Solved!
Go to SolutionNavisphere odd reporting when migrating
HI All,
Just been watching a migrating lun for the past day. While it is migrating across without error I notice the gui is reporting it is in a Queued state. Even thought it says queued I can see the hours remaining drop, and the percent done increase. Should I be concerned, is the gui provide a false status, or is there something wrong with the migration?
Any thoughts would be most appericated.
Cheers
JL
726
1
Posted June 29th, 2010 00:00
I would not be concerned.
It may be just the field did not refresh. try logging out of navisphere and logging back in.
If that does not work try restarting the management server on the storage processors
726
0
Posted June 29th, 2010 04:00
Navisphere (JAVA based) is known for not updating certain fields correctly. Even an "update now" doesn't help sometimes. Loging out en back in should indeed do the trick.
726
0
Posted June 29th, 2010 06:00
A question i have is how big is this Lun you are migrating and what type of Array do you have? It is either very big or you are migrating at a lower setting. I always do my migrations on High or ASAP. even during busy hours, it is not that much of a performance hit on the SP
726
0
Posted June 29th, 2010 08:00
Kenn,
Using the high or ASAP setting can impact a busy array, especially the ASAP setting. I wouldn't recommend using the ASAP setting unless the impact to the array is understood.
And yes you are right, on a lightly utilized array the impact of the ASAP setting is not great, but I have seen the ASAP setting drive up utilization on a busy array.
726
0
Posted June 29th, 2010 08:00
Everyone's workload is different on their array and so for some a migration during business hours is feasible. If you haven't browsed it, there is a document called 'The Influence of Priorities on EMC CLARiiON LUN Management Operations Applied Technology' (yes another long EMC title
). There is a section on the impact of LUN migrations and a nice graph showing how much an ASAP migration can impact an array and a list of the migration rates.
Low 0.9 MB/s
Med. 1.6 MB/s
High 3.4 MB/s
ASAP 350 MB/s (Mirror RAID)
ASAP 163 MB/s (Parity RAID)
You can imagine the resources it takes for the array to increase the migration rate a hundred fold (high vs. ASAP).
726
0
Posted June 29th, 2010 08:00
Sorry i should have siad i do the ASAP during business hours and not busy hours. But i always check SP utilization before i start a migration.
That was sorta why i was asking what type of array he had to sort of see how much CPU power he had to work with.
Thanks
726
0
Posted June 29th, 2010 14:00
Hi All,
thanks for you feed back. In answer to the size it is a 1TB lun being migrated into a 2TB. Both a sata drives. Currently these to luns only host low use development luns, this is only production LUN on there.
I was wondering if it was a Java issue, so had updated my java client. The field didn't refresh post a log off and log on. Will use the cli to track progress.
Cheers
JL
Jim_A1
59 Posts
726
0
Posted June 29th, 2010 06:00
If you ever have an issue with what the GUI is reporting running a naviseccli command should give you the correct state of what it is you are trying to view in the GUI.
In you case if you run naviseccli command -
naviseccli -h -user -password -scope 0 migrate -list
this will list all migrations currently running and their states.
Valid states are queued, migrating, migrated, transitioning or faulted.
Reference the "EMC Navisphere Command Line Interface (CLI) reference 6.29" (chapter 7) for details.
Jim