|
Hi Joe,
Just a quick reply (very buzy currently).
> I'm mostly interested in being able to force a failover. Alexandre's way
> of doing this at the moment is to down an interface or HUP with a changed
> priority weight. I'd like vrrpd to failover following notification from
> an external event (eg smartd saying that a disk is failing, or
> overtemperature...)
> and I'd like the NICs to stay up if they are OK.
Yes, VRRP like HSRP, state transition are drived by interface state. This is
the way
VRRP work.
OTOH, it makes sense to be able to drive VRRP state transition according to
Healthcheck.
This is my goal with THC, to provided threaded checkers for IPVS part but also
for the
VRRP. That way VRRP FSM will be able to handle checkers status to drive protocol
transition.
I must finish a small paper for spec this part. I will post the paper through
mailing list.
Best regards,
Alexandre
|
