From 759c12fc016a6399bb179aa0f930602c87d1e0f8 Mon Sep 17 00:00:00 2001
From: Barak Sason Rofman <bsasonro@redhat.com>
Date: Tue, 24 Nov 2020 12:56:10 +0200
Subject: [PATCH 480/480] DHT/Rebalance - Ensure Rebalance reports status only
once upon stopping
Upon issuing rebalance stop command, the status of rebalance is being
logged twice to the log file, which can sometime result in an
inconsistent reports (one report states status stopped, while the other
may report something else).
This fix ensures rebalance reports it's status only once and that the
correct status is being reported.
Upstream:
> Reviewed-on: https://github.com/gluster/glusterfs/pull/1783
> fixes: #1782
> Change-Id: Id3206edfad33b3db60e9df8e95a519928dc7cb37
> Signed-off-by: Barak Sason Rofman bsasonro@redhat.com
BUG: 1286171
Change-Id: Id3206edfad33b3db60e9df8e95a519928dc7cb37
Signed-off-by: Barak Sason Rofman <bsasonro@redhat.com>
Reviewed-on: https://code.engineering.redhat.com/gerrit/218953
Tested-by: RHGS Build Bot <nigelb@redhat.com>
Reviewed-by: Csaba Henk <chenk@redhat.com>
Reviewed-by: Sunil Kumar Heggodu Gopala Acharya <sheggodu@redhat.com>
---
tests/bugs/distribute/bug-1286171.t | 75 +++++++++++++++++++++++++++++++++
xlators/cluster/dht/src/dht-common.c | 2 +-
xlators/cluster/dht/src/dht-common.h | 2 +-
xlators/cluster/dht/src/dht-rebalance.c | 63 ++++++++++++++-------------
4 files changed, 108 insertions(+), 34 deletions(-)
create mode 100644 tests/bugs/distribute/bug-1286171.t
diff --git a/tests/bugs/distribute/bug-1286171.t b/tests/bugs/distribute/bug-1286171.t
new file mode 100644
index 0000000..a2ca36f
--- /dev/null
+++ b/tests/bugs/distribute/bug-1286171.t
@@ -0,0 +1,75 @@
+#!/bin/bash
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../cluster.rc
+. $(dirname $0)/../../volume.rc
+
+# Initialize
+#------------------------------------------------------------
+cleanup;
+
+volname=bug-1286171
+
+# Start glusterd
+TEST glusterd;
+TEST pidof glusterd;
+TEST $CLI volume info;
+
+# Create a volume
+TEST $CLI volume create $volname $H0:$B0/${volname}{1,2}
+
+# Verify volume creation
+EXPECT "$volname" volinfo_field $volname 'Volume Name';
+EXPECT 'Created' volinfo_field $volname 'Status';
+
+# Start volume and verify successful start
+TEST $CLI volume start $volname;
+EXPECT 'Started' volinfo_field $volname 'Status';
+TEST glusterfs --volfile-id=$volname --volfile-server=$H0 --entry-timeout=0 $M0;
+#------------------------------------------------------------
+
+# Create a nested dir structure and some file under MP
+cd $M0;
+for i in {1..5}
+do
+ mkdir dir$i
+ cd dir$i
+ for j in {1..5}
+ do
+ mkdir dir$i$j
+ cd dir$i$j
+ for k in {1..5}
+ do
+ mkdir dir$i$j$k
+ cd dir$i$j$k
+ touch {1..300}
+ cd ..
+ done
+ touch {1..300}
+ cd ..
+ done
+ touch {1..300}
+ cd ..
+done
+touch {1..300}
+
+# Add-brick and start rebalance
+TEST $CLI volume add-brick $volname $H0:$B0/${volname}4;
+TEST $CLI volume rebalance $volname start;
+
+# Let rebalance run for a while
+sleep 5
+
+# Stop rebalance
+TEST $CLI volume rebalance $volname stop;
+
+# Allow rebalance to stop
+sleep 5
+
+# Examine the logfile for errors
+cd /var/log/glusterfs;
+failures=`grep "failures:" ${volname}-rebalance.log | tail -1 | sed 's/.*failures: //; s/,.*//'`;
+
+TEST [ $failures == 0 ];
+
+cleanup;
diff --git a/xlators/cluster/dht/src/dht-common.c b/xlators/cluster/dht/src/dht-common.c
index 23cc80c..4db89df 100644
--- a/xlators/cluster/dht/src/dht-common.c
+++ b/xlators/cluster/dht/src/dht-common.c
@@ -10969,7 +10969,7 @@ dht_notify(xlator_t *this, int event, void *data, ...)
