News:

We really need your input in this questionnaire

Main Menu

DCI Errors

Started by Nikk, November 21, 2013, 04:10:02 PM

Previous topic - Next topic

Nikk

Hi,

Which paremeter should I raise, if time to time I get dci errors? (see attachment)
DCIPollingInterval = 60
NumberOfDataCollectors = 45
Nodes ~ 80
DCI's ~ 300-350

Should I provide more information?

Thanks in advance,
Nikk

Victor Kirhenshtein

What kind of DCIs are those problematic DCIs? If they are defined as ExternalParameter in agent, the problem may be due to command execution timeout. By default, both server and agent has quite small timeout for data collection (2 seconds) to prevent situation when one long-running external parameter delays collection of other DCIs. You can increase it if needed however. There are short description of relevant configuration parameters: https://www.netxms.org/forum/general-support/external-parameter-help/msg10080/#msg10080.

Best regards,
Victor

Nikk

They are just snmp dci's, and it is not always the same dci's!

Nikk

Victor Kirhenshtein

It could be timeouts or packet drops. I also encounter devices that was not able to respond to multiple consecutive SNMP requests. Can you try to capture SNMP traffic between server and device to see what actually transmitted when errors pops up?

Best regards,
Victor

Nikk

The thing is, that device (ProCurve Switch) is in remote location! I looked into wireshark, and seems like after requests from server, i'm getting back a bunch of empty packets. Maybe switch cannot respond back to multiple requests?

Thanks in advance,
Nikk

Victor Kirhenshtein

What SNMP version you are using? I have a bad experience with ProCurve and SNMP v3 - we had to fall back to v2c in one installation because of intermittent problems with polling and data collection using v3.

Best regards,
Victor

Nikk

Not my case, i'm using v2c :(