CMS webbridge capacity limit

[wearetherobots]

We have a CMS cluster with 2 webbridges and are wondering what is the expected behavior that a user will see when the webbridges are at full, 100 each capacity? We are only using Meeting App/WebRTC for guests to join spaces (no signed in users), so will users still be able to enter meeting ID/passcode and attempt to connect, or get a 404 page, or what have other's seen? With heavy remote use taking place at this time, we would like to be prepared on what is normal and what might need a TAC case and there does not seem to be any documentation I can find on this topic. Thanks

[Haydn von Imhof]

We have asked Cisco this same question with no response yet,

Personally I have seen more than 140 web participants connected to a single CMS1K with no issues.

The limit doesn't seem to be a hard codec one but more a resource and processing limit on the CMS.

As each web participant might be coming in at different resolutions and call speeds you might find the CMS allows more than 100 to connect. I imagine at some point though you will start noticing performance issues but we haven't managed to get to a point where we have that many people connected at once.

None of our users have reported seeing page time our errors or anything like that when we past the 100 limit.

[wearetherobots]

After a few weeks of increasing usage we saw the webbridge max out at 149 connections. Once that is reached users received an "unable to join meeting" error. We also found elsewhere that you can access a page that shows the percentage of utilization (of that 149 number) if you go to the following in a web browser - 

https://your-webbridge-fqdn.domain/api/v1/load 

[maqsood ahmed]

i had a question,, CMS Datasheet announces 350 HD Callls on CMS 2000 ,, what is the WebRTC limits of 149 and 350 Calls from webrtc related ,,,

[Haydn von Imhof]

Further to this if you are using an EXPe and c for external web proxy be away of the below bug, even though its stated the E limit is 100 web sessions you start seeing issues around 70 after which any new users start seeing and error page an cannot log in. TAC can temporarily increase the limits in the back end but if a reboot takes place the original limit will be back in place.

https://bst.cloudapps.cisco.com/bugsearch/bug/CSCvt23674

[manipandey]

is there anyone out there who really get it working seamlessly for 400+ participants over Webrtc using multiple CMS1000 with Expressway E 's doing the TURN.

We can't get it going for 100+ users on webrtc.

Hope to hear from you all.

Thanks and regards 

M

[robertb]

As Haydn said, the bug on the maximum web clients connected via Expressway is set to 70 and can be changed - you'll need to open a TAC case to get it increased to 100.

We have experienced a 404 error on our CMS web page when we have hit our limit if users are joining from external and then had to spin up another Expressway pair. 

As for the Webbridge capacity, I don't think there is a ceiling limit and rather a notice to say that Cisco cant guarantee performance past 100.

Do some tests and capture logs from both Expressway-E and CMS (logbundle) and upload to CSA (link below) to see what kind of WB issues you are facing when you believe you are running at peak periods.

https://cway.cisco.com/csa/ 

If you also have LB enabled in you Callbridge groups, make sure you have loadbalancingUserCalls parameter set to true:

<loadBalanceUserCalls>true</loadBalanceUserCalls>

You may also spin up multiple WB's within virtualization guidelines below:

https://www.cisco.com/c/dam/en/us/td/docs/voice_ip_comm/uc_system/virtualization/virtualization-cisco-meeting-server.html

As the webbridge is merely a passthough to the CB, i don't believe there is much compute required for each WB you decide to deploy.

[robertb]

Update: EXP-E 12.5.9 now support 256 Web clients, apposed to the 70 (100 clients if TAC assists in lifting this via Apache)

hope this helps.

[lior look]

do you know more details about Medium deployment? for x12.6 and above