b=20918 improve log warning

[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 878e594..f0a7601 100644 (file)
--- a/lustre/tests/recovery-random-scale.sh
+++ b/lustre/tests/recovery-random-scale.sh
@@ -18,9 +18,13 @@ CLEANUP=${CLEANUP:-""}
 init_test_env $@
 
 . ${CONFIG:=$LUSTRE/tests/cfg/$NAME.sh}
+init_logging
 
-TESTSUITELOG=${TESTSUITELOG:-$TMP/recovery-random-scale}
+TESTSUITELOG=${TESTSUITELOG:-$TMP/$(basename $0 .sh)}
 DEBUGLOG=$TESTSUITELOG.debug
+
+cleanup_logs
+
 exec 2>$DEBUGLOG
 echo "--- env ---" >&2
 env >&2
@@ -155,13 +159,16 @@ Status: $result: rc=$rc"
 
     if [ $rc -ne 0 ]; then
         print_logs $NODES_TO_USE
+        # we are interested in only on failed clients and servers
+        local failedclients=$(cat $END_RUN_FILE | grep -v $0)
+        # FIXME: need ostfailover-s nodes also for FLAVOR=OST
+        local product=$(gather_logs $(comma_list $(osts_nodes) \
+                                 $(mdts_nodes) $mdsfailover_HOST $failedclients))
+        echo logs files $product
     fi
 
-    if [ $rc -eq 0 ]; then
-        zconf_mount $(hostname) $MOUNT
-    else
-        error "exited with rc=$rc"
-    fi
+    [ $rc -eq 0 ] && zconf_mount $(hostname) $MOUNT
+
     exit $rc
 }
 
@@ -285,9 +292,11 @@ 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.
+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