Hi all.
I have a question that I seem to be having trouble getting an answer from support on but it seems so stupid simple or at least something has to come up.
We are running Zimbra Network Edition 8.8.15 P33.
The situation is I need to restore the data that was on a HSM volume. we had the main volume id 1 and a HSM S3 volume as volume id 4. This S3 volume is no longer operational. We have been backing up things using the NG Backup to a separate volume, which is fine. So I am reading this article to ensure I know how to do a restore.
https://wiki.zimbra.com/wiki/Zimbra_NG_ ... ing_BLOBs)
The command is seemingly straightforward:
zxsuite backup dorestoreblobs [volume id] dryrun false
But what I cannot get clarity on is, if I am restoring volume ID 4 but need to restore it to volume ID 1, am I specifying 1 or 4 in the command for the volume ID? Since it will no doubt take days to complete, I a) don't want to get it wrong and b) don't wnat to mess up anything further.
I appreciate any assistance you might be able to offer.
Thanks,
Rick
Restoring with doRestoreBlobs
-
L. Mark Stone
- Ambassador
- Posts: 2828
- Joined: Wed Oct 09, 2013 11:35 am
- Location: Portland, Maine, US
- ZCS/ZD Version: 10.0.9 Network Edition
- Contact:
Re: Restoring with doRestoreBlobs
My understanding, which I think comports with https://docs.zextras.com/suite/html/cli ... storeblobs, is that the volume ID you specify is the target volume for the blob restore.
Consider that if you have a lost or damaged volume, when you create a new/replacement volume it will likely have a different ID, so it makes sense that you would want to specify the target volume for the restore. To be 110% safe, why not execute with "dryrun true"? Presumably the log file, even if you cancel the dry run job after an hour or two, will provide confirmation and clarity about the target for the restore.
Hope that helps,
Mark
Consider that if you have a lost or damaged volume, when you create a new/replacement volume it will likely have a different ID, so it makes sense that you would want to specify the target volume for the restore. To be 110% safe, why not execute with "dryrun true"? Presumably the log file, even if you cancel the dry run job after an hour or two, will provide confirmation and clarity about the target for the restore.
Hope that helps,
Mark
___________________________________
L. Mark Stone
Mission Critical Email - Zimbra VAR/BSP/Training Partner https://www.missioncriticalemail.com/
AWS Certified Solutions Architect-Associate
L. Mark Stone
Mission Critical Email - Zimbra VAR/BSP/Training Partner https://www.missioncriticalemail.com/
AWS Certified Solutions Architect-Associate
Re: Restoring with doRestoreBlobs
It is a very rare method of restore, in these years we had to use it only two times by now.
From what we've seen it's basically an extended blobchk. It scans all the db entries (or a subset related to the specified volume) and for every missing blob it tries to restore from the backup.
Following this logic the volume ID should be the same, we too had to restore them on HSM but we gave the new volume the same ID of the previous.
As suggested by Mark, a dryrun will give you more details.
From what we've seen it's basically an extended blobchk. It scans all the db entries (or a subset related to the specified volume) and for every missing blob it tries to restore from the backup.
Following this logic the volume ID should be the same, we too had to restore them on HSM but we gave the new volume the same ID of the previous.
As suggested by Mark, a dryrun will give you more details.