COS value on a created account. Until then... Create a new COS and set it up to NOT have any quota. Once you kick off the backup and you see the account is created you can then apply the COS to that account. Call the cos something like no-quota. Here's the steps below.

To copy your cos (assuming your quota cause is the default one, change default to match your production cos your using).

To see all cos's

zmprov gac

Copies cos called default to cos named no-quota

zmprov cpc default no-quota

To remove quota to the new cos no-quota

zmprov mc no-quota zimbraMailQuota 0

To confirm

zmprov gc no-quota | grep -i quota

Now you would start your redirected restore and once you see the account is created, run the below in a separate shell. (example)

zmrestore -a USER@DOMAIN.com -restoreToTime 20080801011800 -lb full-20080726.050017.306 -br -ca -pre restore-

Once restore has kicked off - To apply the no-quota cos to the restored account:

zmprov sac restore-USER@DOMAIN.com no-quota

Restore Of Non-Account Items - Example - COS DL Etc

Cos and DL's are ldap entries basically.

From Backup and Restore

  • zmrestoreldap. This command restores the complete LDAP directory server, including accounts, domains, servers, COS and other data."

You'll see in a DR process, Network_Edition_Disaster_Recovery , that the zmrestoreldap is done before the zmrestoreoffline .

  • zmrestoreldap doesn't have options that allow specific items to be restored (COS, DL's, etc.). It only has option for named accounts (-a). One could try a ldapadd with a ldif of the COS or DL details. One could also take the information on the COS or DL within the ldap file in the backup session to at least have all the variables to manually add it back (via the zmprov command). Your looking at the backups on the LDAP master if your in a multi-server configuration.

Start of ldap entry example to search for:

  • Cos example
    • dn: cn=default,cn=cos,cn=zimbra
  • DL example
    • dn: uid=dl-group,ou=people,dc=mail,dc=domain,dc=com

To compare a current DL with past details, just save out the ldap entry from the backup to a txt file. And then do:

zmprov gdl maillist@domain.com

Make the necessary changes after comparing the two.

Restoring A Calendar (ics)

There seems to be a bug or odd expectation on how this command is currently working. If the appointment exists in the Calendar and the time is different with the same appointment in the ics file your importing - the time of the appointment will not change to the imported ics one. If you delete the event first, then the imported appointment will reflect the correct time.

Here's what I did to reproduce this situation. It seems this has been true for sometime, customer was on 4.5.11 and I was on 5.0.8

Created test account and made two appointments on friday - 9am and 4pm.
 Did a full backup.
 Restored test account to restore_user
 Ran :
     zmmailbox -z -m user@domain.com gru /Calendar > /tmp/calendarA.ics
     zmmailbox -z -m restore_user@domain.com gru /Calendar > /tmp/calendarB.ics
 And then
     diff /tmp/calendarA.ics /tmp/calendarB.ics [no differences]

 Now some tests.
 As user, I deleted the two appointments and then:
     zmmailbox -z -m user@domain.com pru /Calendar /tmp/calendarB.ics
 Refreshed Calendar as User in webclient.
     9am and 4pm appointment shows up.
 I then moved in the webclient the 9am appointment to 11am
 Did another restore:
     zmmailbox -z -m user@domain.com pru /Calendar /tmp/calendarB.ics
 Refreshed Calendar as User in webclient.
     11am and 4pm appointment shows up.
     ** The restore did not move the 11am appointment back to the 9am slot as in /tmp/calendarB.ics
     ** Assumption, this process will not over-write an appointment if it's there - it does not look to the time.
        Let's do a diff of the state of the calendar
        zmmailbox -z -m user@domain.com gru /Calendar > /tmp/calendarC.ics
        diff /tmp/calendarB.ics /tmp/calendarC.ics
          The DTSTAMP and SEQUENCE shows the difference in the time.
 If I delete the 11am appointment and then do the calendarB.ics restore the appointment shows up again at 9am.

I see this same behavior if I also use the web interface to export/import the calendar between the restored account and user one. Even when I import it into a NEW calendar, it even changes the two appointments to reflect the new calendar rather than the default one.

