ocfs2/cluster: Increase the live threshold for global heartbeat

We have seen isolated cases (very few, I might add) of o2hb not detecting all
live nodes on startup. One plausible reasoning for it is that other node had
a hb io delay at the same time. The live threshold set at 2 (as low as it can
be) could be increased to ameliorate the situation.

But increasing the threshold directly affects mount time. Currently it takes
around 5 secs to mount a volume in o2cb cluster with local heartbeat. Increasing
the threshold will make mounts even slower. As the issue itself is rare, we have
left things as they are for the local heartbeat mode.

However we can improve the situation for global heartbeat mode as in that mode,
we start the heartbeat much before the mount.

This patch doubles the live threshold for the start of the first region in
global heartbeat mode.

Addresses internal Oracle bug#10635585.

Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Acked-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Joel Becker <jlbec@evilplan.org>

1 files changed, 12 insertions, 1 deletions

diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 643720209a9..1d28505caff 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -1690,6 +1690,7 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 	struct file *filp = NULL;
 	struct inode *inode = NULL;
 	ssize_t ret = -EINVAL;
+	int live_threshold;
 
 	if (reg->hr_bdev)
 		goto out;
@@ -1766,8 +1767,18 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 	 * A node is considered live after it has beat LIVE_THRESHOLD
 	 * times.  We're not steady until we've given them a chance
 	 * _after_ our first read.
+	 * The default threshold is bare minimum so as to limit the delay
+	 * during mounts. For global heartbeat, the threshold doubled for the
+	 * first region.
 	 */
-	atomic_set(&reg->hr_steady_iterations, O2HB_LIVE_THRESHOLD + 1);
+	live_threshold = O2HB_LIVE_THRESHOLD;
+	if (o2hb_global_heartbeat_active()) {
+		spin_lock(&o2hb_live_lock);
+		if (o2hb_pop_count(&o2hb_region_bitmap, O2NM_MAX_REGIONS) == 1)
+			live_threshold <<= 1;
+		spin_unlock(&o2hb_live_lock);
+	}
+	atomic_set(&reg->hr_steady_iterations, live_threshold + 1);
 
 	hb_task = kthread_run(o2hb_thread, reg, "o2hb-%s",
 			      reg->hr_item.ci_name);