On Wed, Jun 17, 2020 at 3:47 PM Pawel Jakub Dawidek
wrote:
> On 6/15/20 09:18, Matthew Ahrens via openzfs-developer wrote:
> > However, even so, looking up in the BRT for every single zio_free()
> > would be a substantial cost. [...]
>
> After giving it some more thought we could avoid that cost
On 6/15/20 09:18, Matthew Ahrens via openzfs-developer wrote:
> However, even so, looking up in the BRT for every single zio_free()
> would be a substantial cost. [...]
After giving it some more thought we could avoid that cost by leveraging
the fact that we operate on offsets within VDEVs.
We
On Jun 15, 2020, at 1:07 PM, Pawel Jakub Dawidek wrote:
> Exactly. Plus on-disk BRT entry size is 24 bytes and in-memory is 80
> bytes for now (vs. 392 bytes of DDT entry). Smaller structure sizes just
> delays the problem, of course, but sorting can be a big win.
>
> Also note that this table
Thank you Matt for your comments, answers inline.
On 6/15/20 09:18, Matthew Ahrens via openzfs-developer wrote:
> Cool! Couple of questions/observations:
>
> Do I understand correctly that the new data structure you're proposing
> (the BRT) maps from DVA to refcount?
Correct.
> If so, and we
’
for the space of the file) — but it seems like it’d be good to confirm this
is the expectation.
From: Allan Jude
Reply: openzfs-developer
Date: June 15, 2020 at 1:38:49 PM
To: developer@lists.open-zfs.org
Subject: Re: [developer] Manual dedup, aka --reflink support.
If we used a bit
If we used a bit in the block pointer, similar to the one we have for
dedup, we would only need to example the BRT in zio_free() if the BP had
the bit set. Of course the problem with that idea is that the 'original'
file won't have that bit set in its BP, so you'd need to search the BRT
to ensure
Cool! Couple of questions/observations:
Do I understand correctly that the new data structure you're proposing (the
BRT) maps from DVA to refcount?
If so, and we can keep this data structure sorted on disk (by DVA), we
would be more likely to get multiple useful entries when reading one block
Pawel spent a fair amount of time discussing this with me, which is good 'cause
I
apparently had been confused.
The idea and implementation he suggests sounds reasonable to me, and will
(finally!) allow offline dedup :).
Sean.
> On Jun 13, 2020, at 12:52 PM, Pawel Jakub Dawidek wrote:
>
>