Martin Bene
2003-05-16 06:40:59 UTC
Hi Dan,
having the advantage of being MUCH easier to install.
Also, I'd suggest doing things slightly differently from your specification:
- give each node a node IP address. The node retains this address at all
times, it is NEVER moved/migrated by the cluster manager. This IP is just for
management purposes and to make configuration easier - it's the default setup
for both failsafe and heartbeat. No public services use this address.
- give each node an additional high availability IP address. these addresses
are manageged by the cluster manager (i.e moved to the remaining node in case
of a failure). All services the node provides use these addresses.
Bye, Martin
however, i need, as explained, both machines active at the same time...
active-active... not active-standby
What you want can be done both with failsave and with heartbeat, the lateractive-active... not active-standby
having the advantage of being MUCH easier to install.
Also, I'd suggest doing things slightly differently from your specification:
- give each node a node IP address. The node retains this address at all
times, it is NEVER moved/migrated by the cluster manager. This IP is just for
management purposes and to make configuration easier - it's the default setup
for both failsafe and heartbeat. No public services use this address.
- give each node an additional high availability IP address. these addresses
are manageged by the cluster manager (i.e moved to the remaining node in case
of a failure). All services the node provides use these addresses.
Bye, Martin