ONTAP Discussions
ONTAP Discussions
Hi,
IHAC who has several SnapVault secondary volumes. SnapVault is managed by Protection Manager. The destination aggregate is nearly 100% full and he wants to create a new aggregate and move 3 specific volumes into it.
How could this be accomplished with Protection Manager ?
Do I have to snapmirror the volumes into the new aggragte first, do a snapvault update -S and Protection Manager will recognize the changes ?
Any help is appreciated.
Thanks
Thorsten
Solved! See The Solution
Hi Thorsten,
- create the new secondary volume in the new aggregate
[adai]yes
You must probably suspend the dataset so there is no conformance run or scheduled update.
- manually create a VSM relationship between secvol_old and secvol_new
[adai] yes
- quiesce and break the VSM relationship
[adai] yes. Also cleanup the vsm snapshot and snapmirror.conf entry.
- create a new snapvault relationship between prim_vol and secvol_new
[adai] yes but for each qtrees and do a snapvault start -r
- import it into protection manager
[adai] Yes this will work.But we don’t import the old backup version.
- delete the snapvault relationship between prim_vol and secvol_old
[adai] Yes. use snapvault stop.
- be happy...
Regards
adai
Hello,
I am on leave till Monday, 7th September 2009. Please contact support@tek-tools.com for any assistance in the meantime, and I will attempt to respond to your email as soon as possible upon my return.
Regards,
Aravind
Currently you cant specifically move only selected volumes in PM.
PM on automatically reaching the threshold of aggr full will rebaseline the whole relationship from primary to next availble rp.
I am not sure if your work around will work.
Regards
adai
It's possible using the lrep model designed by NetApp. Using lrep you can transfer the baseline transfer to other aggregates or filers.
Adai
Why PM can't recognise that. LREP is supported model from NetApp where baseline transfer are sent through physical means of transfer. Imagine customers having peta bytes of data and you expect them to re-baseline the SV/SM relatioships. It'll take years for them accompolish via WAN.
It turns out volumes have unique IDs associated with them. When you create a new volume and fill it with VSM, it retains it's unique ID. DFM uses the unique id to sort things out when a volume gets renamed, so it's pretty important to us. It also means we can't just assume that if two volumes with the same name show up, they must contain the same data.
Hi Thorsten --
Not, Protection Manager will not automatically figure out that you've copied the secondary volume. There are two problems.
First, the new secondary volume is not a member of your dataset. You need to add it. Also, the new SnapVault relationships are not "managed" by ProtMgr, so the right thing to do would be to import those relationships from the "External Relationships" page.
Second, we track the dataset backups based on a unique ID in the secondary volume. The new volume you create will have a different unique ID so we won't realize the snapshots there are the same as the snapshots in your original secondary volume.
There is good news. We are working on code to automate moving secondary volumes. If all goes well, it will be in the next release.
-- Pete
Hi Pete,
thanks for you answer. But what is the detailed action plan to migrate one or more secondary volumes in a new aggregate ?
From what I understood it could be something like that:
- create the new secondary volume in the new aggreagte
- manually create a VSM relationship between secvol_old and secvol_new
- quiesce and break the VSM relationship
- create a new snapvault relationship between prim_vol and secvol_new
- import it into protection manager
- delete the snapvault relationship between prim_vol and secvol_old
- be happy... 🙂
Is this the correct way or did I miss something ?
I have to do it on Friday so it would really help to have a functional migration plan without a new baseline transfer.
Thanks
Thorsten
Once you import all your old backup versions are not imported.Only the relationships are imported.
Regards
adai
Hi Thorsten,
- create the new secondary volume in the new aggregate
[adai]yes
You must probably suspend the dataset so there is no conformance run or scheduled update.
- manually create a VSM relationship between secvol_old and secvol_new
[adai] yes
- quiesce and break the VSM relationship
[adai] yes. Also cleanup the vsm snapshot and snapmirror.conf entry.
- create a new snapvault relationship between prim_vol and secvol_new
[adai] yes but for each qtrees and do a snapvault start -r
- import it into protection manager
[adai] Yes this will work.But we don’t import the old backup version.
- delete the snapvault relationship between prim_vol and secvol_old
[adai] Yes. use snapvault stop.
- be happy...
Regards
adai
Hi Thorsten,
With DFM 4.0 already out with this capability, did you try the secondary space management feature ?
Regards
adai
Just curious if anyone has tried this as I have customer who will likely be needing it soon....
Hi Andew,
Why dont you try and share us your experience ?
These are some of the advantages of using Secondary Space Management.
Regards
adai