|
I'm not in the IT dept, but I believe the strategy my company uses is something like this:
1. cluster two or four servers all sharing the same SAN disk array and with load-balancer
2. 2PC replication to a standby cluster
3. log-shipping to a remote standby server cluster every five min
4. daily full-backups to remote disk
5. daily full/diff backups to tape
Sounds like overkill, but we haven't had any major problems yet.
|