08-14-2018 01:25 AM - edited 03-19-2019 01:30 PM
Hi Cisco community,
I have a question regarding the deployment of HA for Persistent Chat rooms in IM&P. We recently upgraded our cluster from 10.5.x to 11.5.x. We had two separate Oracle databases. Each of them connected to one IM&P server. All persistent chat rooms were created on 1 IM&P server. Meaning everytime the cluster goes down (e.g. maintenance) Users get a notification that their persistent chat rooms have been deleted. I read through the HA deployment guide of 11.5.1. It says it provides redundancy as well as HA. Here comes the question:
The section "Merge External Databases" reads
If you have upgraded from a release prior to 11.5(1), and you had persistent chat configured with each persistent chat node connecting to a separate external database instance, use this procedure to merge the two databases in a subcluster into a single database that connects to both nodes.
That means I will be left with just one database which has a connection to both IM&P servers. That provides me with HA but the redundancy would be lost. How can I still provide redundancy if the database node fails?
Thank you in advance. Any help is appreciated