Scaling TIMP



Can you tell me about some possible scability/redundancy architectures with TIMP? We have about 1200 employees worldwide, and IM is a critical service for them. How would you recommend deploying TIMP?

There are 3 ways to achieve redundancy/scalability and it all depends on the way you want to configure your network.

1 - use a domain name for each section/department of your company; you would have IM addresses like:
xxx@NYC.mycompany
yyy@LA.mycompany

Install a server for each department and link them through s2s connection

Benefits: distributed architecture; even if connection drops between some departments, users can chat within the departments

2 - use one domain name; all 1200 users on one server.
For redundancy: have DB user integration; have a backup server with data replication and a script to switch from one server to the other

Depending on the way your users chat, it could be possible that one server would not be enough (can be only checked while using the system)

3 - use one domain and decouple front end and back end to scale. This solution allows you to scale nicely up to as many users as you want, but it requires a custom installation of TIMP.