LU-10576 tests: sleep seconds to avoid using cached statfs

In sanity test_803, we check the object usage via "lfs df -i".
But the MDT may return cached statfs if two "df" calls arrive
too close each other (about 1 second). Sleep 3 seconds between
two "df" calls to avoid such trouble.

Test-Parameters: trivial envdefinitions=SLOW=yes testlist=sanity mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs mdscount=2 mdtcount=4
Signed-off-by: Fan Yong <fan.yong@intel.com>
Signed-off-by: Ben Evans <bevans@cray.com>
Change-Id: I9ce4cb6c069a88fe2b93d2d5a6304c96bdb5a0c1
Reviewed-on: https://review.whamcloud.com/31102
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>

diff --git a/lustre/tests/sanity.sh b/lustre/tests/sanity.sh
index c88c65d..7a30605 100755 (executable)
--- a/lustre/tests/sanity.sh
+++ b/lustre/tests/sanity.sh
@@ -18146,7 +18146,7 @@ test_803() {
                        error "Fail to create $DIR/$tdir/dir${idx}"
        done
 
-       sync; sleep 5
+       sync; sleep 3
        echo "before create:"
        $LFS df -i $MOUNT
        local before_used=$($LFS df -i | grep MDT0000_UUID | awk '{print $3}')
@@ -18156,7 +18156,7 @@ test_803() {
                        error "Fail to create $DIR/$tdir/foo$i"
        done
 
-       sync; sleep 5
+       sync; sleep 3
        echo "after create:"
        $LFS df -i $MOUNT
        local after_used=$($LFS df -i | grep MDT0000_UUID | awk '{print $3}')
@@ -18169,6 +18169,7 @@ test_803() {
                        error "Fail to remove $DIR/$tdir/foo$i"
        done
 
+       sleep 3 # avoid MDT return cached statfs
        wait_delete_completed
        echo "after unlink:"
        $LFS df -i $MOUNT