[Gluster-users] glusterfs, pacemaker and Filesystem RA

[Marcel Pennewiß]

On Monday 12 September 2011 15:14:05 Uwe Weiss wrote:
> Hello Marcel, hello Samuel,

Hi,
 
> [something about OCF_CHECK_LEVEL]
> 
> Unfortunately I am not familiar enough with scripting to fix it by myself
> and to contribute it.

I'm not familiar with this, but i'll have a look at it as soon as my 
glusterfs-virtualmachines will be back after maintenance.

> Another item I would like to discuss is a bit more general.
> As Samuel pointed out the Filesystem RA (with native client) needs the
> gluster node it connects to (by using the device attribute of the
> Filsesystem RA) up and running only on startup of the client.
> 
> After that the native client detects by itself if a gluster node is gone or
> not. This is correct so far but in my setup this could be a SPOF.

AFAIR the client gets the volfile after first connect. Afterward it'll run 
even if a node gets down (if you're using replicated setup). So first connect 
/ mount is the critical moment. We're using a virtual IP which is assigned to 
an active node. Works fine here even if only one of the gluster-node is 
online.

Marcel