Location Bandwidth Leak at one location only

[shane.orr]

I am running Call Manager version 8.5.1.11900-21

I have one remote site out of around 20 where location bandwidth leaks. Call Manager does not seem to decrement the bandwidth on some calls when they disconnect. Slowly they creep up to the bandwidth limitation and it has to be forced to resync. I have created a new location in call manager and applied it to this site’s device pool and deleted the original location but that did not help.

TAC seems to be spinning wheels, any ideas?

Thanks in advance,

Shane

[bartdejager]

hi Shane,

we seem to be having the same issue, did you get any more info from TAC, bug ID created maybe ?

we are running: 8.5.1.13900-5

thanks

[shane.orr]

In our case we were hitting this bug and at the time required an ES to resolve.

CSCtj41084 Bug Details

SIP reINVITE glare handling, SipCdpc leak in reqOfferReceived_pending

Symptom: When reINVITE glare happens, 491 will be received, and starts a timer, retry this reINVITE after timer fires. In the meanwhile, remote side will most likely do the same thing. If remote side timer is shorter, and we (SME) will receive reINVITE again before our timer fires. At some condition, this will trigger clear out our cache for our side reINVITE, and we will lose the transparency info in our outgoing reINVITE. And also end to end transaction will be broken in such cases.Conditions: As title, when reINVITE glare happens, that means while we are sending out our reINVITE, we receive reINVITE from remote side.Workaround:No

Status

Fixed

(Resolved)

Severity

3 - moderate

Last Modified

In Last 2 weeks

Product

Cisco Unified Communications Manager (CallManager)

Technology

1st Found-In

8.5(0)

8.5(1)

Fixed-In

8.5(1.98000.4)

8.5(1.98000.7)

8.6(1.10000.1)

8.6(1.10001.1)

8.6(1.10000.43)

8.5(1.13030.1)

Component(s)

cp-sip-trunk

Regression

Y