One Fix

One fix, if the situation allows, is to purge the current Calendar and then import the full ics file. This would be done like this:

zmmailbox -z -v -m ajcody@mail3.internal.homeunix.com ef /Calendar
* ef is for emptyFolder  zmmailbox help folder
*webclient shows all events gone but Calendars are still listed.
zmmailbox -z -v -m ajcody@mail3.internal.homeunix.com pru /Calendar /tmp/calendarB.ics
*webclient show all events with times as expected.

Second Fix

Haven't tried yet, but someone said you should be able to adjust the SEQUENCE number in the appointment and the import process will use that data (data/time) of the newer sequence number appointment.

Archiving User Accounts Out Of Production Use In Zimbra

Backup - Restore Method

To "archive" the user data with the ability of Zimbra later restoring if needed, one would rely on the backup/restore tools.

For example:

zmbackup -f -a user@domain

If you want to make this more "intelligent" later, one could do this:

mkdir /nfs-mount/archived-users/[user-name]
zmbackup -f -a userA@domain -t /nfs-mount/archived-users/userA-domain/

This would give some "intelligent" information for recovery later. Confirms it was an archived account, shows the "real" username, the date of the full backup, and so forth.

ls /nfs-mount/archived-users/userA-domain/
ls /nfs-mount/archived-users/userA-domain/full-20081104.131643.006
  accounts/ session.xml shared_blobs sys

If one would need to "restore" the account later, it would consume a license if the account was "deleted".

zmrestore -a userA@domain -ca -pre restored- -t /nfs-mount/archived-users/userA-domain/ -lb full-20081104.131643.006
Setup A Secondary Zimbra Box For Restores Of Archive Accounts

Your Zimbra license can be installed on multiple machines. One idea that might prove useful in handling these "archive" accounts for those situations when you need to investigate something is to setup a "archive" Zimbra box. You'll want to isolate this box from any "production" use. It will need to be configured to have the "domains" of the archive accounts. You can then use this box to restore the "archive" account and then use the administration tools to investigate the user data.

Use Of REST And Other Tools For Specific User Data

The following page, User_Migration , will shows numerous examples of how to export different types of data from a user account into a neutral file format that one could use for "archive" purposes.

Use Of The REST Command To Export ALL User Data - Version Dependant, 5.0.9+ [I believe]

Example of the commend syntax:

 /opt/zimbra/bin/zmmailbox -z -m user@domain.com getRestURL “//?fmt=tgz” > /tmp/account.tgz

Backup And HSM

Please see Bugs/RFE:

Scripting Out Individual Backups Of Accounts

If you want to do individual backups of accounts using a for-loop, for example, you might want to include the -sync option from zmbackup. zmbackup without this will normal give an error as it passing the next zmbackup command stating that there's a current backup in progress.

Example command in some for-loop script is [without the -sync option]:

sudo -u zimbra /opt/zimbra/bin/zmbackup --fullBackup --account "$acct" --server "$host" --target "$dest"

zmbackup --help shows:

--sync  -sync Runs full backup synchronously. 

The admin guide mentions:

"Full backup is usually run asynchronously. When you begin the full backup, the label of the ongoing backup process is immediately displayed. The backup continues in the background. You can use the zmbackupquery command to check the status of the running backup at any time."

I couldn't find any other indication beyond that to explain in more details the purpose of that flag. But from what is stated above, it does look like the -sync flag will resolve the issues of "backup in progress" when scripting out multiple zmbackup commands.

If not, you could query for "Status: in progress" from the zmbackupquery command.

You can give the zmbackupquery command flags for date/time, label, account, -t target, and so forth [Do a zmbackupquery --help to see the options format].

