The hobbit-alerts.cfg entry for the secondary:
HOST=<PRIMARYSERVER> COLOR=red SERVICE=bbd,conn SCRIPT /usr/local/scripts/hbtSecondaryTakeOver <EMAILADDRESS> MAIL <EMAILADDRESS>
The hbtSecondaryTakeOver script:
echo "Starting at date, $BBHOSTNAME . $BBSVCNAME" >/tmp/blah
MESSAGE="Hobbit: $BBHOSTNAME's $BBSVCNAME service is $BBCOLORLEVEL. Secondary taking over"
export HOST=/usr/bin/hostname
export USER=hobbit
echo "mailing: $RCPT \"$MESSAGE\"" >> /tmp/blah
echo $MESSAGE | /usr/bin/mail $RCPT
HETC="/home/hobbit/server/etc"
cp ${HETC}/hobbit-alerts.cfg.primary ${HETC}/hobbit-alerts.cfg The hobbit-alerts.cfg entry to release (would go in hobbit-alerts.cfg.primary)
HOST=iadhbtp1 COLOR=red SERVICE=bbd,conn RECOVERED SCRIPT /usr/local/scripts/hbtSecondaryRelease <EMAILADDRESS> MAIL <EMAILADDRESS>
And the release script:
echo "Starting at date, $BBHOSTNAME . $BBSVCNAME" >/tmp/blah
MESSAGE="Hobbit: $BBHOSTNAME's $BBSVCNAME service is $BBCOLORLEVEL. Secondary turning down"
export USER=hobbit
echo "mailing: $RCPT \"$MESSAGE\"" >> /tmp/blah
echo $MESSAGE | /usr/bin/mail $RCPT
HETC="/home/hobbit/server/etc"
cp ${HETC}/hobbit-alerts.cfg.secondary ${HETC}/hobbit-alerts.cfg
You'll need to adjust things like, machine names, file/directory locations/name, and email addresses.
Also, You don't want the Release script to run on the primary, so either don't have it there or have some logic in the script that stops execution if it's not the right machine (like if [ "$HOSTNAME" != "<secondaryserver>"] ; then exit fi)
Paul Root Lead Internet Systems Eng Qwest Network Services
From: Raymond, David [mailto:David.Raymond at ca.ebarrette.com] Sent: Wednesday, February 09, 2011 2:46 PM To: xymon at xymon.com Subject: RE: [xymon] RE: xymon replication or clustering
Are you able to send me the how to for this. My skill is a little limited and that could help me.
thx
David Raymond From: Root, Paul [mailto:Paul.Root at qwest.com] Sent: Wednesday, February 09, 2011 14:49 To: 'xymon at xymon.com' Subject: [xymon] RE: xymon replication or clustering
I have a secondary server watching the primary. It has 1 think to do in hobbit-alerts.cfg, watch the primary. If the secondary looses contact with the primary it will run a script.
That script is to copy in the primaries hobbit-alerts.cfg file in place. It is put on the secondary automatically with rdist.
Also, in the primaries hobbit-alerts.cfg file, there is a watch that looks for a recovery, at which time, another script is run to reverse the process.
Pretty simple
Paul Root Lead Internet Systems Eng Qwest Network Services
From: Raymond, David [mailto:David.Raymond at ca.ebarrette.com] Sent: Wednesday, February 09, 2011 1:04 PM To: xymon at xymon.com Subject: [xymon] xymon replication or clustering Hi,
We have one xymon in our Head Office, and other one in branch office. In my Head office, we want to have 2 xymon to replicate result, if one failed, the other take place.
An employe have start something, but he left month ago. I see I the hobbit he did, this line in bb-hosts : summary Hobbit-PRD.All 172.30.1.1 http://hobbit/hobbit group servers 127.0.0.1 Hobbit-PRD # bbd http://Hobbit-PRD/
So the big question is, it is possible to do this ? Can we have to xymon than are in "cluster"
And if we want to get our branch office?
thx
David Raymond
This communication is the property of Qwest and may contain confidential or privileged information. Unauthorized use of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender by reply e-mail and destroy all copies of the communication and any attachments.
This communication is the property of Qwest and may contain confidential or privileged information. Unauthorized use of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender by reply e-mail and destroy all copies of the communication and any attachments.