Possible SNMP issue

[sirhc]

maybe try to factory reset the switch that is having issues...

Sure we can give that a try. It's going to take us about a week before we can do this because we have some SLA customers being fed off these switches. Ill update you everyone once we have done that.

Factory default and set back up from scratch, is pointless if you export, default, and import the old config.

[Nitroxide]

That's not really an answer guys. I have this problem on all my switches. Can you please look into it further.

[davidbaker]

Okay so here is an update:

Defaulted the in-service Switch and reloaded it did not fix the issue. I re-configured a new switch from scratch with all the same settings and it's also doing that same thing in Solarwinds. I opened a ticket with Solarwinds and concluded that Solarwinds is just reading the data being sent to it from the Netonix switch.

We just deployed another one of these switches to a new location and it is also having the same problem..

Before when I tried a new Switch straight out of the box with a basic config and reported that it had no problem. It appears I jumped the gun on that as well. I didn't give Solarwinds enough time to collect all the poll data. I can confirm that now the new switch right out of the box with a very basic config is also not reporting SNMP correctly.

So is there anything else we can try?

[Dave]

thanks for doing the leg work on this....Eric contact you to figure out how we solve this...

[Eric Stern]

If thats the case then you should be able to see the incorrect data coming from the switch by using another query tool like SnmpB. If you can collect one or two reads that show the incorrect values that would be very helpful. Just copy/paste from SnmpB and attach them. Or email to eric@netonix.com.

[davidbaker]

Solarwinds contacted me back and asked if there was a new MIB file for the Netonix (Other people have called in about this issue). I gave them the one you guys have available for download and also gave them the MIB walk. So by next Tuesday (12/20) they will put out an updated MIB database with the new Netonix mibs in and see if that fixes the issue.

Meanwhile, the workaround is to go and edit the interface in Solarwinds and change the interface speeds manually.

Still doesn't explain why this all started after changing firmwares.....

Will wait until Tuseday to see if this MIB database update fixes the issue.

[Nitroxide]

any updates on this?

[Nitroxide]

Hello???

[davidbaker]

Just a quick update as I failed to check back to the forums! Anyways, Solarwinds was not able to determine any problems in their software after a number of Wireshark traces and diagnostic with their DEV team. These interface changes are false notifications that the Netonix is reporting to Solarwinds after Solarwinds polls the device via snmp. So with my annoyance level at an all time high, we just disabled the interface change event messages in Solarwinds just to make the problem stop displaying on the main screen. You have to go into your SQL database to disable this event....

So no the problem isn't fixed, but it's out of sight and out of mind.

[davidbaker]

I hate to revisit this yet again, but a recent update with a Solarwinds patch made all these interface changes show up yet again. I contacted Solarwinds to see if they have another workaround to just filter out the interface change issue and after doing some more digging, it is determined for sure that Solarwinds is reporting what Netonix is responding to snmp polls. Here are the logs from the snmpwalk:


.1.3.6.1.2.1.2.2.1.5.1 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.2 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.3 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.4 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.5 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.6 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.7 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.8 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.9 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.10 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.11 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.12 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.13 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.14 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.15 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.16 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.17 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.18 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.19 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.20 = GAUGE32: 100000000
.1.3.6.1.2.1.2.2.1.5.21 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.22 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.23 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.24 = GAUGE32: 1000000000
.1.3.6.1.2.1.2.2.1.5.25 = GAUGE32: 0
.1.3.6.1.2.1.2.2.1.5.26 = GAUGE32: 1000000000
So my question is, what OID should Netonix be reporting for its interface speed? These all appear to be reporting over a Tb when its all either 1Gb or 100Mb back to my server.