CMM Locking up totally untill hard reset (x-10)

I have a CMM Micro in a location that is “BH Centeral” of my system. This CMM is locking up. The times are always at 5ish, 7ish and 10ish in the evenings. On weekends it deviates and happens at different times durning the day. To me, these times seem to be in the peek hours of use. There are NO errors in the log file of the CMM or the BHs. I have IP monitor that monitors the network with pings every 5 mins. at these times, all traffic stops.

The only thing I can think of that is new is a few weeks ago, we installed a bridge wave 60Ghz as our primary link which greatly increaced our bandwidth. so it’s our building (cmm micro/master) -> bridgewave->next building with cmm1(slaves) ->Canopy 5.7 BH20 -> building with cmm micro I am talking about. the bridgewave is hooked up to our main switch with its own cable and so is the cmm. the next building the bridgewave is hooked up to the same switch as the cmm is… the in the building that is rebooting, the cmm is the only “switch”.

I’m not sure if that has anything to do with the reboots, but it’s the ONLY change in my structure. and to think about it, it didn’t start right after that, but soon after that link update.

So, is there any reason we would think the CMM would be rebooting? I know the ground is great in that building as well. perhaps its the BH20 going to the 3rd building… but there are not errors on that either other than a few currupt data counts at 1 old format beacon count on the RF STATS ---- and 1 rx overrun with a few out erros on the ethernet stats (this shows on both sides actually).

Any ideas?

Oh, one last thing, the cmm that requires a reboot to reestablish communications is definatly grounded properly. Everything is in a small building with a grounding bar all the way around that leads to stakes outside (about 12ft from the building).

You are sure that it is rebooting? Have you checked the system up time? How many mac addresses are in the bridge? Do you use vlans?

sorry, what I meant to explain was it locks up “until” I reboot it via x-10 trhough the phone line. Once I do that, all links come back up.

Hw do I check the MAC table of the CMM? At this time, in the BH20 going to this CMM there are under 400 entries and no errors. However, I’ll have to check before a reboot if possible. Come to think of it, I don’t see any lost pings on the BH on that tower… and the port MIB shows some errors…

CLIP

RxFCSErrors 4
RxFragments 2
RxSymbolError 1

so maybe it’s the rebooting of the BH that brings it all back up… crap.

Vlans… I only use the uplink feature… I had two ports selected, but on ly one used (BH to the next tower of coarse) and one for when I go out there with a laptop. I disabled in on the laptop interface just for the heck of it.

Well I have check the records on our monitor and the cmm stops responding, and the BH hooked to it doesn’t stop. so the BH link to this CMM is staying up however, no pings are going past it.

just an update…

Anyone?

Consider the number of mac addresses in the bridge to be twice the number of SMs you have plus other network elements like AP/BH/CMM.

Today we flooded one CMM with large number of mac addresses and it stopped responding. When we stopped the flood it came back.

If I was you I would change the cable of the BH to SM, use only 100 mbps full duplex for the BHs and try to see if the CMM is still up when it freezes again. If it’s still up I would use WireShark sniffer to see what happens.

Also consider using management vlans, one for AP/BH/CMM, other for SMs and others for users. This looks like a big job, but it will relive you from stress in the future.

Hey!

Have you tried hard-coding the link speed in the CMM and the new backhaul? Just a thought. Along that same line, is there any chance that the new high-powered radio is causing some interference on the wire itself, i.e, is the new equipment mounted above that which was pre-existing?

they are both set to auto, and they are negotiating at 100full. I will try hard coding.

The new unit on the building before the one with the problem is actually 20ft bellow the other units (canopy BHs)

Other than that, I will be doing some whireshark I recon. I will definately take the advise on the vlans… sooner than later.