Crowd Data Center FAQ

For detailed comparison of Crowd Data Center and Crowd Server, see Crowd Server and Data Center feature comparison.

See Crowd documentation for detailed description of the migration process.

In order for hot failover to be possible, we recommend that you have a minimum of two cluster nodes. Adding more nodes will increase availability in case of node failures, and might increase throughput of some operations (notably authentication requests, if Crowd is handling authentication for your applications).

Because Crowd Data Center provides full active-active clustering, if a node goes down a failover can happen in seconds. You do not have to wait for a "cold standby" to start up, for all the data to replicate consistently, or for traffic redirection to propagate in the DNS. All this is handled automatically and seamlessly by Crowd Data Center and your load balancer, without intervention by the system administrator.

In Crowd Data Center, user sessions are stored in the database so all operations related to user sessions can be handled by any node in the cluster. In case a node fails, the load-balancer will route traffic to other nodes preventing users from noticing any functionality loss. The same applies for sessions in the Crowd web interface (if you were in the process of doing a multi-step operation, like configuring a new application, you might need to restart that process).

You can choose any load balancer you like provided it support the following:

• cookie-based session affinity ("sticky sessions")
• determining node availability based on a healthcheck URL
• SSL termination to offload HTTPS encryption and decryption

If you have no specific preference for your load balancer, you can find instructions for Apache and HAProxy on the Crowd Data Center load balancer configuration examples page.

No. Crowd Data Center does not allow cluster nodes with mixed versions in the same cluster. Even if you upgrade nodes one at a time, there must always be a period of downtime between the old and new versions.

However, when properly managed, upgrades of a Crowd Data Center instance should take no more downtime than with an equivalent Crowd Server instance, and can often be considerably less. In Crowd Data Center, nodes can be taken offline one by one to perform the upgrade, leaving only the shutdown time of the last node on the old version and the startup time of the first node on the new version inside the downtime window. In Crowd Server instances this is only possible if you have provisioned an identical backup server for the purpose.