mon/OSDMonitor: prime all pgs if we estimate we're approaching that anyway

If we have enough noteworthy OSDs it is likely we will touch all or most
PGs anyway, at which point it is faster and simpler to check all pgs.

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

diff --git a/src/common/config_opts.h b/src/common/config_opts.h
index aeb58f7b19f11846a8a533d8d4701ff8c7544ff5..b903dcf177ff01ad5a5f2b68676adf48d9d80e40 100644 (file)
--- a/src/common/config_opts.h
+++ b/src/common/config_opts.h
@@ -284,6 +284,7 @@ OPTION(mon_osd_allow_primary_temp, OPT_BOOL, false)  // allow primary_temp to be
 OPTION(mon_osd_allow_primary_affinity, OPT_BOOL, false)  // allow primary_affinity to be set in the osdmap
 OPTION(mon_osd_prime_pg_temp, OPT_BOOL, true)  // prime osdmap with pg mapping changes
 OPTION(mon_osd_prime_pg_temp_max_time, OPT_FLOAT, .5)  // max time to spend priming
+OPTION(mon_osd_prime_pg_temp_max_estimate, OPT_FLOAT, .25) // max estimate of pg total before we do all pgs in parallel
 OPTION(mon_osd_pool_ec_fast_read, OPT_BOOL, false) // whether turn on fast read on the pool or not
 OPTION(mon_stat_smooth_intervals, OPT_INT, 6)  // smooth stats over last N PGMap maps
 OPTION(mon_election_timeout, OPT_FLOAT, 5)  // on election proposer, max waiting time for all ACKs

diff --git a/src/mon/OSDMonitor.cc b/src/mon/OSDMonitor.cc
index 16d1102ddc7b3cecf760aa970e338f23f6a7fa41..ee4b1e877f1dbcdb1dedbc9b96022051444580bc 100644 (file)
--- a/src/mon/OSDMonitor.cc
+++ b/src/mon/OSDMonitor.cc
@@ -1014,6 +1014,23 @@ void OSDMonitor::maybe_prime_pg_temp()
   if (!all && osds.empty())
     return;
 
+  if (!all) {
+    unsigned estimate =
+      mapping->get_osd_acting_pgs(*osds.begin()).size() * osds.size();
+    if (estimate > mapping->get_num_pgs() *
+       g_conf->mon_osd_prime_pg_temp_max_estimate) {
+      dout(10) << __func__ << " estimate " << estimate << " pgs on "
+              << osds.size() << " osds >= "
+              << g_conf->mon_osd_prime_pg_temp_max_estimate << " of total "
+              << mapping->get_num_pgs() << " pgs, all"
+              << dendl;
+      all = true;
+    } else {
+      dout(10) << __func__ << " estimate " << estimate << " pgs on "
+              << osds.size() << " osds" << dendl;
+    }
+  }
+
   OSDMap next;
   next.deepish_copy_from(osdmap);
   next.apply_incremental(pending_inc);