LU-13535 lfsck: fix possible PFL layout corruption

[fs/lustre-release.git] / lustre / tests / sanity-lfsck.sh

diff --git a/lustre/tests/sanity-lfsck.sh b/lustre/tests/sanity-lfsck.sh
index 1987a48..f19d464 100644 (file)
--- a/lustre/tests/sanity-lfsck.sh
+++ b/lustre/tests/sanity-lfsck.sh
@@ -79,8 +79,8 @@ SHOW_LAYOUT="do_facet $SINGLEMDS \
                $LCTL get_param -n mdd.${MDT_DEV}.lfsck_layout"
 SHOW_LAYOUT_ON_OST="do_facet ost1 \
                $LCTL get_param -n obdfilter.${OST_DEV}.lfsck_layout"
-MOUNT_OPTS_SCRUB="-o user_xattr"
-MOUNT_OPTS_NOSCRUB="-o user_xattr,noscrub"
+MOUNT_OPTS_SCRUB="$MDS_MOUNT_OPTS -o user_xattr"
+MOUNT_OPTS_NOSCRUB="$MDS_MOUNT_OPTS -o user_xattr,noscrub"
 MOUNT_OPTS_SKIP_LFSCK="-o user_xattr,skip_lfsck"
 
 lfsck_prep() {
@@ -2189,6 +2189,28 @@ test_18b() {
        cancel_lru_locks mdc
        cancel_lru_locks osc
 
+       # dryrun mode only check orphans, not repaie
+       echo "Trigger layout LFSCK --dryrun to find out orphan OST-object"
+       $START_LAYOUT --dryrun -o -r ||
+               error "Fail to start layout LFSCK in dryrun mode"
+       wait_all_targets_blocked layout completed 2
+
+       local PARAMS=$($SHOW_LAYOUT | awk '/^param/ { print $2 }')
+       [ "$PARAMS" == "dryrun,all_targets,orphan" ] ||
+               error "Expect 'dryrun,all_targets,orphan', got '$PARAMS'"
+
+       local orphans=$(do_facet mds1 $LCTL get_param -n \
+                       mdd.$(facet_svc mds1).lfsck_layout |
+                       awk '/^inconsistent_orphan/ { print $2 }')
+       [ $orphans -eq 3 ] ||
+               error "Expect 3 found on mds1, but got: $orphans"
+
+       # orphan parents should not be created
+       local subdir
+       for subdir in $MOUNT/.lustre/lost+found/*; do
+               [ ! "$(ls -A $subdir)" ] || error "$subdir not empty"
+       done
+
        echo "Trigger layout LFSCK on all devices to find out orphan OST-object"
        $START_LAYOUT -r -o || error "(1) Fail to start LFSCK for layout!"
 
@@ -4772,7 +4794,7 @@ test_31b() {
 
        #define OBD_FAIL_LFSCK_BAD_NAME_HASH    0x1628
        $LCTL set_param fail_loc=0x1628 fail_val=1
-       createmany -d $DIR/$tdir/striped_dir/d $MDSCOUNT ||
+       createmany -d $DIR/$tdir/striped_dir/d $((MDSCOUNT * 5)) ||
                error "(2) Fail to create file under striped directory"
        $LCTL set_param fail_loc=0 fail_val=0
 
@@ -4785,13 +4807,14 @@ test_31b() {
        local repaired=$(do_facet mds2 $LCTL get_param -n \
                         mdd.$(facet_svc mds2).lfsck_namespace |
                         awk '/^name_hash_repaired/ { print $2 }')
+       echo "repaired $repaired name entries with bad hash"
        [ $repaired -ge 1 ] ||
                error "(5) Fail to repair bad name hash: $repaired"
 
        umount_client $MOUNT || error "(6) umount failed"
        mount_client $MOUNT || error "(7) mount failed"
 
-       for ((i = 0; i < $MDSCOUNT; i++)); do
+       for ((i = 0; i < $((MDSCOUNT * 5)); i++)); do
                stat $DIR/$tdir/striped_dir/d$i ||
                        error "(8) Fail to stat d$i after LFSCK"
                rmdir $DIR/$tdir/striped_dir/d$i ||
@@ -5764,6 +5787,35 @@ test_39()
 }
 run_test 39 "LFSCK does not break foreign dir and reverse is also true"
 
+test_40a() {
+       [[ $MDSCOUNT -ge 2 ]] || skip "needs >= 2 MDTs"
+
+       check_mount_and_prep
+       $LFS mkdir -i 1 $DIR/$tdir/dir1
+       $LFS setstripe -E 1M -c1 -S 1M -E 128M -c2 -S 4M -E eof $DIR/$tdir/dir1
+
+       touch $DIR/$tdir/dir1/f1
+       local layout1=$(get_layout_param $DIR/$tdir/dir1/f1)
+
+       echo "Migrate $DIR/$tdir/dir1 from MDT1 to MDT0"
+       $LFS migrate -m 0 $DIR/$tdir/dir1
+
+       echo "trigger LFSCK for layout"
+       do_facet $SINGLEMDS $LCTL lfsck_start -M ${MDT_DEV} -t layout -r
+
+       wait_update_facet $SINGLEMDS "$LCTL get_param -n \
+               mdd.${MDT_DEV}.lfsck_layout |
+               awk '/^status/ { print \\\$2 }'" "completed" 32 || {
+               $SHOW_LAYOUT
+               error "(2) unexpected status"
+       }
+
+       local layout2=$(get_layout_param $DIR/$tdir/dir1/f1)
+
+       [[ "$layout1" == "$layout2" ]] || error "layout lost after lfsck"
+}
+run_test 40a "LFSCK correctly fixes lmm_oi in composite layout"
+
 # restore MDS/OST size
 MDSSIZE=${SAVED_MDSSIZE}
 OSTSIZE=${SAVED_OSTSIZE}