Restoring From Backup w Warm Standby Question

[catherinelam]

My environment contains two EC2s: one primary and one warm standby. Due to a series of unfortunate events, our database on the warm standby got corrupted and phantom would not start on it. Luckily, we had a volume backup in AWS of the SOAR directory, so it was saved. 

However, after some research afterwards, we found a different method of backing up: https://docs.splunk.com/Documentation/SOARonprem/6.2.2/Admin/BackupOrRestoreAndWarmStandby

I think I'm being dense and overthinking it, but the article mentions a "primary warm standby", a "primary" + a "secondary" + a "warm standby" later on in the article. How many servers are in this configuration? I am not understanding how it is being set up and what the secondary is referencing. Also, what is a "primary warm standby"?

Would this article be helpful in the situation I described above with my failed warm standby? 

[phanTom · Accepted solution]

@catherinelam 

A warm standby is only ever 2 servers, 1 Parent & 1 Child. The Parent syncronises to the Child via postgres sync and rsync for shared files. The failover is still manual but can be scripted if you have the right probe setup on the LB to check and alert when the primary becomes unavailable. 

Personally I think using AWS functionality to restore will give you a quicker time to recovery.

[catherinelam]

Could you clarify what the documentation meant when it said "secondary" and "warm standby primary", if a warm standby only has two servers? I am curious. 

Just wanted to consider my options for backups and present them. 🙂 

[phanTom]

@catherinelam "warm standby" is the architecture and Primary / Secondary is the server role. One is only active at any one time.