Cluster failure

[saranraj456]

How to find the cluster failure proactively ,except the options like auto support?

[ekashpureff]

Proactively ? What's your definition of proactive ? It implies you do things to prevent a cluster failover.

Make sure you maintain your NetApp. Keep it cool, provide clean power, the list could go on.

Enabling autosupport to notify you of health events, watching out for BURTS regarding your equipment,

using SNMP traps and monitoring are all good proactive practices.

But an autosupport message to notify you of a cluster failover is the primary way to know that it happened.

I hope this response has been helpful to you.

At your service,

Eugene E. Kashpureff
ekashp@kashpureff.org
Senior Systems Architect / NetApp Certified Instructor
http://www.linkedin.com/in/eugenekashpureff

(P.S. I appreciate points for helpful or correct answers.)

[saranraj456]

Sorry for the word pro-active.

I just want to know the alternate thing for autosupport alert?

And also ,Could you explain me about the SNMP trap?

Thanks,

Saran

[muhammad_i_pasha · Accepted solution]

Alternative to autosupport could be DFM alerts. For this you need to set the dfm server as your snmp trap host on filer and set an alert on dfm server for traps.

But best and quickest way is autosupport message.

[saranraj456]

We are having a centralized DFM server to the entire setup.

The DFM alerts are lagging when compared to autosupport messages.

Is there any way to improvise this?

[shaunjurr]

Hi,

You can setup your DFM server (or HPOV or whatever else you might have) to accept snmp traps from your filers.  Then setup DFM to fire off a mail or some pager mail.  The lag you are seeing is often because it tries a few times to reach the filer before it declares it dead (or some DFM bugs...).  You may also have lags in your mail system that won't be helped even if you send your traps to the DFM server.  All of this is documented in the DFM documentation.

Hope this helps a little.