Finding Message Blob In Users Backup

Get the zimbraId of the user, this is what is used in the backup session directory layout:

$ zmprov ga admin@`zmhostname` zimbraId
# name admin@zcs806.DOMAIN.com
zimbraId: e46a828b-cdda-4635-98ab-31b8ac0129b6

Get the mailboxId, this is what is used in the zmvolume primary message volume directory layout:

$ zmprov gmi admin@`zmhostname`
mailboxId: 1
quotaUsed: 77794874

Finding a message in the primary message volume:

$ cd /opt/zimbra/store/0/1/msg/9/

$ ls -la 40511-49376.msg
-rw-r----- 1 zimbra zimbra 1372 Mar 16 02:01 40511-49376.msg

Finding the users directory in a full backup session directory layout:

$ find /opt/zimbra/backup/sessions/full-20140426.080019.153/ -name e46a828b-cdda-4635-98ab-31b8ac0129b6 -print

Locating the message in the backup session - this example is where the backups are using the zip option:

$ cd /opt/zimbra/backup/sessions/full-20140426.080019.153/accounts/e46/a82/e46a828b-cdda-4635-98ab-31b8ac0129b6/blobs/

$ ls
blobs-1.zip  blobs-2.zip  blobs-3.zip  blobs-4.zip

$ for i in `ls /opt/zimbra/backup/sessions/full-20140426.080019.153/accounts/e46/a82/e46a828b-cdda-4635-98ab-31b8ac0129b6/blobs/`; do echo $i ; unzip -l $i | grep 40511-49376.msg; done

     1372  04-26-2014 01:00   1/9/sha256_bKH8aeWJSh9bhzv5zsXrweA7+jZ88NUllkfV+7m9cJo=_40511-49376.msg1

Backing Up Backups - 3rd Party Tools And Software - Dealing With Directories With Hard Links

Description Of Hard Links

Zimbra uses hard links and special attention needs to be given to this fact. See hard links if your not familiar with hard links and their difference to symbolic links. Not all 3rd party backup software will handle or respect hard links. Many unix commands will need special flags to maintain hard links. When hard links are respected and also "copied" to the new location you could find your data usage become a large multiplication of the original size.

An good thread I found on the topic of preserving hard links for copy/move/backup operations is here:

I'll be summarizing the comments and including additional information I find on the topic below, based upon the command or software being used.

Zimbra and Single Instance Storage - Hard Links

If hard links are possible, we use them. The message must be identical and on to users on the same mailstore and stored on zmvolumes where hard linking is possible. Hard links exist only on the same partition. Postfix has a veriable, default_destination_recipient_limit , which will cause large recipient emails to be delivered in a way where they aren't identical [ see Ajcody-Hardlinks-And-Postfix-default_destination_recipient_limit for more details ].

Mike gives a good description of it here:

Redoing Hard Links RFE

For 8.0.2+ , see this option - zmdedupe :

Easy Way To SEE Hard Links In Use

I sent a test email with 5 accounts listed in the To field, that is shown below with the inode listing of 2133394. the first column of the ls [because of the i option] is listing the inode number of the file. I included the -type f because the . and .. will show directories using the same inode as the 'name' of the directory as listed.

# pwd

# find . -links +1 -type f -exec ls -lai '{}' \; | sort
2133387 -rw-r----- 2 zimbra zimbra 1600 May  7 05:30 ./0/13/msg/0/281-4320.msg
2133387 -rw-r----- 2 zimbra zimbra 1600 May  7 05:30 ./0/27/msg/0/261-3021.msg
2133394 -rw-r----- 5 zimbra zimbra 1789 May  7 05:53 ./0/13/msg/0/282-4322.msg
2133394 -rw-r----- 5 zimbra zimbra 1789 May  7 05:53 ./0/14/msg/0/412-8476.msg
2133394 -rw-r----- 5 zimbra zimbra 1789 May  7 05:53 ./0/15/msg/0/284-9052.msg
2133394 -rw-r----- 5 zimbra zimbra 1789 May  7 05:53 ./0/25/msg/0/260-4021.msg
2133394 -rw-r----- 5 zimbra zimbra 1789 May  7 05:53 ./0/27/msg/0/268-3033.msg
2133404 -rw-r----- 2 zimbra zimbra 1518 May  7 05:22 ./0/13/msg/0/280-4318.msg
2133404 -rw-r----- 2 zimbra zimbra 1518 May  7 05:22 ./0/27/msg/0/260-3018.msg
Why Does A Message Blob With The Same Message-ID Have Multiple Inodes

Most likely, the message has a large recipient list and because of a variable in postfix, default_destination_recipient_limit , causes multiple deliveries of the message to have slight differences between them - for example, the times in the headers.

