[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Google
  Web www.spinics.net

RE: [ogfs-users]RE: Problems with opengfs + opendlm on RHEL 3



> 
> Yes, I downloaded the source for GFS (actually we first tried 
> the binary release
> before it was open sourced) but we ended up looking at 
> opengfs because they do not
> yet support 2 node clusters (minimum is 3).  I've read 
> several pieces of documentation
> that suggest they are working on a solution for this 
> (seemingly popular!) cluster setup
> so I will stay tuned until it is supported.   Our hardware 
> does not easily lend itself to 
> setting up a 3rd node (there are only 2 available connections 
> to the shared storage).
> 

You piqued my interest to read the GFS 6.0 Admin doc:

http://www.redhat.com/docs/manuals/csgfs/admin-guide/

I don't think the 3rd node needs to be attached to shared storage ...

The 3-node minimum is for the GULM (a descendent of OpenGFS' memexp)
lock server, when used in a redundant failover mode (RLM).  It just
needs an odd number of *lock* servers simply in order to determine
quorum (e.g. 2 out of 3).

The filesystem nodes are rather independent of the lock servers; you can
run the filesytem on the same or different nodes as the lock servers.
It looks like the cluster management stuff expects to be told about
*all* of the nodes, filesystem and/or lock server, so all 3 nodes would
be considered part of the cluster, whether or not each mounts the
filesystem.

It might help your understanding to realize that there is no lock
information on the filesystem disk!  All of the lock data is stored in
the lock servers' memory, and shared via LAN, not disk.  Apparently, the
hot-standby lock server nodes sniff the LAN traffic to stay updated ...
this is a guess on my part.  

Note that OpenGFS' memexp has a mode in which it can sync lock data to
disk (presumably shared, so another memexp node could access it if the
first memexp node died), but I've never tried using this mode.  Note
that this disk is *not* the filesystem disk!

See Admin doc sections:

1.3.2, 1.5, 6.5, 11.1.3, and probably a few others.

Let us know how you do.

-- Ben --

Opinions are mine, not Intel's



> 
> Thanks for you reply!
> 
> -Marc Swanson-
>  


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_idG21&alloc_id040&opÌk
_______________________________________________
Opengfs-users mailing list
Opengfs-users@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/opengfs-users


[Site Home]     [Kernel list]     [Security]     [Bugtraq]     [Photo]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [DVD Store]     [Linux Clusters]     [Linux RAID]     [Linux Resources]

Powered by Linux