Unmarking a failed CPG

[Gakecs]

So after a CPG fails, specific state description "Copy Space Grow Failed," how might I remove this tag? I have acknowledged all alerts, fixed the problem (thanks to some great responses here), and my CPG has been compacted, grown, shrunk, etc etc after all this. All the VVs point to the same copy space, which also now has plenty of room to grow. How can I get this to stop showing a failed tag? I've got a support request in... but it's been hours... and I'm a really impatient guy. Google is failing me.

[Gakecs]

For anyone else that runs into this problem...

If your capacity utilization is above 85% the CPG will not come out of failed status without issuing the "setsysmgr restart" command through the CLI.

[chanchon]

For anyone else that runs into this problem...

If your capacity utilization is above 85% the CPG will not come out of failed status without issuing the "setsysmgr restart" command through the CLI.

i guess it is a risky action!
i ran into this problem and i don't know what i should do for this problem!!
i am scared to run "setsysmgr restart" because in the Reference manual it has been said that could be result in data loss!!!
any comment ?

[Architect]

Personally i think that a sysmgr restart should never be needed to fix such a simple warning/error.
usually the state is automatically resolved when all other errors are set to fixed.

So if you do a showalert, set all displayed errors to fixed (after really fixing them ofcourse), and finally run a checkhealth -svc to double check if the array is really error-free, the message should disappear from the IMC.

[RitonLaBevue]

Did you try "compactld -f -pat *" ?

[chanchon]

Personally i think that a sysmgr restart should never be needed to fix such a simple warning/error.
usually the state is automatically resolved when all other errors are set to fixed.

So if you do a showalert, set all displayed errors to fixed (after really fixing them ofcourse), and finally run a checkhealth -svc to double check if the array is really error-free, the message should disappear from the IMC.

there is no alert but by running checkhealth -svc i encountered some error not about CPGs!

[chanchon]

Did you try "compactld -f -pat *" ?

I ran compact LD on management console and it didn't work , trying "compactld -f -pat *" doesn't ave any risk ?

[Architect]

what errors did checkhealth gave you?
Compacting LDs doesn't really make sense to me in solving this, but he might know something i don't.

[RitonLaBevue]

It seems that compactcpg and compact don't provide same results on failed CPGs.

Architect, you know everything bout 3PAR then ?

[Architect]

Nobody knows everything on 3PAR...

Tunesys should normally fix LDs that do not match the CPG configuration, but when it does not, a manual tuneld might be needed for the LDs that are created with degraded specifications.

I'm not aware a compactld will fix an LD that is created with degraded specification, but that is why i wrote that you might know something that i don't.

if there are still LDs with a degraded speficiation, the checkhealth should mark them, that is why i asked for the errors checkhealth gave.