[sheepdog] async flush in the object cache

[Liu Yuan]

On 06/12/2012 11:47 PM, Christoph Hellwig wrote:
> On Fri, May 18, 2012 at 10:00:22AM +0800, Liu Yuan wrote:
>> This is supposed to be an interim solution for the inefficiency of
>> current recovery against heavy IO. Currently, without it, we can't run
>> massive nodes at all, even a single node event will put the cluster into
>> useless. I'd like to remove it until we have settle down all the
>> inefficiency problem in recovery code.
> 
> What issues are still outstanding?  I'd really be a lot happier if this
> violation of block device semantics didn't make it into a release.
> 

There are still known problems about object deletion, which result in
flush failure. I am going to handle it soon. If no other noticeable
problems, I'd like too remove it from sheep. Well, seems object cache
reclaim should be included in the release to make the object cache
seriously useful, though.

Thanks,
Yuan