Am Son, 04 Jun 2000 schriebst Du:
>Hi,
>
>Sorry to disturb you again with my Slackware ;) but I desesperatly cannot
>reach the "step 3 " of the drbd script without failure... :
>
>www:/etc/rc.d/init.d# ./drbd start
>Loading DRBD module...ok
>Configuring DRBD resource drbd0...ok
>Waiting for DRBD resource drbd0 to resynchronize...failure!
>
You are right, this message is a bit missleading.
The fact is that, if the other side is not in primary state,
resynchronization is "failing", because there is nothing to
resynchronize :)
A better message would be "not-needed" or "skiped", but this
emerged from a RedHat environment and RH's action function can
only print "OK" or "failure" :)
>And there is no delay to synchronize, the failure appears immediatly..
>The network seems Ok, Im using 192.168.150.3 and 4 for the IP of the 2
>servers.
Well, I was asuming, that there was nothing to resynchronize. To test it
you could try:
1) Bring up both nodes.
2) Simulate a failure on the master node.
3) Modify the data on the slave node (which should offer the service by now)
4) Restart the master node. (now resynchronization should work)
Without heartbeat:
1) Connect the device-pair.
2) Unload the module on one machine (The failure of the master node :)
3) Modify the data on the other (datadisk start && modify fs)
4) Run the "drbd" script on the machine without module again.
(Now you should see a working resynchronize)
>So my question is, like always: the problem appears because i'm on a
>Slackware, or cause my config wasn't OK ?
I do not think that it is Slackware specific.
>I'm working to modify your program to make it "Debian and Slack-friendly "
>;) I'll tell you when that will be done..
>
That would be great.
PS: Bist du aus dem deutschsprachigen Teil der Schweiz ?
-Philipp
>Hi,
>
>Sorry to disturb you again with my Slackware ;) but I desesperatly cannot
>reach the "step 3 " of the drbd script without failure... :
>
>www:/etc/rc.d/init.d# ./drbd start
>Loading DRBD module...ok
>Configuring DRBD resource drbd0...ok
>Waiting for DRBD resource drbd0 to resynchronize...failure!
>
You are right, this message is a bit missleading.
The fact is that, if the other side is not in primary state,
resynchronization is "failing", because there is nothing to
resynchronize :)
A better message would be "not-needed" or "skiped", but this
emerged from a RedHat environment and RH's action function can
only print "OK" or "failure" :)
>And there is no delay to synchronize, the failure appears immediatly..
>The network seems Ok, Im using 192.168.150.3 and 4 for the IP of the 2
>servers.
Well, I was asuming, that there was nothing to resynchronize. To test it
you could try:
1) Bring up both nodes.
2) Simulate a failure on the master node.
3) Modify the data on the slave node (which should offer the service by now)
4) Restart the master node. (now resynchronization should work)
Without heartbeat:
1) Connect the device-pair.
2) Unload the module on one machine (The failure of the master node :)
3) Modify the data on the other (datadisk start && modify fs)
4) Run the "drbd" script on the machine without module again.
(Now you should see a working resynchronize)
>So my question is, like always: the problem appears because i'm on a
>Slackware, or cause my config wasn't OK ?
I do not think that it is Slackware specific.
>I'm working to modify your program to make it "Debian and Slack-friendly "
>;) I'll tell you when that will be done..
>
That would be great.
PS: Bist du aus dem deutschsprachigen Teil der Schweiz ?
-Philipp