|
On 2000-08-13T09:58:23,
Julian Anastasov <ja@xxxxxx> said:
> - For VS/NAT the real servers have to switch to the new router,
> with or without their knowledge.
Yes. heartbeat, FailSafe (almost any clustering solution) will take over the
IP addresses too.
> To support or not to support:
> - the state of the masq modules/applications must be replicated too
If we finally got around to use conntrack from Netfilter, we could implement
this in a common layer of code which replicates the conntrack status.
In any case those too implementations will share a lot of code, so once one is
done, it shouldn't be too difficult.
> - smoothly move all services from one director to the backup one,
> i.e. the support is not restricted for failover cases only. For
> example, to install a new kernel or to make other changes that
> lead to downtimes.
This is easy and automatically supported - just trigger the failover manually.
> - move only some of the services to another box but continue to
> serve the rest of the services.
uhoh. This may be more difficult. I don't know if we have to support this in
the first iteration.
Sincerely,
Lars Marowsky-Brée <lmb@xxxxxxx>
Development HA
--
Perfection is our goal, excellence will be tolerated. -- J. Yahl
|
