diff options
| author | zhoushicheng <madaozhou@gmail.com> | 2015-04-10 12:10:26 +0800 |
|---|---|---|
| committer | Vijay Bellur <vbellur@redhat.com> | 2015-05-19 08:34:25 -0700 |
| commit | 8f788528e64c4c13e16f7ad2d9f667a3813e08cc (patch) | |
| tree | d9a2030ba391340fdacee35037f3685afc85a962 /tests/basic | |
| parent | 64d2a5a25b64601958c0f47ec2e6356226f97d4b (diff) | |
test: Fix sparse file self heal test
This patch solves problems caused by XFS with speculative preallocation feature on :
Test EXPECT "1" has_holes $B0/${V0}0/big2bigger would fall when XFS has not freed the preallocated blocks.
It is caused by XFS speculative preallocation feature. The test would pass if this feature is disabled.
Speculative preallocation can speed up under linux 3.8(and later).
Otherwise, the test would pass by dropping cache manually to speed up speculative preallocation.
As in http://review.gluster.org/#/c/10411/, using "( cd $M0 ; umount $M0 )" to drop caches, which is
better than "echo 3 > /proc/sys/vm/drop_caches".
drop caches operation was added in test:
tests/basic/afr/sparse-file-self-heal.t
BUG: 1206461
Change-Id: Ie2c9d1b92fa8307c44498752fdd100eb86f9689c
Signed-off-by: zhoushicheng <madaozhou@gmail.com>
Reviewed-on: http://review.gluster.org/10253
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Diffstat (limited to 'tests/basic')
| -rw-r--r-- | tests/basic/afr/sparse-file-self-heal.t | 10 |
1 files changed, 10 insertions, 0 deletions
diff --git a/tests/basic/afr/sparse-file-self-heal.t b/tests/basic/afr/sparse-file-self-heal.t index 1bc915e062c..a362494d95e 100644 --- a/tests/basic/afr/sparse-file-self-heal.t +++ b/tests/basic/afr/sparse-file-self-heal.t @@ -51,6 +51,11 @@ EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 1 TEST gluster volume heal $V0 full EXPECT_WITHIN $HEAL_TIMEOUT "0" afr_get_pending_heal_count $V0 +#If the file system of bricks is XFS and speculative preallocation is on, +#dropping cahce should be done to free speculatively pre-allocated blocks +#by XFS. +( cd $M0 ; umount $M0 ) # fail but drops kernel cache + big_md5sum_0=$(md5sum $B0/${V0}0/big | awk '{print $1}') small_md5sum_0=$(md5sum $B0/${V0}0/small | awk '{print $1}') bigger2big_md5sum_0=$(md5sum $B0/${V0}0/bigger2big | awk '{print $1}') @@ -116,6 +121,11 @@ EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 1 TEST gluster volume heal $V0 full EXPECT_WITHIN $HEAL_TIMEOUT "0" afr_get_pending_heal_count $V0 +#If the file system of bricks is XFS and speculative preallocation is on, +#dropping cahce should be done to free speculatively pre-allocated blocks +#by XFS. +( cd $M0 ; umount $M0 ) # fail but drops kernel cache + big_md5sum_0=$(md5sum $B0/${V0}0/big | awk '{print $1}') small_md5sum_0=$(md5sum $B0/${V0}0/small | awk '{print $1}') bigger2big_md5sum_0=$(md5sum $B0/${V0}0/bigger2big | awk '{print $1}') |