Link Up/Link Down

[Mike Drobick]

I saw a posting about this last year, but it didn't look like there was an answer. I have a few ProCurve 2626s and 2650s that even after setting: snmp-server host a.b.c.d "snmp_string" Not-INFO, I still get tons of Link Up and Link Down informational alerts in HP SIM. I've set up an Event collection and Delete Event task to keep these under control, but I'd like to stop them from happening entirely if possible. I did flash the firmware to H_08_72 on all switches, but it hasn't fixed anything.

Does anyone have any suggestions? Is the command supposed to do what I want it to and is just broken, or is that used for something else?

Thanks.

Mike

[Kell van Daal]

712 is Cold Start
713 is Link Down
714 is Link Up
The value 4=logandtrap while 1=none

I just realized you might want the CLI command to do that.

setmib 1.3.6.1.2.1.16.9.1.1.3.713 -i 1
setmib 1.3.6.1.2.1.16.9.1.1.3.714 -i 1

It does not prompt with "are you sure?"
followed by "are you sure you are sure?"

Well, if you change your mind, you can set it back to the defaults with:
setmib 1.3.6.1.2.1.16.9.1.1.3.713 -i 4
setmib 1.3.6.1.2.1.16.9.1.1.3.714 -i 4

;)


This response is shamelessy copied from two answers Les Ligetfalvy gave in a thread.

[Les Ligetfalvy · Accepted solution]

Thanks for the credit and for jumping in to help.

[Mike Drobick]

Kell, I'll give you some points for referring me to that solution...that definitely knocked out the SIM problem. Les...since you are the originator, I'll give you points as well.

So, now the SIM isn't reporting them anymore, can anyone give me the answer to the real problem...why does the switch still send them out? I've gone all the way to Critical and they were still being sent. It seems like the snmp-server command should filter out these messages before they are sent to the SIM server where I am now ignoring the trap.

Anyone?

Thanks for your help guys!!!

Mike

[Les Ligetfalvy]

If you are talking about the actual port toggles, they can be caused by a number of things. Some switch models have seen a reduction in the number of events with newer firmware but the source of the toggles may be external to the switch.

One possible cause is bad power or cabling while another is power management. Then it is entirely possible that the clients are rebooting, causing the toggles.

While killing the messenger does get rid of the annoyances of the log getting cluttered, it can mask the underlying cause. I was able to reduce the number of toggles by swapping out bad jumpers, cleaning up dirty power, and disabling power management on the NICs to the point that I can tolerate them in my syslog. I do block them from my traplog though.

[Mike Drobick]

I do believe I have seen a reduction in flapping since upgrading from 60 to 72 on the firmware, but I guess I have two concerns.

1) Kinda as you've answered...why are they flapping all the time. I guess I could troubleshoot, although I no longer have any traps to refer to. But I'm talking like 330000 Link Up/Link Down messages since Sept. 17th. I do think they were reduced with the firmware, but maybe to about 100 per hour.

2) The bigger question...if I have the command snmp-server host a.b.c.d "snmp_string" Critical, why does it still send out the trap. I guess I would have to assume the switch considers Link Up or Link Down a Critical alert, but then why does SIM list it as Informational? I know on the BayStack 450-24T that is at the core, I could turn off the Link Up/Link Down trap and now my SIM server doesn't see it anymore.

I guess in all, while I would like the flapping to stop, I would REALLY like to have the switch stop telling me that they are flapping until I can get around to troubleshooting. It just seems like that function isn't working correctly. Or, I just don't understand it which is entirely possible.

Thanks again.

Mike