osd: cap adjusted max mon report interval at 2/3 of timeout

This ensures that we don't throttle back mon reports so much that
the mon times out out due to no pg stat reports.  Since there is
little value is having a lower max anyway, just set this at an
upper bound (relative to the mon's timeout value).

Signed-off-by: Sage Weil <sage@redhat.com>

diff --git a/src/common/config_opts.h b/src/common/config_opts.h
index ca0ce58833901d775a0028c542398d86a6da493c..f5ea116940c53ea3954e6a6de7e89e77d8a15d73 100644 (file)
--- a/src/common/config_opts.h
+++ b/src/common/config_opts.h
@@ -656,7 +656,7 @@ OPTION(osd_pg_max_concurrent_snap_trims, OPT_U64, 2)
 OPTION(osd_heartbeat_min_healthy_ratio, OPT_FLOAT, .33)
 
 OPTION(osd_mon_heartbeat_interval, OPT_INT, 30)  // (seconds) how often to ping monitor if no peers
-OPTION(osd_mon_report_interval_max, OPT_INT, 120)
+OPTION(osd_mon_report_interval_max, OPT_INT, 600)
 OPTION(osd_mon_report_interval_min, OPT_INT, 5)  // pg stats, failures, up_thru, boot.
 OPTION(osd_mon_report_max_in_flight, OPT_INT, 2)  // max updates in flight
 OPTION(osd_pg_stat_report_interval_max, OPT_INT, 500)  // report pg stats for any given pg at least this often

diff --git a/src/osd/OSD.cc b/src/osd/OSD.cc
index 7f538c1ac1a217dab07e39186423d6b8b81d00ca..85fee628da327203d94b923f35b79de257b8b068 100644 (file)
--- a/src/osd/OSD.cc
+++ b/src/osd/OSD.cc
@@ -4065,7 +4065,10 @@ void OSD::tick_without_osd_lock()
     pg_stat_queue_lock.Lock();
     double backoff = stats_ack_timeout / g_conf->osd_mon_ack_timeout;
     double adjusted_min = cct->_conf->osd_mon_report_interval_min * backoff;
-    double adjusted_max = cct->_conf->osd_mon_report_interval_max * backoff;
+    // note: we shouldn't adjust max because it must remain < the
+    // mon's mon_osd_report_timeout (which defaults to 1.5x our
+    // value).
+    double max = cct->_conf->osd_mon_report_interval_max;
     if (!outstanding_pg_stats.empty() &&
        (now - stats_ack_timeout) > last_pg_stats_ack) {
       dout(1) << __func__ << " mon hasn't acked PGStats in "
@@ -4079,7 +4082,7 @@ void OSD::tick_without_osd_lock()
            stats_ack_timeout * g_conf->osd_stats_ack_timeout_factor);
       outstanding_pg_stats.clear();
     }
-    if (now - last_pg_stats_sent > adjusted_max) {
+    if (now - last_pg_stats_sent > max) {
       osd_stat_updated = true;
       report = true;
     } else if ((int)outstanding_pg_stats.size() >=