|
Howdy all-
Just try to get ldirectord running on my director to detect real server
failure and I am a little confused about how to setup it...
I am running the ldirectord included with ipvs 0.8.0. I was trying to
follow the instructions on
http://www.linuxvirtualserver.org/HighAvailability.html to set it up,
but I think that I have either misread the directions or I am missing
something obvious with my setup.
I copied the ldirectord.cf file to /etc/ha.d/ and edited for my LVS
setup as follows:
# Global Directives
checktimeout=5
checkinterval=30
virtual=192.168.2.1:80
real=192.168.2.2:80 gate 5
real=192.168.2.3:80 gate 5
service=http
request="index.html"
receive="Test Page"
scheduler=rr
protocol=tcp
Ldirectord correctly configures the ipvs tables on my director using
the information above, but after that it does not remove the servers
when they have failed (i.e. index.html does not exist on either of
them). Ldirectord does show up as a running process and correctly
creates and deletes the ipvs table when started and stopped.
What am I missing? It is an LVS-DR, all linux 2.4.5 boxes, running
apache and tomcat. Would keepalived be a better choice (I like the fact
that I can do layer 3,4 and 5 checks with keepalived, but from the docs
you need to be running ipchains)?
Thanks again,
Zack
|
