Whamcloud - gitweb
LU-11721 lod: limit statfs ffree if less than OST ffree 53/34453/4
authorAndreas Dilger <adilger@whamcloud.com>
Sun, 3 Feb 2019 00:11:00 +0000 (17:11 -0700)
committerOleg Drokin <green@whamcloud.com>
Sat, 8 Jun 2019 02:32:37 +0000 (02:32 +0000)
commit805b98c8a9812a2c2bf77b54d30e1bc40aff299e
tree7cc3ee814c3b6c8731a5d704f1a5c73a642793ff
parenta188106dfd9df0e58982401de957cc7136999f56
LU-11721 lod: limit statfs ffree if less than OST ffree

If the OSTs report fewer total free objects than the MDTs, then
use the free files count reported by the OSTs, since it represents
the minimum number of files that can be created in the filesystem
(creating more may be possible, but this depends on other factors).
This has always been what ll_statfs_internal() reports, but the
statfs aggregation via the MDT missed this step in lod_statfs().

Fix a minor defect in sanity test_418() that would let it loop
forever until the test was killed due to timeout if the "df -i"
and "lfs df -i" output did not converge.

Fixes: b500d5193360 ("LU-10018 protocol: MDT as a statfs proxy")
Fixes: 263e80f4572b ("LU-11721 tests: wait for statfs to update ...")

Lustre-change: https://review.whamcloud.com/34167
Lustre-commit: a829595add808d0fb09bab525c500d0aa6955883

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Id8d7b7edfd854f1ec30bfbbb85f04b0c973ebbe5
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Nikitas Angelinas <nangelinas@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/34453
Tested-by: Jenkins
Tested-by: Maloo <maloo@whamcloud.com>
lustre/include/obd_class.h
lustre/llite/llite_lib.c
lustre/lmv/lmv_obd.c
lustre/lod/lod_dev.c
lustre/tests/sanity.sh