[Linuxha-users] Caught in a loop

[Simon Edwards]

Hello Michael,
	From what you're saying when the use clhaltapp on node1 both Apache
daemons stop? If that is the case you will need to modify the script
that is in place for the" stopscript" command for the application in
the /etc/cluster/apache/appconf.xml. First thing to do is to test it-
with the apache application running on node1 as well as your local
Apache, run whatever the "stopscript" is. It should only stop the
clustered Apache instance. If no httpd processes are running you now
know that this script needs to be modified in some way.

	Possibly both use the same pid file (/var/run/httpd.pid)?

Regards,
Simon

On Tue, 2005-01-25 at 14:21 +1000, Michael Mansour wrote:
> Hi Simon,
> 
> While my test cluster is running, and the test apache app is up, I made a 
> modification to node1:/etc/cluster/clconf.xml
> 
> After this I tried to build the cluster (clbuild --verbose), with the 
> following error reported at the end:
> 
> INFO  25/01/2005 04:08:55 Successfully used TCP ping for ENBD/DRBD traffic 
> between both nodes.
> ERROR 25/01/2005 04:08:55 Node x.x.x.x is already running a cldaemon process!
> ERROR 25/01/2005 04:08:55 Running the command whilst a cluster daemon is
> ERROR 25/01/2005 04:08:55 already running is NOT sensible - you not only
> ERROR 25/01/2005 04:08:55 have this built as a cluster - it is still running!
> 
> Ok, so I'll try and force a shutdown of everything (clhalt --force):
> 
> ERROR 25/01/2005 04:09:22 The cluster configuration file
> ERROR 25/01/2005 04:09:22 /etc/cluster/clconf.xml appears to have been changed
> ERROR 25/01/2005 04:09:22 but the changes have not yet been validated.
> ERROR 25/01/2005 04:09:22 Please run the clbuild(1M) command first before
> ERROR 25/01/2005 04:09:22 running this command again.
> ERROR 25/01/2005 04:09:22 Please note that if the cluster is already running
> ERROR 25/01/2005 04:09:22 you will need to use the --force argument. This will
> ERROR 25/01/2005 04:09:22 not affect running applications.
> ERROR 25/01/2005 04:09:22 clhalt exiting with error code 6.
> 
> Ok, let's try and halt the apache app (clhaltapp -A apache -V):
> 
> ERROR 25/01/2005 04:09:49 The cluster configuration file
> ERROR 25/01/2005 04:09:49 /etc/cluster/clconf.xml appears to have been changed
> ERROR 25/01/2005 04:09:49 but the changes have not yet been validated.
> ERROR 25/01/2005 04:09:49 Please run the clbuild(1M) command first before
> ERROR 25/01/2005 04:09:49 running this command again.
> ERROR 25/01/2005 04:09:49 Please note that if the cluster is already running
> ERROR 25/01/2005 04:09:49 you will need to use the --force argument. This will
> ERROR 25/01/2005 04:09:49 not affect running applications.
> 
> Ok, let's check the cluster (clstat):
> 
> ERROR 25/01/2005 04:10:06 The cluster configuration file
> ERROR 25/01/2005 04:10:06 /etc/cluster/clconf.xml appears to have been changed
> ERROR 25/01/2005 04:10:06 but the changes have not yet been validated.
> ERROR 25/01/2005 04:10:06 Please run the clbuild(1M) command first before
> ERROR 25/01/2005 04:10:06 running this command again.
> ERROR 25/01/2005 04:10:06 Please note that if the cluster is already running
> ERROR 25/01/2005 04:10:06 you will need to use the --force argument. This will
> ERROR 25/01/2005 04:10:06 not affect running applications.
> 
> Ok, let's try a force build (clbuild -V --force):
> 
> INFO  25/01/2005 04:10:27 Successfully used TCP ping for ENBD/DRBD traffic 
> between both nodes.
> ERROR 25/01/2005 04:10:28 Node x.x.x.x is already running a cldaemon process!
> ERROR 25/01/2005 04:10:28 Running the command whilst a cluster daemon is
> ERROR 25/01/2005 04:10:28 already running is NOT sensible - you not only
> ERROR 25/01/2005 04:10:28 have this built as a cluster - it is still running!
> 
> Hmmm.. how about a forced halt of the app (clhaltapp -A apache -V --force)
> 
> ERROR 25/01/2005 04:11:06 The cluster configuration file
> ERROR 25/01/2005 04:11:06 /etc/cluster/clconf.xml appears to have been changed
> ERROR 25/01/2005 04:11:06 but the changes have not yet been validated.
> ERROR 25/01/2005 04:11:06 Please run the clbuild(1M) command first before
> ERROR 25/01/2005 04:11:06 running this command again.
> ERROR 25/01/2005 04:11:06 Please note that if the cluster is already running
> ERROR 25/01/2005 04:11:06 you will need to use the --force argument. This will
> ERROR 25/01/2005 04:11:06 not affect running applications.
> 
> I'm pretty much left with no choice but to do a:
> 
> scp node2:/etc/cluster/clconf.xml node1:/etc/cluster
> 
> and then halt the app and then the cluster, make my modification to clconf.xml 
> and then build again.
> 
> Is it meant to be this way?
> 
> Michael.
> 
> 
> _______________________________________________
> Linuxha-users mailing list
> Linuxha-users at linuxha.net
> http://linuxha.net/mailman/listinfo/linuxha-users_linuxha.net
> 
>