topic Re: Predictive Drive Failure in Insight Remote Support

Predictive Drive Failure

shocko — Tue, 09 Aug 2011 15:00:04 GMT

Guys,

I'm using WBEM on some servers. On one of them, I have noticed it sent SIM an indication that one drive was in a state of predictive failure. Should this produce and auto support call via IRS/WEBES? i.e. should predictive drive failures trigger a call at the HP backend?

Re: Predictive Drive Failure

Leo Plewa — Fri, 02 Sep 2011 07:08:11 GMT

hello,

i saw the same behavior on our systems.

Predictive failed drives did not produce a case.

Could this behavior be changed ?

What is the reason that no case will be produced ?

Thanks for any hints

Re: Predictive Drive Failure

Brad Cunningham — Fri, 02 Sep 2011 21:02:35 GMT

Was this Windows or HPUX for the system that reported the failure?

You need to ensure that there is a WEBES subscription in place also. This will be pointed back to the SIM server on port 7906.

When you look at WEBES is it a valid entity “not orange in color”?

Does it have a valid WBEM protocol listed, site, contact and entitlement filled out properly?

If all this is OK we should have seen something in the WEBES list of events. There's an outside chance that this was a non-actionable events but if the drive failed I would think this was not the case. You might try sending some test events through to ensure WEBES and the RSC client for RSA is working correctly.

Regards brad

Re: Predictive Drive Failure

Leo Plewa — Mon, 05 Sep 2011 07:38:15 GMT

For me i saw this problem on windows 2003 servers with PSP 8.60 and linux servers with RH5 and PSP 8.70.

All servers are using SNMP.

SNMP test traps are always received.

In case of a predictive failed disk i did not receive an SNMP Trap and so no case is created.

For full defective HDDs traps and cases are generated.

Re: Predictive Drive Failure

shocko — Mon, 05 Sep 2011 09:35:16 GMT

Hi guys,

I've figured this one out. The answer is YES, predictive drive failures should generate support calls. The problem I had was that webes was not subscribed to SIM and so wasn't analyzing WBEM events received from systems when a predictive failure was detected. Created a subscription to SIM WMI mapped from webes and now auto supports are generated when webes analyses the WBEM events sent to SIM.