Cisco MCU 4520 unexpected crash

mguguvcevski · ‎05-08-2013

Hello community,

I was wondering if anyone has come across the same issue on a MCU 4520:

trap: pid 62.34 (mcu): user write DSI trap @ 0xc7e5008 by 0x1e9ab6c (DSISR 0x42000000, err=12)

UVM: pid 62.34 (mcu), uid 0 killed: out of swap

[2] Killed ./mcu

EXIT 137

[1367763601] libunbound[793:0] error: tube msg write failed: Broken pipe

May 5 14:20:01 reboot: rebooted by root

I am suspecting

CSCuc09395 - MVC Zaphod crash with out of swap backtrace

but it is different.

Thanks

Patrick Pettit · ‎05-08-2013

Hi. Im assuming youve opened a support case on this? What version you running on MCU would help. Out of swap may mean a memory issue. Be nice to know what was happening on the Mcu at the time it occurred. You using TMS? Any other monitoring systems polling the MCU? Need some more details if possible.

If you decide to open a case, download diagnostic.zip file and load it to the case. Also any details you have on what was happening before the crash is also helpful to know.

Sent from Cisco Technical Support Android App

mguguvcevski · ‎05-09-2013

Hello Patrick,

Yes a TAC case has been logged SR 625837053 but for few days now it seems it is deadlocked somewhere between TAC and development I guess.

Nothing was going on on the MCU of the time of occurence, as it crashed and rebooted during off hours.

There is no TMS in this setup.

Hope that the diagnostics would mean more to TAC, as they seem to me like a memory dump of certain kind.

Thanks

Patrick Pettit · ‎05-10-2013

Hi mguguvcevski. I don't think its a deadlock between TAC and Dev. We just see some odd messages going on in your even log prior to.

Seeing a lot of these messages in your console and the zip file:

161953.637 CONFERENCE : Info : new SIP participant 19273

161953.652 SIP : Info : Incoming call

161953.656 SIP : Warning : Invalid Request URI type

161953.657 SIP : Warning : Invalid Request URI type

161953.659 CONFERENCE : Info : new SIP participant 19274

161953.674 SIP : Info : Incoming call

161953.678 SIP : Warning : Invalid Request URI type

161953.679 SIP : Warning : Invalid Request URI type

161953.680 CONFERENCE : Info : new SIP participant 19275

These are printing every second almost prior to when the restart occurred. If you start to see these messages again, would be nice to know who is originating them and how. Not saying this is the cause of the problem, but seeing this over and over in the span of less then a second is troublesome to me.

Plus there are other items in the diagnostic zip file that seems to suggest that all SIP connections where taken, but can't see how as no one was really connected at the time.

What I'd like to know is where these requests are originating from. If you see these messages in your event log on the MCU would be great, to run SIP/H323 Log to se where they are originating from. Maybe someone spammed your box trying to make calls or something which caused these messages.

This is me thinking out loud here. If you can find out who this was, would be great, and maybe we can assimilate somehow in the lab to try and see how the MCU got short on memory.

VR

Patrick