ignite recovery from rx7640 to rx2660

[Carme Torca]

Hi,

I have problems whith one rx2660 server. I restored one ignite from rx7640 in this server (rx2660). And now I have problems with oracle database, I don't know what is happening but the server hangs and it doesn't do a crash, I have to generate it.

I don't know if its something because of the ignite restore.

Could anyome help me?

What I could do?

Thanks a lot of,
Carmen.

[Vivek_Pendse]

Did you get any while restoring ignite backup on rx2660?
If not, then you need monitor the memory/cpu/io resources either from glance or sar or top to check what exactly happens?

Thanks,
Vivek

[Torsten. · Accepted solution]

Not sure if this really works, because the rx2660 is much newer and different hardware, so you need to check if all drivers (e.g. CPU support, SAS support, NIC support etc.) is up to date.

[Vivek_Pendse]

If you have SAS drivers installed before taking ignite backup then it won't be an issue for restoration. But, here you have already restored the backup. Just updating the patch bundle (HWE, QPK) is enough.

[Carme Torca]

One question. The ignite is from a vpar in the rx7640. And the rx2660 has not vpars, maybe the problem is it? Have I to reinstall all in rx2660 or maybe could I solve it?

Thanks a lot of,
Carmen.

[Vivek_Pendse]

That is not an issue.
You can migrate from vpar to physical m/c.

[klb]

I've had similar issues caused by

1) Hard NFS mounts occurring at boot against hosts that don't exist or against network that doesn't exist or is down

1.5) bad entry high in /etc/inittab with :wait: and it never returns control back to init(1m).

2) /etc/resolv.conf and /etc/nsswitch.conf configs that don't give up on a DNS lookup when the DNS servers aren't available/down/offline.

3) runaway memory hog causing thrashing so severe nothing gets done

4) runaway fork() caller that fills up the entire process table not allowing any new processes. usually this is a non-root session and can be worked around to get back to normal. But, if root is doing it...bad.

5) generally mucked up file permissions after a bad chown -R or chgrp -R call.

6) mass file delete as root with -R from / that was stopped after a little while :)

... and I'm sure there are others.

Hth,

-klb

[Carme Torca]

Finally, I have updated the server with all last recommended and drivers and the servers works ok now.

Thanks a lot of!!!