3 ms·
I've found that for RDS, a writer instance and a hot standby reader instance with automatic failover work pretty well. When a failover happens, you're usually l
by jniedrauer 6y ago
I've found that for RDS, a writer instance and a hot standby reader instance with automatic failover work pretty well. When a failover happens, you're usually looking at about 30 seconds of downtime, which is "good enough" for most purposes.
- icedchai 6y ago30 seconds is pretty good. I worked on an "enterprise" system running AIX and HACMP (IBM's HA software.) A failover event would take minutes... and this was on the same local network.