Write cache, is write cache, is write cache?
Bruce Cran
bruce at cran.org.uk
Mon Jan 24 15:47:45 UTC 2011
On Mon, 24 Jan 2011 06:42:36 -0800
Jeremy Chadwick <freebsd at jdc.parodius.com> wrote:
> In the case of ZFS, why would all data be written to the disk every
> single time there's a write(2) operation? Performance-wise that makes
> absolutely no sense. So there is absolutely going to be a "window of
> failure" that can happen, and mirroring/raidz can recover from that,
> as a result of the checksum "stuff".
Very few people would expect data to be on disk after every write(2),
but they should expect it to be on disk after every fsync(2) - my
understanding is that databases depend on that for correct operation.
--
Bruce Cran
More information about the freebsd-fs
mailing list