Issues with CMS cluster media link between callbridges

[Nick Halbert-Lillyman]

I've got an environment with 2 x CMS Callbridges and a single database server.

Servers are clustered (both callbridge and database) and are in the same Callbridge group.

Calls from CUCM work fine and one of the Callbridges is picked to host a space, but the media link between each CMS fails, giving a log message like this:

remote media link to call bridge "CMS2" failed for call 1c5f5044-1532-4b8e-bfee-4281d7c3c001, coSpace 07e8b499-9d77-43e2-b863-f7243924de44

For CUCM calls, this isn't an issue as the call is just directed to the correct CMS using the Callbridge group, but for WebRTC calls it means a user can't join a space if it's hosted on a different CMS.

Any ideas where to look for the media link failure?  Both Callbridges have heartbeat to each other and the database is in Sync.

CMS version is 2.2.9.

[Wayne DeNardi]

Have you checked your deployment against the Deployment Guide, especially the port requirements (Table 14, Page 171) for the ports that are required between each of the Call Bridges and the Call Bridges and Database server.  

[Nick Halbert-Lillyman]

Hi Wayne,

In this instance all three servers are on the same subnet, so there's no ports blocked between them.

I've been reading through the deployment guide but I can't find anything specific to the media distribution link, from what I gather it should just "work" (assuming the ports are open) if the callbridge clustering is up, unless I am misunderstanding.

[Wayne DeNardi]

It should pretty much "just work" as you you say.  Have you tried removing the troublesome CallBridge from the cluster and then re-adding it?

[Nick Halbert-Lillyman]

Just tried but no luck :(

[Wayne DeNardi]

Looks like you're not the only one who's had this issue - see remote media link to call bridge failed

[Patrick Sparkman]

Might be worth opening a case with TAC, not much info on this issue from what I've seen, and if it's been a problem for several users, than it might be a bug or something else.

[Jaime Valencia]

Have you configured the outbound rules between each CMS so they can dial each other?

[Nick Halbert-Lillyman]

Hi Jamie,

No I haven't configured anything for them to call each other - I've been using this document  www.cisco.com/go/pdihttps://www.cisco.com/c/dam/en/us/td/docs/conferencing/ciscoMeetingServer/Deployment_Guide/Version-2-2/Cisco-Meeting-Server-2-2-Scalable-and-Resilient-Deployments.pdf

I've done dial rules for SIP endpoints dialling in, however I didn't see anything about Callbridges dialling each other?  Any idea what I need to configure for this or are you able to point me in the right direction?>  I'm sure I just missed something in there...

[Wayne DeNardi]

It's mentioned very vaguely in the document - the need to link the call bridges is described in section 5.3 of the guide.  Near the bottom of this section it mentions that "you need to configure dial plans for inter-peer calls between the clustered Call Bridges".  It does keep suggesting you see the "Deploying the Call Bridges" section which refers you back to the section you're already looking at which isn't helpful.

If you look at the more recent 2.3 guide - section 6.9.1, it provides more detail on how to do this.

[Nick Halbert-Lillyman]

Thanks Wayne, I did see that section but it doesn't actually describe the inbound/outbound dial rules that would be needed... does that mean I just have to match an inbound call of anything terminating on the local Callbridge and forward it to the other Callbridge? And if so, do I need to do anything to prevent a loop?

[Wayne DeNardi]

If you look at the more recent 2.3 guide, section 6.9.1 (page 56) provides more details on this.

[Nick Halbert-Lillyman]

Thanks Wayne that does indeed give a much better explanation, I was using the 2.2 document as this was the version I was deploying it on.

I'll give that configuration a go and report back, thanks all.

[Nick Halbert-Lillyman]

Just to close the loop on this, Wayne I finally got a chance to implement your suggestion and it did indeed fix the issue, thanks!

Short answer: Configure an Outbound dial rule with both the domain and proxy as the callbridge IP address, make a rule for each Callbridge in the cluster.

[lleste.ext1]

Hi Nick,

I have the same issue as you. I configure the outbound dial rule with both the domain and proxy as the callbridge IP address but i still have my issue. Do you configure anything else like incoming dial rule?

Thanks in advance