Please see the following:

Hard Links Used Within Zimbra Backup Directory - Sessions

Note, this describes an example when zimbraBackupMode equals Standard rather than Auto-Grouped

In regards to the backup sesions, hard links are only used for data that is in the shared_blobs directories from the various full backup sessions you have.

If you goto [below is an example used throughout]:

cd /opt/zimbra/backup/sessions/full-[some full dir]/shared_blobs/[some path]/t9RRjTIdwAZ3k,iJSWo0DxFKCbs=/

Then do a:

ls -li blob.dat
3869621 -rw-r----- 2 zimbra zimbra 5419 Jul 15 01:00 blob.dat

The first number is the inode of the file, which will be the same for all other hard links in use.

cd /opt/zimbra/backup/sessions
find . -inum 3869621 -print
  /opt/zimbra/backup/sessions/full-[some full dir #1]/shared_blobs/[some path]/t9RRjTIdwAZ3k,iJSWo0DxFKCbs=/blob.dat
  /opt/zimbra/backup/sessions/full-[some full dir #2]/shared_blobs/[some path]/t9RRjTIdwAZ3k,iJSWo0DxFKCbs=/blob.dat

Notice also, that the directory path naming scheme is also used for the user backup paths. The directory name that the blob.dat is in, will also be used in the user paths.

For example:

/opt/zimbra/backup/sessions/full-[some full dir]/shared_blobs/[some path]/t9RRjTIdwAZ3k,iJSWo0DxFKCbs=/blob.dat

if I then goto and do:

cd /opt/zimbra/backup/sessions/full-[some full dir]/
find . -name t9RRjTIdwAZ3k* print

I'll see matches in various user backups, with:

full-[some full dir]/accounts/c2e/23b/[zimbra ID of user]/blobs/3/0/t9RRjTIdwAZ3k,iJSWo0DxFKCbs=260-2041.msg4

Note though, those "user" files are not links [hard or soft] and they are of zero length in size.

Reference To Bugs Tied To Backups And Links:

Bacis Unix Commands


cp [copy] Man page:

  • http://linux.die.net/man/1/cp
    • -a, --archive : same as -dpR
      • -d : same as --no-dereference --preserve=link
      • -p : same as --preserve=mode,ownership,timestamps
      • -R, -r, --recursive : copy directories recursively
  • -l, --link : link files instead of copying
  • -L, --dereference : always follow symbolic links
  • -P, --no-dereference : never follow symbolic links
  • -p : same as --preserve=mode,ownership,timestamps
    • --preserve[=ATTR_LIST] : preserve the specified attributes (default: mode,ownership,timestamps) and security contexts, if possible additional attributes: links, all
    • --no-preserve=ATTR_LIST : don't preserve the specified attributes
  • -s, --symbolic-link : make symbolic links instead of copying

No explicit mention of "hard links".

Formats from blog url above [I'll test these later and confirm/deny]:

  • cp -rpv
  • cp -av --preserve=all . /mnt/new
  • Simple `cp -a` using cp (GNU coreutils) 5.97 on my debian does the job quite nicely, I just checked. No need for the --preserve=all option, -a implies --preserve=link. It didn't seem to take too long either, but I would be surprised if it was very much better than rsync. Much easier to remember though.

mv [move] Man page:


tar [tape archive] Man page:

  • http://linux.die.net/man/1/tar
    • --check-links : warn if number of hard links to the file on the filesystem mismatch the number of links recorded in the archive
    • -h, --dereference : don't dump symlinks; dump the files they point to
      • No explicit reference to "hard links".

Please see the following on hard-links and GNU/Tar:


dd [copy and covert - cc was reserved for the C complier] Man page:

No explicit mention of "hard links" in Man page.

rsync and nice

rsync and nice Man page:

A reasonable syntax to use for rsync is [ taken from Ajcody-Notes-Server-Move#The_Actual_Steps ]:

nice +19 rsync -avzH -e ssh --progress /opt/zimbra/ root@NEWHOSTIP:/opt/zimbra


nice -20 rsync -avzH -e ssh --delete --progress /opt/zimbra/ root@NEWHOSTIP:/opt/zimbra 
Notice the use of --delete in the last one.
find And cpio

find and cpio Man pages:

