posix locking on OCFS2

[steve]

On 21/11/14 11:47, Ralph Böhme wrote:
> On Fri, Nov 21, 2014 at 09:31:27PM +1100, Amitay Isaacs wrote:
>> [Changed subject line]
>>
>> On Fri, Nov 21, 2014 at 1:08 PM, Chan Min Wai <dcmwai at gmail.com> wrote:
>>
>>> Dear Martin,
>>>
>>> Since we have touch the lock.
>>> I've some experience with it where I'd lock are define.
>>>
>>> I point the lock to the shared ocfs2 cluster.
>>>
>>> CTDB Will not start and kept on asking for lock.
>>>
>>> Which is something I'm not sure.
>>>
>>> I follow this guide.
>>>
>>> http://linuxcostablanca.blogspot.com/2014/07/samba4-cluster-for-ad-drbd-ocfs2-ctdb.html?m=1
>>>
>>> The different is that my ocfs2 are shared storage between the 2 node and
>>> thus no Drbd.
>>>
>>> Does the lock really work on this scenario?
>>>
>>> Thank you.
>>>
>>> Ps sorry to cut in as such.
>>>
>>> Regards,
>>> Min Wai, Chan
>>>
>>>
>> CTDB recovery lock requires posix file locking support on cluster file
>> system.  Looking at the OCFS2 documentation, it seems that OCFS2 does not
>> support posix file (fcntl) locking yet.
>>
>> Here is a snippet from OCFS2 wiki, (
>> https://oss.oracle.com/osswiki/OCFS2/NewFeaturesList.html)
>> Cluster aware flock(2)
>>
>>     -
>>
>>     *Required Kernel:* Linux 2.6.26
>>     -
>>
>>     *Required Tools:* Any
>>
>> The flock(2) system call is now cluster aware. File locks taken on one node
>> from userspace will interact with those taken on other nodes. All flock(2)
>> options are supported, including the kernels ability to cancel a lock
>> request when an appropriate kill signal is recieved by the user.
>> Unfortunately, POSIX file locks, also known as lockf(3) or fcntl(2) locks
>> are not yet supported in a cluster manner. We hope to have that ready in an
>> upcoming version of Ocfs2.
>
> possibly a documentation bug? Few lines below it states:
>
>    Cluster aware POSIX file locks (fcntl(), lockf())
>    * Required Kernel: Linux 2.6.28
>    * Required Tools: Any - userspace cluster stack required
-------------^^
Is CTDB a, 'userspace cluster stack'?
>
>    POSIX locks are now cluster aware. Locks taken on one node will
>    interact with those taken on another node. Due to the group
>    communication required to make these locks coherent, a userspace
>    cluster is required.
>
> This resembles my memory from back then when I was testing a Netatalk
> cluster with ocfs2. :)
>
> -Ralph
>
Ah, so it does work. But it doesn't. Or at least it's supposed to. Does 
it? It doesn't work for us and Chan at least.

Thanks,
Steve