Whamcloud - gitweb
LU-3455 mdt: sync when evicting client 67/11767/14
authorHongchao Zhang <hongchao.zhang@intel.com>
Mon, 20 Oct 2014 16:28:17 +0000 (00:28 +0800)
committerOleg Drokin <oleg.drokin@intel.com>
Mon, 29 Dec 2014 17:20:36 +0000 (17:20 +0000)
commitfabd01968d0fac83e4855da5a5dfe8f68c038ceb
tree8307853a89dbc75fc34c14b7e9161482a3c52aa9
parent2b294992edce5af7b79d4300ed3aa1ea6a8db850
LU-3455 mdt: sync when evicting client

If some client is evicted from the MDT, then the underlying OSD
should be synchronized, otherwise the recovery will be failed
for the replay requests were dropped alongside with the client.

in test_20b in replay-single.sh, the occupied disk space will be
released only after DMUs are committed for ZFS, then it should
wait some more time before calling 'df' to get the disk usage info.

Test-Parameters: alwaysuploadlogs \
envdefinitions=SLOW=yes,ENABLE_QUOTA=yes,ONLY=20b \
mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs \
clientcount=4 osscount=2 mdscount=2 \
mdssizegb=10 ostsizegb=10 austeroptions=-R failover=true iscsi=1 \
testlist=replay-single,replay-single,replay-single,replay-single,replay-single

Change-Id: Ic126bd3c58a6d2da1c69c7231e88a8977417d37b
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Reviewed-on: http://review.whamcloud.com/11767
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
lustre/tests/replay-single.sh
lustre/tests/test-framework.sh