Re: [Xen-devel] [PATCH v10] remus drbd: Implement remus drbd replicated disk

[Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>]

Hongyang Yang writes ("Re: [PATCH v10] remus drbd: Implement remus drbd 
replicated disk"):
> Can we use the following approach:
>    The interface for per checkpoint remains async. but in the implementation,
> because the syscalls are fast enough, we can simply make it a sync call and
> call the async callback when done?

Yes, that is a perfectly legitimate thing to do.  Many of libxl's
internal asynchronous interfaces are sometimes synchronous in that
way.

Callers of such an interface in libxl must typically be aware that the
callback may happen reentrantly.  This should be mentioned as part of
the API specification.  For an example see the comment just above
libxl__bootloader_run.

I think this is probably the right thing to do for the network
buffering syscalls.


For the drbd checkpoint wait system call, we have sadly still to come
up with a good solution.  Shriram tells us that this call might, in an
error case, block for a significant fraction of a second.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel