Active IQ Unified Manager Discussions
Active IQ Unified Manager Discussions
Hi,
Query from my customer; support haven't been able to assist.
Is NetApp DFM capable of alerting on “interface down” or “SAS channel down”. If it is – then we have an issue:
Specifically - Unplugging a network cable or SAS cable does not get recognised in DFM. We have waited a number of hours AND change the monitoring time to 1 minute.
I have seen events (I think triggered on DFM start-up):
An Error event at 30 Jan 16:57 GMT Standard Time on interface e0c on Active/Active Controller controller.local: Interface Status Down.
DFM – 5.02.7873
DOT - 8.0.2 7-Mode
Anything else we can try?
Thanks,
Tony
Hi Tony,
Hope you are well.
Can you check if the filer(s) are set to send SNMP traps to the DFM server? Check the output of the 'snmp' command on the filer(s). Traphost should include the hostname or ip of the DFM server and 'init' should be set to '1'. If not...
snmp traphost add <hostname | IP address>
snmp init 1
...should do the trick.
Regards,
Craig
Thanks Craig - I've asked the customer to try this and report back. I'll post an update when I get one.
Cheers annd thanks,
Tony
Hi Craig,
That made no difference, I'm afraid. Does anyone else have any other ideas?
Thanks all,
Tony
DFM/UM cannot alert on a link break, although I am not sure that is causing your lack of alerts.
Check the evevnttype list for any events that might be useful in logging an event, which are mandatory for generating an alert/triggering an alarm. I do not recall any SAS specific events but do not have CLI access at the moment.
Hi Tony,
Can you check if there is TRAP defined for the same ? If so then Craigs suggestion of enabled DFM host as trap receiver should get you this event.
The only down side is that all event raised for traps are of the severity Information and they dont show up in the events report.
Pls use events-history report to view the same or change the trap severity to make it show up in the normal report.
The default event severity definition in OCUM for Traps is Information. But the same can be modified.
Default Severity Definition for all trap events is Information :
+++++++++++++++++++++++++++++++++++++++++++++++++
C:\>dfm eventtype list | findstr /i trap-received
alert-trap-received Information alert-trap-received
critical-trap-received Information critical-trap-received
emergency-trap-received Information emergency-trap-received
error-trap-received Information error-trap-received
information-trap-received Information information-trap-received
notification-trap-received Information notification-trap-received
warning-trap-received Information warning-trap-received
C:\>
Modify them as follows:
++++++++++++++++++++
C:\>dfm eventtype modify -v Warning alert-trap-received
Modified event "alert-trap-received".
C:\>dfm eventtype modify -v Critical critical-trap-received
Modified event "critical-trap-received".
C:\>dfm eventtype modify -v Emergency emergency-trap-received
Modified event "emergency-trap-received".
C:\>dfm eventtype modify -v Error error-trap-received
Modified event "error-trap-received".
C:\>dfm eventtype modify -v Warning warning-trap-received
Modified event "warning-trap-received".
C:\>
Now you will receive the trap with appropriate severity:
++++++++++++++++++++++++++++++++++++++++++++++++
C:\>dfm eventtype list | findstr /i trap-received
alert-trap-received Warning alert-trap-received
critical-trap-received Critical critical-trap-received
emergency-trap-received Emergency emergency-trap-received
error-trap-received Error error-trap-received
information-trap-received Information information-trap-received
notification-trap-received Information notification-trap-received
warning-trap-received Warning warning-trap-received
The drawback or side effect:
======================
When a trap other than information severity is generated the object status of the filer changes from green to orange, yellow or red.
And even when the condition is rectified the object status doesn’t return back to green as there is no neutralization event.
Now to overcome and make the object status green resolve the event by clicking on the resolve now.
Regards
adai
This has bitten me before and I never found a solution in DFM either. I eventually wrote a Perl script run that scans the filer message log every 15 minutes and sends me an email if it sees (among other things):
netif.linkUp
netif.linkDown
So if a link goes down, I get an email that says something like this:
Fri Mar 29 12:09:45 EDT [<filer name>: netif.linkDown:info]: Ethernet c0a: Link down, check cable.
I still guess there is no option to monitor a Nic Port going down. Can someone please let us know if there is any alarm to monitor the same.