Ocfs2 users - Mar 2011 - byte-range locking

I trying to run ctdb (samba) on ocfs2 (drbd, dual-primary, heartbeat,
2.6.38-rc6-git3). ctdb produced "... not locked when recovering!"
message. I
found messages about incomplete byte-range locking in drbd prior to 1.4, but
nothing concrete about latest versions. First looking to sources is unsure. Are
byte-range locking "complete" now and I may to find just a bug (or
even
misconfiguration), or this code whole absent?

-- 
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/

Hi,

On Fri, Mar 04, 2011 at 06:55:48PM +0200, Dzianis Kahanovich
wrote:
> I trying to run ctdb (samba) on ocfs2 (drbd, dual-primary, heartbeat,
> 2.6.38-rc6-git3). ctdb produced "... not locked when recovering!"
message. I
> found messages about incomplete byte-range locking in drbd prior to 1.4,
but
> nothing concrete about latest versions. First looking to sources is unsure.
Are
> byte-range locking "complete" now and I may to find just a bug
(or even
> misconfiguration), or this code whole absent?
byte-range locking is implemented in the file system, but it requires a
userspace component (and a userspace cluster stack) in order to work. As far
as I know, SLES11 (via the HA extension) is the only distro currently
shipping all the bits to make it work right.

Since you're building a mainline kernel, I don't really know what distro
you're on.
	--Mark

--
Mark Fasheh