|
thanks
aris......I have already made a work around regarding my problem....heres my new
script..I put this one on the crontab to run every min.
#!/bin/sh
host="192.168.0.201";
cnt=1;
filename="`ping
-c $cnt -A $host | grep -o "Unreachable"`";
inactive1="Unreachable";
echo
"$date";
echo "$host";
echo "$filename";
echo "$inactive1";
if [
$filename = $inactive1 ]; then
/root/down2ndserver
mv
/root/down2ndserver
/root/down2ndserver.bakup
else
echo secondary server $host is up
fi
exit
The content of the script "2ndserver" is to restart the heartbeat. Once
the cron will run again for the 2nd time and so on..., heartbeat will not
restart cause 2ndserver script was move already.........by the way thanks for
your time and help....
Ed
can you give us a detailed info regarding your
heartbeat setup
like heartbeat version, ha.conf, how do you conduct the
test,
etc.
i'm confused. afaik, heartbeat "alone" doesn't
perform
load-balancing. my understanding of your problem is that
your
workstation is still being pointed to the backup proxy
eventhough it's dead
(kindly elaborate). this can be caused
by an unreleased IP resource on the
backup proxy or by an
incorrect dns entry (try using the IP instead of a
FQDN).
HTH
Eduardo Mapalo wrote:
Hi Aris,
Thanks for ur quick response, I don't have any DNS installed in my cluster.
My only option is to create a script to monitor server B...if server B is
down the script must run to restart the heartbeat on server A. Here's my
script for that..and it is running however may problem pa....I only need to
run the script once on server A....How can I do that?...cron is not
advisable....include the script on the mon services is to bad. I've observed
that both mon and cron will run the script every min.......which cause my
browser to hang-up while browsing.....any suggestion......??
#!/bin/sh
host="192.168.0.201";
cnt=1;
filename="`ping -c $cnt -A $host | grep -o "Unreachable"`";
inactive1="Unreachable";
echo "$date";
echo "$host";
echo "$filename";
echo "$inactive1";
if [ $filename = $inactive1 ]; then
/etc/init.d/heartbeat restart
else
echo secondary server $host is up
fi
exit
-----Original Message-----
From: Ariz C. Jacinto [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, May 12, 2004 8:07 AM
To: [EMAIL PROTECTED]; Philippine Linux Users Group Mailing List
Subject: Re: [plug] help with heartbeat
Eduardo Mapalo wrote:
Hello to all,
I have setup one fail-over clustered machine which are server A and server
B...running on squid services. My problem is I don't know how to disable
the
rr (round robin) which heartbeat do. When the primary is down, no problem
secondary will take-over all the services and theres no such round robin
scenario. But once the secondary server is down...thats the problem...round
robin still exist. Once heartbeat pointed to server B I cannot browse
anything on the web..cause there's a definite time for heartbeat to stay
poiting at server B which is dead already....I need to restart pa the
heartbeat on server A to disable round robin....pls help.
Heartbeat is not responsible for the round-robin, it's the DNS.
secondly, heartbeat is used not to monitor the "backup" but to
monitor the "Primary"
create a script that will monitor the nodes in a cluster. if one
of them are dead then remove it on the round robin's list.
HTH
|
--
Philippine Linux Users' Group (PLUG) Mailing List
[EMAIL PROTECTED] (#PLUG @ irc.free.net.ph)
Official Website: http://plug.linux.org.ph
Searchable Archives: http://marc.free.net.ph
.
To leave, go to http://lists.q-linux.com/mailman/listinfo/plug
.
Are you a Linux newbie? To join the newbie list, go to
http://lists.q-linux.com/mailman/listinfo/ph-linux-newbie
