From 57f3262baa7d8931176a81cde05bc057facfc3b6 Mon Sep 17 00:00:00 2001
From: Alexander Boyko <alexander.boyko@hpe.com>
Date: Wed, 27 Apr 2022 04:36:49 -0400
Subject: [PATCH] LU-15788 lmv: try another MDT if statfs failed

With lazystatfs option statfs could fail if MDT0 is offline.
This leads to MPICH->IOR fail during FOFB tests. A client
could get statfs data from different MDT at DNE setup.

HPE-bug-id: LUS-10581
Signed-off-by: Alexander Boyko <alexander.boyko@hpe.com>
Change-Id: Icec83aba0c3ddbc749b782787b1b52faadf34a3e
Reviewed-on: https://review.whamcloud.com/47152
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Alexander Zarochentsev <alexander.zarochentsev@hpe.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
---
 lustre/lmv/lmv_obd.c           | 14 ++++++++++----
 lustre/tests/recovery-small.sh | 23 +++++++++++++++++++++++
 2 files changed, 33 insertions(+), 4 deletions(-)

diff --git a/lustre/lmv/lmv_obd.c b/lustre/lmv/lmv_obd.c
index 3a1405b..5b8c03d 100644
--- a/lustre/lmv/lmv_obd.c
+++ b/lustre/lmv/lmv_obd.c
@@ -1239,6 +1239,7 @@ static int lmv_statfs(const struct lu_env *env, struct obd_export *exp,
 	__u32 i;
 	__u32 idx;
 	int rc = 0;
+	int err = 0;
 
 	ENTRY;
 
@@ -1260,6 +1261,10 @@ static int lmv_statfs(const struct lu_env *env, struct obd_export *exp,
 		if (rc) {
 			CERROR("%s: can't stat MDS #%d: rc = %d\n",
 			       tgt->ltd_exp->exp_obd->obd_name, i, rc);
+			err = rc;
+			/* Try another MDT */
+			if (flags & OBD_STATFS_SUM)
+				continue;
 			GOTO(out_free_temp, rc);
 		}
 
@@ -1273,7 +1278,7 @@ static int lmv_statfs(const struct lu_env *env, struct obd_export *exp,
 			 * clients can be mounted as long as MDT0 is in
 			 * service */
 			*osfs = *temp;
-			break;
+			GOTO(out_free_temp, rc);
 		}
 
 		if (i == 0) {
@@ -1286,11 +1291,12 @@ static int lmv_statfs(const struct lu_env *env, struct obd_export *exp,
 			osfs->os_granted += temp->os_granted;
 		}
 	}
-
-	EXIT;
+	/* There is no stats from some MDTs, data incomplete */
+	if (err)
+		rc = err;
 out_free_temp:
 	OBD_FREE(temp, sizeof(*temp));
-	return rc;
+	RETURN(rc);
 }
 
 static int lmv_statfs_update(void *cookie, int rc)
diff --git a/lustre/tests/recovery-small.sh b/lustre/tests/recovery-small.sh
index ad27535..dbaa788 100755
--- a/lustre/tests/recovery-small.sh
+++ b/lustre/tests/recovery-small.sh
@@ -3280,6 +3280,29 @@ test_149() {
 }
 run_test 149 "skip orphan removal at umount"
 
+test_150() {
+	[ $MDSCOUNT -lt 2 ] && skip "needs >= 2 MDTs"
+	local lazystatfs
+	local max
+
+	lazystatfs=$($LCTL get_param -n llite.$FSNAME-*.lazystatfs | head -1)
+	max=$($LCTL get_param -n llite.$FSNAME-*.statahead_max | head -1)
+
+	$LCTL set_param llite.$FSNAME-*.lazystatfs=1
+	$LCTL set_param llite.$FSNAME-*.statahead_max=0
+	stack_trap "$LCTL set_param llite.$FSNAME-*.lazystatfs=$lazystatfs" EXIT
+	stack_trap "$LCTL set_param llite.$FSNAME-*.statahead_max=$max" EXIT
+	# stop a slave MDT where one ons stripe is located
+	stop mds1 -f
+
+	stack_trap "start mds1 $(mdsdevname 1) $MDS_MOUNT_OPTS && \
+		wait_recovery_complete mds1 && clients_up && true" EXIT
+
+	df $MOUNT || error "statfs failed"
+	return 0
+}
+run_test 150 "statfs when MDT0 offline with lazystatfs option"
+
 complete $SECONDS
 check_and_cleanup_lustre
 exit_status
-- 
1.8.3.1