[dm-devel] Re: clustered snapshots

[Mikulas Patocka]

Hi

> I've just looked at the last group of patches briefly.  So far, I like what I
> see.  I especially like the concept of locking locally first, and then
> switching to cluster.  If I add cluster lock caching - this will further speed
> things up.
> 
> I'd prefer to have the UUID that is passed in on the CTR line be more general
> - leaving the possibility for future features/flags, rather than just plopping
> an extra argument on the end.

OK, so propose some command line syntax.

> I also need to look over the
> 'reread_exceptions' stuff, as this seems a little light...  I assume you
> didn't just take my patch to allow rereads because by adding an additional
> function to the API, you simplify some of the logic?

I actually tested reread (in no-cluster situation). The trick is that we 
must pass the requests to a thread because if we invoke request from a 
request routine, it gets queued and never finished until the request 
routine finishes. The code is already there to handle dropped snapshot, so 
I just reused it.

For the actual reread I resuse most of existing code.

> Also, any progress on adding the cluster locking to the your shared 
> snapshots?
> 
> brassow

I haven't started with it, but I'll do it the same way, it doesn't seem 
hard. I am more scared of the userspace support for clustering (do you 
have any?)

Mikulas