b=22797 notify users that recovery-[random,double]-scale tests are not functional...

[fs/lustre-release.git] / lustre / tests / recovery-random-scale.sh

diff --git a/lustre/tests/recovery-random-scale.sh b/lustre/tests/recovery-random-scale.sh
index 2458aff..57fe798 100644 (file)
--- a/lustre/tests/recovery-random-scale.sh
+++ b/lustre/tests/recovery-random-scale.sh
@@ -18,6 +18,7 @@ CLEANUP=${CLEANUP:-""}
 init_test_env $@
 
 . ${CONFIG:=$LUSTRE/tests/cfg/$NAME.sh}
+init_logging
 
 TESTSUITELOG=${TESTSUITELOG:-$TMP/$(basename $0 .sh)}
 DEBUGLOG=$TESTSUITELOG.debug
@@ -42,11 +43,16 @@ LOAD_PID_FILE=${LOAD_PID_FILE:-$TMP/client-load.pid}
 
 remote_mds_nodsh && skip "remote MDS with nodsh" && exit 0
 
+[[ $FAILURE_MODE = SOFT ]] && \
+    log "WARNING: $0 is not functional with FAILURE_MODE = SOFT, bz22797"
+
 build_test_filter
 
 check_and_setup_lustre
 rm -rf $DIR/[df][0-9]*
 
+max_recov_time=$(max_recovery_time)
+
 # the test node needs to be insulated from a lustre failure as much as possible,
 # so not even loading the lustre modules is ideal.
 # -- umount lustre
@@ -131,7 +137,7 @@ summary_and_cleanup () {
     local result=PASS
     [ $rc -eq 0 ] || result=FAIL
 
-    log "Duraion:                $DURATION
+    log "Duration:                $DURATION
 Server failover period: $SERVER_FAILOVER_PERIOD seconds
 Exited after:           $ELAPSED seconds
 Number of failovers before exit:
@@ -189,12 +195,8 @@ fi
 start_client_loads $NODES_TO_USE
 
 echo clients load pids:
-if ! do_nodes $NODES_TO_USE "set -x; echo \$(hostname): && cat $LOAD_PID_FILE"; then
-    if [ -e $DEBUGLOG ]; then
-        exec 2<&-
-        cat $DEBUGLOG
+if ! do_nodesv $NODES_TO_USE "cat $LOAD_PID_FILE"; then
         exit 3
-    fi
 fi
 
 START_TS=$(date +%s)
@@ -291,9 +293,12 @@ while [ $ELAPSED -lt $DURATION -a ! -e $END_RUN_FILE ]; do
     # the value ( SERVER_FAILOVER_PERIOD - MINSLEEP )
     if [ $sleep -lt $MINSLEEP ]; then
         reqfail=$((reqfail +1))
-        log "WARNING: failover, client reintegration and check_client_loads time
-exceeded SERVER_FAILOVER_PERIOD - MINSLEEP !
-Failed to meet interval $reqfail times ( REQFAIL=$REQFAIL ); have sleep=$sleep"
+        log "WARNING: failover, client reintegration and check_client_loads time exceeded SERVER_FAILOVER_PERIOD - MINSLEEP !
+Failed to load the filesystem with I/O for a minimum period of $MINSLEEP $reqfail times ( REQFAIL=$REQFAIL ).
+This iteration, the load was only applied for sleep=$sleep seconds.
+Estimated max recovery time : $max_recov_time
+Probably the hardware is taking excessively long to boot.
+Try to increase SERVER_FAILOVER_PERIOD (current is $SERVER_FAILOVER_PERIOD), bug 20918"
         [ $reqfail -gt $REQFAIL ] && exit 6 
     fi