Possible syntax use to try [recursive copy of current working directory] :

mkdir /path/to/dest
cd /path/to/DATA
find . -print | cpio -Bpdumv /path/to/dest

Issue to note: "...cpio didn't properly preserve timestamps on directories."

dump And restore

dump and restore Man pages:

ditto - For Mac

ditto Man page:


xfs_copy Man page:

LVM Tricks
LVM Snapshots

See Back Up (And Restore) LVM Partitions With LVM Snapshots

SAN Snapshot

Possible note of caution I've seen from someone, "When implementing snapshots for ZCS, you should do the snapshot across all ZCS LUNs for a single host at the same time using a consistency group (for netapp, I believe this means cg-start/cg-commit)."

Please see:

Cloud Backups
Amazon S3 , Amazon EC2 , SecoBackup And/Or Tar

I've not used Amazon S3 or SecoBackup. I have no idea about the pricing structure of Amazon S3 and how differing solutions might cause price differences. What I think would be a reasonable approach:

  • Adjust zimbra cron to:
    • run zmbackup as normally scheduled but then include:
      • tar and gzip "new" backup that was made to a "staging" partition.
      • Setup Secobackup [CLI method for cron] to then copy this tar'd/gzip'd file to the Amazon S3 cloud.
      • Remove local tar'd/gzip'd file from staging partition.

I purpose the tar'd/gzip'd step because I doubt there's a way to avoid the hard link issue with SecoBackup/Amazon S3. Why pay multiple times for the same data?

Some information a customer reported to me:

S3 does not work as a normal filesystem and you cannot mount it; hence 
it wouldn't normally work.

However, there are various projects out there which let you use S3 as 
a local POSIX-compliant file system.

Possible options:

jungle disk
Amazon EBS

To cut a long story short, PersistentFS came out on tops - it worked 
extremely well - however did not work with Zimbra at all once I set it 
up as the store (/opt/zimbra/store)

The problem is that the filesystem while it is POSIX compliant does not 
have support for hard linking (Which is what Zimbra does with tmp incoming 
messages to the store).  -- [bug 43019 below]

So, overall it's not really possible to do it right now with S3.

They should have hard link support soon.


3rd Party Backup Tools And Software - Generally Not Apart Of The Basis OS

Amanda And Zmanda

Zmanda Home Page:

Amanda Home Page - Advanced Maryland Automatic Network Disk Archiver:

Found this, "Hard links. Maintains the integrity of hard links during backup."


Arkeia Home Page:

Found this, "The Arkeia solution accommodates full and incremental backups, scheduled or on demand, and preserves directory structure, registry, symbolic links and special attributes."

Also, a customer supplied what Arkeia's support sent them to the question. Here is that response:

When Arkeia encounters a hard link (a regular file with more than one reference to it):
  • if it is the first time we see this link, we backup the file and keep the inode number and the path to the file in a memory hash table.
  • if we have already seen the file, we backup the fact that it is a hard link and the target of the link. The files data are not backed up again.

Backup Exec - Symantec / Veritas

Backup Exec Home Page:

Unable to find any reference about the Linux/Unix Agent and the Backup Exec server being able to handle or not symbolic and hard links.

Backup Exec "server" is only available on Windows. One might inquire with Symantec if you can "swap out" your current investment in "Backup Exec" and use their NetBackup product. This supports hard links and the "server" can run on Windows or other *nixes. See Ajcody-Backup-Restore-Issues#NetBackup_-_Veritas.2FSymantec


Bacula Home Page:

Found this:

When enabled (default), this directive will cause hard links to be backed up. However, the File daemon keeps track of hard linked files and will backup the data only once. The process of keeping track of the hard links can be quite expensive if you have lots of them (tens of thousands or more). This doesn't occur on normal Unix systems, but if you use a program like BackupPC, it can create hundreds of thousands, or even millions of hard links. Backups become very long and the File daemon will consume a lot of CPU power checking hard links. In such a case, set hardlinks=no and hard links will not be backed up. Note, using this option will most likely backup more data and on a restore the file system will not be restored identically to the original.



