I created a set of tickets for what I think needs to happen:
http://tracker.ceph.com/issues/14031. There's a doc linked from that
ticket with more details on the design.
-Sam
On Fri, Nov 13, 2015 at 10:26 AM, Tianshan Qu wrote:
> Hi:
>
> sorry for half send.
>
> I'm
Lazily persisting the intermediate entries would certainly also work,
but there's an argument that it needlessly adds to the write
transaction.
Actually, we probably want to avoid having small writes be full stripe
writes -- with a 8+3 code the difference between modifying a single
stripelet and
This scheme fundamentally relies on the temporary objects "gracefully"
transitioning into being portions of full-up long-term durable objects.
This means that if the allocation size for a temporary object significantly
mismatches the size of the mutation (partial stripe write) you're creating a
Hi:
sorry for half send.
I'm working on this recently. I think we should distinguish the scene
of sequence write and small write.
The first, overwrite is enough. The second small write's typical
optimal is a log based write, which do not need ec's reconstruct and
avoid parity chunk's extra
On Thu, 12 Nov 2015, Samuel Just wrote:
> I was present for a discussion about allowing EC overwrites and thought it
> would be good to summarize it for the list:
>
> Commit Protocol:
> 1) client sends write to primary
> 2) primary reads in partial stripes needed for partial stripe
> overwrites