Whamcloud - gitweb
LU-10463 osd-zfs: use 1MB RPC size by default 57/30757/3
authorAndreas Dilger <andreas.dilger@intel.com>
Sat, 6 Jan 2018 01:39:06 +0000 (18:39 -0700)
committerOleg Drokin <oleg.drokin@intel.com>
Sat, 20 Jan 2018 06:19:14 +0000 (06:19 +0000)
commitaf34a876d2ebde2b4717c920683c7fc8b5eae1cf
tree96c82617e93d7c37357670d23d1383bc3cc4f773
parent1e6bd608a89ddaa3c731102f3721b26a47c28741
LU-10463 osd-zfs: use 1MB RPC size by default

Revert back to using 1MB RPC size for ZFS back-end storage, if it
is not otherwise specified, and as long as the ZFS recordsize is
1MB or smaller.  Continue to use the ZFS recordsize if it is larger.

For ldiskfs, continue to use 4MB RPC size, unless the bigalloc
feature is enabled and has a larger chunksize.

Testing has shown that while 4MB RPC size is good for ldiskfs, it
does not improve ZFS performance, and increases IO variability in
some cases.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I4b306843667bfd960ad07ecc3886a696fd3ebbe5
Reviewed-on: https://review.whamcloud.com/30757
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
lustre/include/dt_object.h
lustre/ofd/ofd_dev.c
lustre/osd-ldiskfs/osd_handler.c
lustre/osd-zfs/osd_handler.c