I have a couple of questions about SnapProtect - they may or may not be linked so I'll post them both together.
Scenario DoT 7-Mode 8.1.3 - SnapMirror and Vault fully licensed. 2 physical 2220 filers both HA pairs PRISAN-A,B and DRSAN-A,B.
OCUM 5.2 , SnapProtect 10.4SP4+.
This is a new install. A test vfiler object has been created 'cifs_vfiler' and its configured with vfilerDR between PRISAN-B and DRSAN-B - snapmirror is functioning normally every 15 minutes.
I have created policies, added NAS clients (including a specific one for the vfiler I'm trying to backup) and added arrays into SnapProtect, under the Storage Policy I have Primary (Snap) , followed by CIFSTESTVAULT (with a source of Primary (Snap).
All looks good so far. If I run a backup job, immediate full, and goto advanced - untick the 'Skip Catalog Phase...' then the backup job takes the correct action, makes a snapshot on the primary and indexes the data ready for a restore - I can then browse the restore area. However ...
Q1) If I run a snapshot job with the default option (skip catalog phase) then when I try to do a 'browse from snapshot' on restore then an error appears "Please check if browse path [/vol/xxx/.snapshot/SP_2_12_6xxx] is valid, snapshot [SP_2_12_6xxx] is present, and volume [/vol/xxx] is valid and online. If I do a list snaps from the controller or the SnapProtect interface I can see the snapshot ID SP_2_12_6xxx that is being referenced without any issues? so it appears as if NAS live-browse isn't working - unless I have got the wrong end of the stick!
The second issue is the vault copy, I have gone into NMC and created two secondary resource pools on the DRSAN nodes pointing to aggregates with plenty of free space.
In the SP I have assigned the secondary resource pool to the vault job without any errors.
Q2) Each time I attempt to run a vault copy I get an error and the job just attempts to restart but doesn't seem to 'do' anything to the secondary storage or begin any vaulting procedures.
Error message is:
"Error occurred while adding list of snapshots to Dataset id , Source Copy . Error Message = [DFM Server database is still not refreshed with new qtrees or new luns that have been added to volume [cifstestvol] on file server [cifstest_vfiler] OR with new volumes on file server [cifstest_vfiler]. Continuing without waiting for DFM refresh to complete may result in new members not getting added to the DFM dataset. Please wait for some more time and retry the operation"
I would be grateful for any suggestions on either of these, suffice to say the system is not under any load at all at the moment. Have I missed something obvious?
Nothing concrete as yet although I have now made some progress on my own, although the job failed on the first and second go when I ran it a third (and subsequent) time(s) it did work?
Also there was an erroneous DNS entry for one of the vFilers and although I have everything configured with IP addresses things did start to work better once this incorrect entry was resolved. I remember reading in one of the deployment guides that correct name resolution is key to how all the various entities work together in the workflow.
Any further updates and I'll post :-)
I understood what happened in my case.
Before I run my snapvault baseline I changed the volume name (I changed because the volume name had a typo). The DFM recognized that I had renamed the volume, but the SnapProtect did not recognize that my renamed volume because I already had some local snaps (Snap Copy) for that datastore/volume.
I followed this steps:
1) I removed the datastore from the subclient;
2) I ran the Snap Copy again;
3) I ran the Vault Copy (it is running right now, but of course not for that datastore);
5) I will delete all old backups (I can do this because it is a new implementation);
4) I will add my datastore again to the subclient;
6) I will run Snap Copy again;
7) I will try to run Vault Copy again.
If it does not work I have another strategy a little bit more complex.
Flavio Onofre de Souza