David Binger wrote:
This is an interesting point, and it makes me wonder if
there would be interest having the fsync behavior vary on
a per-transaction basis instead of a per-storage basis.
Maybe the client submitting transactions that are just
Session-like changes could include a message to the
Chris Withers wrote:
David Binger wrote:
This is an interesting point, and it makes me wonder if
there would be interest having the fsync behavior vary on
a per-transaction basis instead of a per-storage basis.
Maybe the client submitting transactions that are just
Session-like changes could
OK, cool. Anybody feel like implementing it?
I might, but I will not have time before february, and people who know
the ZODB.code already will probably make a faster and better job.
--
Lennart Regebro, Nuxeo http://www.nuxeo.com/
CPS Content Management http://www.nuxeo.org/
Hi,
Lennart Regebro wrote:
OK, cool. Anybody feel like implementing it?
I might, but I will not have time before february, and people who know
the ZODB.code already will probably make a faster and better job.
Can you put it on line in some wiki? Maybe I have some time before
February. ;)
Christian Theune wrote:
Hi,
Lennart Regebro wrote:
OK, cool. Anybody feel like implementing it?
I might, but I will not have time before february, and people who know
the ZODB.code already will probably make a faster and better job.
Can you put it on line in some wiki?
Specifically:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Chris Withers wrote:
+1 from me too, this feels like a really good proposal :-)
+1. Sessions might be a good example of a storage using
'fsync-behavior=None': you would get the speedup, and the ability to
survive a restart, without needing to