UM3 on Debian 3.1 with kernel 2.6.8

Hello,

I have just installed UM3 on two debian 3.1 with kernel 2.6.8 to provide
SMTP Filtring with load balancing (postfix/amavis/spamassassin/clamav).
These two servers are the real servers too.

I haven't patch my kernel. It's the debian version.

It seems that all work as fine, but some times (1/3 tries) the service
bound on the virtual IP address take a lot of time to answer. By example,
if you open a telnet session on the virtual IP on the SMTP port, the
welcome message can take 4 ou 5 seconds. If I call directly one of the
nodes, the problem doesn't occur.

For additionnal information, I have followed this installation procedure :
http://www.debianhelp.co.uk/ultramonkey.htm

Below my configuration files :
/etc/ha.d/ha.cf

logfile /var/log/ha-log # File to write other messages to
logfacility     local0 # Facility to use for syslog()/logger
keepalive 2 # keepalive: how long between heartbeats?
deadtime 10 # deadtime: how long-to-declare-host-dead?
warntime 10 # warntime: how long before issuing "late heartbeat" warning?
initdead 10
auto_failback on
# nice_failback on # nice_failback:  determines whether a resource will
fail #back to its "primary" node, or remain on whate$
udpport 694 # What UDP port to use for communication?
bcast   eth0    # Linux What interfaces to heartbeat over?
mcast eth0 225.0.0.1 694 1 0 # Set up a multicast heartbeat medium
#       Tell what machines are in the cluster
#       node    nodename ...    -- must match uname -n
node    mynode01
node    mynode02

respawn hacluster /usr/lib/heartbeat/ipfail
apiauth ipfail gid=haclient uid=hacluster


/etc/ha.d/haresources

mynode1 \
ldirectord::ldirectord.cf \
LVSSyncDaemonSwap::master \
IPaddr::192.168.250.90/24/eth0/192.168.250.255

/etc/ha.d/ldirectord.cf

# Global Directives
checktimeout=10
checkinterval=10
#fallback=127.0.0.1:25
autoreload=yes
logfile="/var/log/ldirectord.log"
#logfile="local0"
quiescent=yes

## Virtual Service for SMTP
virtual=192.168.250.90:25                 ## IP VIRTUEL
        #fallback=127.0.0.1:25
        real=192.168.250.91:25 gate ##mynode01
        real=192.168.250.92:25 gate ##mynode02
        service=smtp
        scheduler=wlc   ##mode fonctionnement du cluster
        #persistent=600
        protocol=tcp
        checktype=negotiate

Thanks for help,
Regards,
Kinaï


-- 
Ultra Monkey - http://www.ultramonkey.org/
To UNSUBSCRIBE, email to [email protected], with a body:
unsubscribe ultramonkey-users [email protected]
where "[email protected]" is YOUR email address.