if ((cmd == GF_DEFRAG_CMD_STATUS) ||
(cmd == GF_DEFRAG_CMD_STATUS_TIER) ||
(cmd == GF_DEFRAG_CMD_DETACH_STATUS))
- gf_defrag_status_get(conf, output);
+ gf_defrag_status_get(conf, output, _gf_false);
else if (cmd == GF_DEFRAG_CMD_START_DETACH_TIER)
gf_defrag_start_detach_tier(defrag);
else if (cmd == GF_DEFRAG_CMD_DETACH_START)
diff --git a/xlators/cluster/dht/src/dht-common.h b/xlators/cluster/dht/src/dht-common.h
index 9ec5b51..92f1b89 100644
--- a/xlators/cluster/dht/src/dht-common.h
+++ b/xlators/cluster/dht/src/dht-common.h
@@ -1252,7 +1252,7 @@ dht_fxattrop_cbk(call_frame_t *frame, void *cookie, xlator_t *this,
int32_t op_ret, int32_t op_errno, dict_t *dict, dict_t *xdata);
int
-gf_defrag_status_get(dht_conf_t *conf, dict_t *dict);
+gf_defrag_status_get(dht_conf_t *conf, dict_t *dict, gf_boolean_t log_status);
void
gf_defrag_set_pause_state(gf_tier_conf_t *tier_conf, tier_pause_state_t state);
diff --git a/xlators/cluster/dht/src/dht-rebalance.c b/xlators/cluster/dht/src/dht-rebalance.c
index d49a719..16ac16c 100644
--- a/xlators/cluster/dht/src/dht-rebalance.c
+++ b/xlators/cluster/dht/src/dht-rebalance.c
@@ -2720,7 +2720,6 @@ gf_defrag_migrate_single_file(void *opaque)
iatt_ptr = &entry->d_stat;
if (defrag->defrag_status != GF_DEFRAG_STATUS_STARTED) {
- ret = -1;
goto out;
}
@@ -3833,7 +3832,6 @@ gf_defrag_fix_layout(xlator_t *this, gf_defrag_info_t *defrag, loc_t *loc,
list_for_each_entry_safe(entry, tmp, &entries.list, list)
{
if (defrag->defrag_status != GF_DEFRAG_STATUS_STARTED) {
- ret = 1;
goto out;
}
@@ -4863,7 +4861,7 @@ out:
LOCK(&defrag->lock);
{
status = dict_new();
- gf_defrag_status_get(conf, status);
+ gf_defrag_status_get(conf, status, _gf_true);
if (ctx && ctx->notify)
ctx->notify(GF_EN_DEFRAG_STATUS, status);
if (status)
@@ -4998,7 +4996,7 @@ out:
}
int
-gf_defrag_status_get(dht_conf_t *conf, dict_t *dict)
+gf_defrag_status_get(dht_conf_t *conf, dict_t *dict, gf_boolean_t log_status)
{
int ret = 0;
uint64_t files = 0;
@@ -5095,34 +5093,35 @@ gf_defrag_status_get(dht_conf_t *conf, dict_t *dict)
gf_log(THIS->name, GF_LOG_WARNING, "failed to set time-left");
log:
- switch (defrag->defrag_status) {
- case GF_DEFRAG_STATUS_NOT_STARTED:
- status = "not started";
- break;
- case GF_DEFRAG_STATUS_STARTED:
- status = "in progress";
- break;
- case GF_DEFRAG_STATUS_STOPPED:
- status = "stopped";
- break;
- case GF_DEFRAG_STATUS_COMPLETE:
- status = "completed";
- break;
- case GF_DEFRAG_STATUS_FAILED:
- status = "failed";
- break;
- default:
- break;
- }
+ if (log_status) {
+ switch (defrag->defrag_status) {
+ case GF_DEFRAG_STATUS_NOT_STARTED:
+ status = "not started";
+ break;
+ case GF_DEFRAG_STATUS_STARTED:
+ status = "in progress";
+ break;
+ case GF_DEFRAG_STATUS_STOPPED:
+ status = "stopped";
+ break;
+ case GF_DEFRAG_STATUS_COMPLETE:
+ status = "completed";
+ break;
+ case GF_DEFRAG_STATUS_FAILED:
+ status = "failed";
+ break;
+ default:
+ break;
+ }
- gf_msg(THIS->name, GF_LOG_INFO, 0, DHT_MSG_REBALANCE_STATUS,
- "Rebalance is %s. Time taken is %.2f secs", status, elapsed);
- gf_msg(THIS->name, GF_LOG_INFO, 0, DHT_MSG_REBALANCE_STATUS,
- "Files migrated: %" PRIu64 ", size: %" PRIu64 ", lookups: %" PRIu64
- ", failures: %" PRIu64
- ", skipped: "
- "%" PRIu64,
- files, size, lookup, failures, skipped);
+ gf_msg("DHT", GF_LOG_INFO, 0, DHT_MSG_REBALANCE_STATUS,
+ "Rebalance is %s. Time taken is %.2f secs "
+ "Files migrated: %" PRIu64 ", size: %" PRIu64
+ ", lookups: %" PRIu64 ", failures: %" PRIu64
+ ", skipped: "
+ "%" PRIu64,
+ status, elapsed, files, size, lookup, failures, skipped);
+ }
out:
return 0;
}
@@ -5299,7 +5298,7 @@ gf_defrag_stop(dht_conf_t *conf, gf_defrag_status_t status, dict_t *output)
defrag->defrag_status = status;
if (output)
- gf_defrag_status_get(conf, output);
+ gf_defrag_status_get(conf, output, _gf_false);
ret = 0;
out:
gf_msg_debug("", 0, "Returning %d", ret);
--
1.8.3.1