Andreas Kurz wrote:
Hello Madison,
Your drbd configuration seems to work fine, but keep in mind that you
can only mount a drbd device that is in prmary state. So you have to
umount the filesystem on the master, switch drbd devices to secondary
mode and switch to primary mode on the slave ... but that can make
heartbeat for you ;-)
You configured your system ....mmmmh ... not in the easiest way ;-)
Oh if you knew how many time I heard that... :)
... I also configured DRBD8 today and the integration into heartbeat
is planned for next Monday. I configured it like this:
1) download latest tar.gz from drbd.org and compiled rpm packages on CentOS
2) installed packages on both nodes, doing an "modprobe drbd" do see
the module loads flawlessly
3) created a /etc/drbd.conf, copied to both nodes, tested with
"drbdadm -d up all"
4) executed "drbdadm up all" on both nodes ... watched kernel messages
with "dmesg"
5) executed "drbdadm -- --overwrite-data-of-peer primary all" on first node
6) created mountpoint on both nodes, created file system on primary
node on device /dev/drbd0, mounted filesystem ... works
7) activated /etc/init.d/drbd init script at system startup ... on
centos: chkconfig drbd on.... so the drbd devices get activated on
system startup in secondary mode, on both nodes
To enable the filesystem on the slave, the master has to be set to
secondary mode:
master# umount /dev/drbd0
master# drbdadm secondary all
slave# drbdadm primary all
slave# mount /dev/drbd0 /mnt
Works fine ... unmounted file system, made all secondary ... next step
heartbeat configuration ;-)
You normally don't need all these "low-level" drbd commands ....
drbdadm does that for you. At system startup the drbd init script
prepares the drbd devices and Heartbeat does the activation of the
primary and all the rest for you.
Regards,
Andreas
Thank you kindly for your reply!
After posting I found this page:
https://services.ibb.gatech.edu/wiki/index.php/Howto:Software:DRBD
I am half-way through following their steps on my test server. Even if
it works, your steps look easier, so I will try wiping my test servers
and trying the steps you mentioned. I will report back on Monday how it
goes.
Thank you!!
Madison
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