BRU Home Page:

Found this in one of their manuals, you should confirm with them based upon the product version you'll be using:

Special Files - BRU will save and restore all types of filesystems and files with their proper ownership, access attributes, creation dates, and modification dates. BRU can be used to move an entire directory hierarchy from one system to another, with all files, including directories, block special files, character special files, fifos, hard links, and symbolic links reproduced with all attributes intact.
Lone-Tar - Lone Star Software Corp.

Lone-Tar Home Page:

Found the following references, though it's not explicit in stating "hard links" :

BACKUP FEATURES - Backs up everything including device files, empty directories, links, symbolic links, Virtual Files and NFS mounted file systems.
RESTORE FEATURES - Fast Seek File restore of files, sym-links, device files, and linked files.


NetBackup - Veritas/Symantec

NetBackup Home page:

Found the following:

Hard links to directories :
On most UNIX systems, only the root user can create a hard link to a directory. Some systems do not permit hard links and many vendors recommend that these links be avoided. NetBackup does not back up and restore hard-linked directories in the same manner as files:
  • During a backup, if NetBackup encounters hard-linked directories, the directories are backed up once for each hard link.
  • During a restore, NetBackup restores multiple copies of the hard-linked directory contents if the directories do not already exist on the disk. If the directories exist on disk, NetBackup restores the contents multiple times to the same disk location.
Hard links to files :
A hard link differs from a symbolic link in that a hard link is not a pointer to another file. A hard link is two directory entries that point to the same inode number.
If the backup selection list includes hard-linked files, the data is backed up only once during a backup. NetBackup uses the first file name reference that is found in the directory structure. If a subsequent file name reference is found, it is backed up as a link to the name of the first file. Backup up only the link means that only one backup copy of the data is created, regardless of the number of hard links. Any hard link to the data works.
For more information and examples, see “Hard links to files (NTFS volumes or UNIX)” on page 173.


NetVault - ORBiT

NetVault Home page:

Found this in their "NV Backup Administrators Guide" - pdf only:

[screen shot of GUI check box] The Attempt to Restore Hard Links’ option as revealed in the Restore Options tab on a Linux/UNIX-based version of the File System Plugin.
Attempt to Restore Hard Links (Linux/UNIX-based O/S, ONLY) - During a backup, when the first occurrence of a hard link is found, the complete data will be backed up. For all other occurrences, only the link is backed up. This data can only be restored when the first occurrence exists; trying to restore subsequent occurrences without the presence of the first causes the job to fail. Selecting this option will attempt to locate the full sequence so that all occurrences of the hard link will be restored.

rsnapshot Home page:

Tivoli - IBM

Tivoli Home Page:

Found the following:

Understanding how hard links are handled
When you back up files that are hard-linked, Tivoli Storage Manager backs up each instance of the linked file. For example, if you back up two files that are hard-linked, Tivoli Storage Manager will back up the file data twice.
When you restore hard-linked files, Tivoli Storage Manager attempts to reestablish the links. For example, if you had a hard-linked pair of files, and only one of the hard-linked files is on your workstation, when you restore both files, they will be hard-linked. The files will also be hard-linked even if neither of the files exists at the time of restore, if both of the files are restored together in a single command. The one exception to this procedure occurs if you back up two files that are hard-linked and then break the connection between them on your workstation. If you restore the two files from the server, Tivoli Storage Manager will respect the current file system and not re-establish the hard link.
Attention: If you do not back up and restore all files that are hard-linked at the same time, problems will occur. To ensure that hard-linked files remain synchronized, back up all hard links at the same time and restore those same files together.


Other Related Items


freedup Man page:

freedup Homepage:

I've never used this tool, but from the description it seems it might come in handy for some circumstances.

"Establishes hard or symbolic links between identical files. Search all given file system trees for identical files and link them to the most frequently referenced inode or if equally referenced to the inode of the first file tree. If the devices differ a symbolic link is used instead of a hard link. Symbolic links will not replace files, when at least one of the directory trees is not starting with a '/'."
Tape Devices

Many times, "drivers" aren't needed for tape devices for linux, but many administrators are unaware of this and never give it a test. Instead, they just assume the device doesn't work and the tape vendor isn't supporting it because they "didn't publish" a driver for it.



Jump to: navigation, search