Disaster Recovery configurations require full rebuild of the failed environment.
How can testing be performed without that requirement?
How can valid configuration be determined outside of a real failure to observe functional recovery from the secondary systems environment?
Confidence in fail over recovery is lost without testing results known prior to a possible real failure.
Most organizations are unable to test this in production environments.
CA Performance Management r3.6 and newer releases
- Disaster Recovery assumes the failed machines are no longer functional and must be replaced. This is why the configuration and recovery requires a full reinstall as it's part of the tested process.
- The only way to validate the function is to test it in a non-disaster scenario where the original machines can be rebuilt to become the new DR standby environment.
- If the systems are on VMs snapshots are always a great option for use in a recovery scenario.
- If they are physical machines, we have other customers who set up a dual environment. Not always a great option for larger organizations and networks, but when environment 1 meets its end, environment two is sitting ready to go.
- Good backups of the Vertica DR DB and of the MySql DBs on the CAPC server are also critical to successful recovery options.
Per Product Management there is nothing on the current product Roadmaps related to changes for Disaster Recovery functionality beyond the current feature set. Their current focus is on High Availability solutions and functionality.
Anything new for Disaster Recovery at this time would have to be considered an Enhancement Request which should be submitted as an Idea posting in the CA Performance Management Community site.