LMS 3.2.1 sending more pings than thought usual

[dlongworth]

Hi,

I'm looking for some assistance or tips with troubleshooting LMS 3.2.1 and the ping/ICMP traffic it transmits.

A lot of my devices are receiving a lot more pings from LMS than I would have anticipated.

I don't run PING sweeps in Device Discovery or CM-UT. I've even disabled DFM polling in a hope to trace the source of these PINGs.

Does anyone have a list of which modules use PINGs so I can turn them off and track down the offender.

I really only want to manage the known devices I already have via SNMP alone. I don't require LMS to be PINGing for discovery or reachability purposes.

Any help is appreciated.

Kind Regards

David

[Gaganjeet Chug]

Hi,

You have all already check all the possibilities that will be generating ICMP traffic. Next step should be to stop the dameon manager and observe if the ICMP traffic is still there more than usual. And do you have any third party application installed on LMS server ?

Thanks,

Gaganjeet

[dlongworth]

Thx Gaganjeet,

It was DFM.

It just didn't stop pinging until after a server reboot, despite applying the new settings.

I've got a core switch with multiple SVIs but I only want DFM to monitor the Loopbacks. What's the most effective method to do this, would you know? Or is DFM designed to monitor ALL IP addresses on a device.

Regards

David

[Gaganjeet Chug]

Hi,

Thanks for the udpate. DFM is actually designed to monitor ALL IP address on a device for effective monitoring.

So was it DFM that is pinging the device even when the daemon manager was stopped. There was a bug in LMS 3.2 where sm_server didnt get killed even if the daemon manager was stopped but that got fixed in LMS 3.2.1 that is your current version.

i know about a script where you can unmanage few ip address from DFM but that will include a complete subnet and also i tried that back in LMS 2.6.

Many Thanks,

Gaganjeet

[dlongworth]

Hi Gaganjeet,

Not quite correct.

DFM was still pinging after I disabled ALL polling parameters and saved/applied. I had not restarted the daemon. I didn't believe I had to.

After that I tried to delete all the devices from DFM (Auto-allocation was also disabled) but they would not delete.

(I have another LMS server where DFM will not learn devices from DCR and comes up with the 'TreeManager' error but all fixes in Forum failed. Could this be the same bug?)

After this I restarted daemon manager (LMS) and it stopped pinging.

Is this normal operation for DFM?

I'm interested in the script if you can send it to me. Basically my management loopbacks are in one or two subnets only and we have no desire for the user vlans to be pinged. Besides loopbacks will never go down unless the switch does.

Basically the distribution-switch cpu reaches 100% utilisation as all its SVIs are pinged simultaneously. Changing the time intervals extends the polling but this has the negating effect of delaying alerts and warnings, etc.

I'm trying to find the happy balance but DFM is not very flexible. It's all or nothing.

Many thanks for your help. Please let me know how I can get that script.

David