9971s stuck on gateway after SRST event

[Scott Jones]

Recently I've had a string of WAN 'blips' that caused the phones to fall back to the gateway.  Most of them re-register back to CallManager after the outage is fixed, but it seems as though a few get stuck on the gateway...i've watched the timer count down and it just restarts the countdown. The SCCP phones, though, go right back with no issue...

Has anyone come across this before?  I've never dealt with 9971s until coming to the current environment.

Phone FW:  sip9971-9-3-1-10

IOS ver.:  15.1(4)M4 (3900 series platform)          

[Rob Huffman · Accepted solution]

Hi Scott,

The firmware "first found in" does not match exactly, but it

sure sounds like this bug;

CSCuc85032 - Fallback to CUCM may fail intermittently

Description

Symptom:
Phone registered on SRST may fail to fallback to CUCM.
Conditions:
Phone is used for days.
Workaround:
Reset phone from phone UI.

Details

First Found in:                          (1)

9.3(2)

Status:

Terminated

Last Modified:

Jan 27,2013

Product:

Cisco Unified IP Phones 9900 Series

Platform:

Dependent

Severity:

3 - moderate

Cheers!

Rob

"Clocks go slow in a place of work
Minutes drag and the hours jerk" 

-The Clash

[Scott Jones]

Spot on...i find it a little disconcerting that they won't fix the issue though.  every time it has happened, it's been to someone fairly high up in the organization...murphy's law i suppose.

thanks again, as always.

SJ

[Rob Huffman]

Hey Scott,

You are always welcome my friend!

Of course it's the "BIG wigs".....that's the way these things

always go and why the bug shows terminated is beyond me as well!

On another note, I read your post about the sound quality on the 9971's

and came across this bug that I thought you might be interested in. It

details a "special" firmware designed for issues where g.729 is being used;

CSCud13359 - 9971 Acoustic Fine-Tuning for certain customers - Handsets ONLY

Cheers!

Rob

"Clocks go slow in a place of work
Minutes drag and the hours jerk" 

-The Clash

[Scott Jones]

Interesting!  That might help a little bit...a lot of our issues are speakerphone related though, unfortunately.  Seems to be a couple of design flaws on the speaker itself and the mic placement or type or something.  We're going to try 7975s and see if that makes our people any happier.  Although, the Jabra SPEAK 410 UC works very nicely on the 9971s...only issue is if they use a headset too.

[markbatts]

HI Scott,

We have exactly the same issue , and can reproduce this quite regularly.Did you bother talking to Cisco about this?

cheers

MArk

[Scott Jones]

Hey Mark--yes, actually, i've yelled a few times about this...In my case, i'm in a particularly difficult place with it because we had to have custom firmware developed to address sound issues.  The problem is the bug identified by Rob won't be addressed in a suitable fashion for my company until next summer in the 9-4 releases when they roll the sound fixes into the general release.  However, I was told this has been addressed in the latest/greatest firmware that comes with 9.1 CallManager (I think it's 9-3-4...).

Give that load a shot and see.  We deployed it in APAC as they were a net-new cluster and haven't had any past experience with the sound issues, and so far, haven't seen a single instance of the phones getting stuck in SRST.

good luck!

SJ

[markbatts]

Hi Scott , thanks for the prompt reply,

the bug i'm interested in is the SRST bug CSCuc85032  , when you refer to the bug pointed out by rob do you mean this one or the sound one.

I'm toying with going to 9-3-4 as thats the latest one ,  can i just confirm that you have had wan failures with this code and have not seen any issues with SRST getting stuck?

thanks again

mark

[Scott Jones]

Ah--sorry for not being clear on that.  Yes, 85032 is the bug I meant.  We have had a couple of planned failovers, where we essentially cut the site off to simulate a failure, and there have also been one or two actual failures since we stood the cluster up earlier this year as well, all with none of the issues referenced in the bug.

[markbatts]

Hi Scott , thats great info , just one more thing , am i right in thinking that Cisco are actually going to officially fix this bug 85032 in 9.4?

[Scott Jones]

If I understood the info I got from the BU correctly, it is fixed in the latest firmware now (not sure if the bug ID was ever updated...).  9-4 is supposed to have the audio adjustments in it, to bring it in-line with the special load that I have to run to keep my users quiet (the special load is based on 9-2, so no SRST fix at this time...).

HTH