Whamcloud - gitweb
LU-10912 mdc: use large xattr buffers for old servers 51/32251/2
authorJohn L. Hammond <john.hammond@intel.com>
Fri, 13 Apr 2018 15:57:28 +0000 (10:57 -0500)
committerJohn L. Hammond <john.hammond@intel.com>
Thu, 3 May 2018 21:04:08 +0000 (21:04 +0000)
commit843e31073bc154cbe16f7f61a6a94d7eae2679ad
tree19502d00e154ae1bd20fda2f9daabbc46d9a00a6
parent4eb9ba01430acf7f16d9a801c8cb6b39b2c75a27
LU-10912 mdc: use large xattr buffers for old servers

Pre 2.10.1 MDTs will crash when they receive a listxattr (MDS_GETXATTR
with OBD_MD_FLXATTRLS) RPC for an orphan or dead object. So for
clients connected to these older MDTs, try to avoid sending listxattr
RPCs by making the bulk getxattr (MDS_GETXATTR with OBD_MD_FLXATTRALL)
more likely to succeed and thereby reducing the chances of falling
back to listxattr.

Lustre-change: https://review.whamcloud.com/31990
Lustre-commit: a95b85efdc97136d0fbf6959acf62919319901ed

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ia96323c47c91a44495b73be2d95705298c7f7ac9
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Minh Diep <minh.diep@intel.com>
Reviewed-on: https://review.whamcloud.com/32251
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
lustre/mdc/mdc_locks.c