|
On Mon, 2004-05-17 at 14:06, Patrick LeBoutillier wrote:
My question is: if heartbeat starts and brings up the alias on eth0 and
then
the network eth0 is connected to fails, will the director still be able to
deal with the packets even if they start coming in from eth1?
Here's something I'm going to try:
- Switched from LVS-DR to LVS-TUN to allow virtual IP to be in a
different network than
realserver IP
- Setup virtual IP interface on the tunl0 interface (192.168.82.1) on
each machine
- Setup real IP interface on the lo interface (192.168.80.1 and
192.168.81.1 respectively)
on each machine
- Created "noop" heartbeat plugin so that heartbeat leaves the virtual
IP alias alone
(on tunl0) on fallback:
haresources: cluster.domain.tl noop::192.168.82.1 ldirectord ...
Heartbeat therefore no longer manages the virtual IP alias
Then:
- Create 2 configurations for gated:
1) announce 192.168.8[01].0/24 (config "real_only")
2) announce 192.168.8[01].0/24 and192.168.82.0/24 (config "real_dir")
- Setup heartbeat so that it starts gated with "real_dir" config when
the ressources
are acquired and starts gated with "real_only" config when it looses
the resources.
That way gated always announces the realserver networks (so that the
director can reach
the realservers) via both interfaces, but only the active director
announces the virtual
IP network (again via both interfaces).
Seems like it should work, I will try it out tomorrow.
Any comments?
Patrick
--
=====================
Patrick LeBoutillier
Laval, Quebec, Canada
| 