CMS Call bridge and DB cluster resilence.

[Guilherme Mascarenhas]

Hi,

Please, anybody know where I can find the information below?

When clustering CMS on DB and Call Bridge.

1. What happen when a CMS server gets isolated from the other call bridge servers in the cluster?

2. What happen when a CMS server gets isolated from the other database servers in the cluster?

3. Anybody know what is the database running on CMS?

Your help is much appreciated.

Thank you,

[Patrick Sparkman · Accepted solution]

If a CMS becomes disconnected from the either the Call Bridge or database cluster, it will no longer be active in the cluster until it's able to reconnect.  I don't know the specifics about the database itself, but all that is needed to run a separate external database is simply install another CMS and only configure the database, in this scenario, the other CMS components such as the Call Bridge etc doesn't even need to be licensed or activated unless you plan on using them.

[Guilherme Mascarenhas]

Hi Patrick,

Thank you, please do you know if there is any documentation about these scenarios?

Please one more scenario. Considering all servers below in a call bridge cluster.

The information below is the latency between the sites.

	Site1	Site2	Site3	Site4	Site5
DB Cluster Master	3	1	2
Site1	x	101	148	200	188
Site2	101	x	54	212	198
Site3	148	54	x	168	236
Site4	200	212	168	x	76
Site5	188	198	236	76	x

What issues can happen between sites 2 and 4 and 3 and 5 where the latency is over 200ms?

Thanks a lot.

[Patrick Sparkman]

Take a look at the CMS Scalable and Resilient Deployment Guides on how clustering works.

Regarding latency, the recommendation is 200ms between database servers, latency between Call Bridges and master database should be kept to a minimum as well.  There is only one master database active at any time, all other databases will become slaves/hot-standbys, there is a limit of 5 database nodes in a cluster.