I read an article recently on avoiding fallback in distributed systems.[0]
Is it more appropriate to call the strategy in this case fallback, or failover? Since the secondary server wasn't running in production until the first one failed, it sounds like fallback?
Perhaps higher reliability strategies would have been instead of having a secondary server, just have more mirrored disks on the main server, to reduce the likelihood of the array being compromised?
Alternatively, to run both the primary and secondary servers in production all the time. But that would presumably merely move the single point of failure to the proxy?
Is it more appropriate to call the strategy in this case fallback, or failover? Since the secondary server wasn't running in production until the first one failed, it sounds like fallback?
Perhaps higher reliability strategies would have been instead of having a secondary server, just have more mirrored disks on the main server, to reduce the likelihood of the array being compromised?
Alternatively, to run both the primary and secondary servers in production all the time. But that would presumably merely move the single point of failure to the proxy?
[0] https://aws.amazon.com/builders-library/avoiding-fallback-in...