UCSM response slow, hung sometimes, restart is the only solution?

[shiva_yerram]

Hi,

We have multiple(>80) clients connecting to UCSM and retrieve responses (xml api), sometimes UCSM wont respond and hung.

We are restarting FICs to fix this problem. I'm not sure this is the right way, do you recommend to restart or fix something better way?

Thanks

[padramas]

Shiva,

We need more information about slowness

1)  N/W topology between UCSM client and FI

2)  Has this worked before ?

3)  Are you going through any HTTP proxy ?

4)  Does it happen all the time and for every one ?

5)  UCSM version and FI model

Padma

[shiva_yerram]

Padma,

1. Clients (VMs) and FI are three hops distance with 4 ms latency.

2. Yes, its still been working, but issue only happens if UCSM is loaded with requests and connections.

3. Nope

4. We saw this behavior to heavy loaded UCSs

5. 1.4/2.0 and 6120/6140

Thanks,

Shiva

[Robert Burns]

Shiva,

Is there a common version of Java being used?  If so can you advise which versions?

Thanks,

Robert

[shiva_yerram]

Robert,

we are using 1.6 for client side api access.

Thanks,

Shiva

[Steven Williams]

Having the same problem with our UCS servers. TAC has told me this is a problem with the 1.4(1) and 1.4(2) BIOS. Supposedly there is a fix coming late next week. So basically the CIMC and SSH to the UCS will not respond and power cycling is the only way to fix this. If you are running VMWare on these UCS hosts, your vCenter will also lockup when trying to manage them. There is a fix for that from VMware:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1030265

So while I may be able to manage my VM's from vCenter, I am still unable to get to my CIMC, but atleast I can get to my VM's.

HTH