[Pacemaker] DRBD and fencing

[Martin Aspeli]

Dejan Muhamedagic wrote:
> Hi,
>
> On Wed, Mar 10, 2010 at 11:10:31PM +0800, Martin Aspeli wrote:
>> Dejan Muhamedagic wrote:
>>> Hi,
>>>
>>> On Wed, Mar 10, 2010 at 09:02:48PM +0800, Martin Aspeli wrote:
>>>> Lars Ellenberg wrote:
>>>>
>>>>> Or, if this is as infrequent as you say it is, have those blobs in a
>>>>> regular file system on a regular partition or LV, and replace every
>>>>> "echo>    blob" with "echo>    blob&&    csync2 -x blob" (you get the idea).
>>>> Unfortunately, that'd mean modifying software I don't really have
>>>> control over. :-/
>>> If they are infrequent, running csync2 as a cron job once a
>>> minute would do, right?
>> I don't think that'd be safe for this particular application. The
>> use case is CMS file uploads.
>>
>> It's more likely that you'll see three writes in one minute and then
>> no writes for 3 days than that you'll see the same volume of writes
>> spread over the same length of time.
>>
>>> ocfs2 introduces an extra level of complexity. You don't want
>>> that unless really necessary.
>> How would that complexity manifest?
>
> By trying to support fs read/write operations on multiple hosts
> in parallel perhaps :)

Sorry, I meant more: what kind of problems would I expect to see if I 
adopted OCFS2? I'd expect some slow-down in throughput, write 
especially, but how much? What else? Is it likely to crash? Is there 
some kind of maintenance overhead?

Martin

-- 
Author of `Professional Plone Development`, a book for developers who
want to work with Plone. See http://martinaspeli.net/plone-book