Clearing a loop-protect snmp trap

[mlein]

Hi folks,

I've got a bunch of Procurve 2520s and I am setting up monitoring via nagios. I've enabled the loop-protect functionality on one of the switches and found a nagios plugin to check the switch via snmp and have successfully tested the loop-protect functionality and nagios monitors it fine. The problem I have is that the MIB in question doesn't clear back to a non-loop value once the loop is removed or even well after (it's been a couple hours now). So i'm wondering how to clear the value without rebooting the switch. I'd think that it would automatic but I guess not.

The commands I used in the CLI of the switch were:

loop-protect all

loop-protect trap loop-detected

loop-protect disable-timer 60

write memory

and the MIBs are 1.3.6.1.4.1.11.2.14.11.5.1.12.1.5.2.1.1.2.1 to 1.3.6.1.4.1.11.2.14.11.5.1.12.1.5.2.1.1.2.28 The MIB has a value of 2 when there's no loop detected and 1 when there is.

Thanks,

Mark

Edited to add an extra command for the switch (trap loop-detected)

[mlein]

Figured it out. Just have to go back into the CLI on the switch and run the following commands:

no loop-protect all

loop-protect all

That resets the loop-protect counters. It would be nice to find a way for the switch to do it itself though.

[paulgear]

Hi Mark,

I haven't looked at loop-protect for a while, but if it's anything like BPDU protection, the preferred way to reset after fixing the loop is to disable and then re-enable the interface.

It seems strange that it doesn't reset properly after your timer is expired, though.  Have you checked that you're on the latest firmware?  I had one 2520 in a network of mostly 2600 & 2800 switches, and it seemed to have the most flaky SNMP implementation of all the switches, even though it was one of the newer models.

[Richard Brodie_1 · Accepted solution]

The MIB is fairly explicit that's the way it's meant to work. I can see the logic, in that if you have a circuit breaker that trips it stays tripped until manually reset.

"This object will be set to TRUE when a loop is detected on
the port. The value of this object will be reset to FALSE when
the port's hpicfBridgeLoopProtectPortEnable is set to
FALSE."

snmpset 1.3.6.1.2.1.1.7.11.12.12.1.5.2.1.1.1.n to False (2) then True (1) will do it programatically.

[mlein]

Thanks for the reply Paul,

I did try disabling and re-enabling the port via the web console and that didn't clear the loop counter. I didn't try disabling and re-enabling via the CLI though.

The switches are running the latest and greatest firmware (14.30) too. I see there's an "Early Availability" release (15.09) available but i'm guessing that's the equivilent to beta which doesn't do me a whole lot of good in a production environment. Guess I can put it on one spare 2520 I have to test it though.

[mlein]

Ahh, well that solves that then. Thanks Richard.