j5 high availability can be achieved by making use of a cold standby system on redundant hardware. Database replication and file system replication is used to ensure that the standby system has the latest application and data and is ready to start processing if the production server becomes unavailable. Failover then involves starting the j5 services on the standby server and updating the DNS or IP address to switch the production traffic to that system.
Use of the staging server for standby
Most j5 customers operate a staging server - a separate physical or virtual machine with a j5 system used to test changes prior to going into production. Ideally, the staging server should be separate from the standby server, but if necessary, it is possible to host the standby j5 system on the same server as the staging j5 system, with these considerations:
-
The standby and staging j5 systems must be entirely separate, i.e. separate ports, separate services, separate j5 application folders, separate j5 system folders, and separate databases.
-
The standby server must have the capacity to handle production volumes simultaneously with testing volumes, unless the intention is to suspend testing on the staging server while it is used for production failover.
-
This approach has additional risk, since running different j5 systems adds complexity, and a shared staging server may not be as clean or as available as a dedicated standby server.