mon/OSDMonitor: further improve prepare_command_pool_set E2BIG error message

d2c0fe9b5319a4404965c40ec92e291802ef30f6 improved this error message,
but it can be improved further by suggesting that the pg_num be increased in
smaller increments.

This commit is not cherry-picked from master because 4d3407ba93b2c15dd86aa7f97bd9f29c3508c60e
(merged for Nautilus) removed the error message in question.

Fixes: http://tracker.ceph.com/issues/39353
Signed-off-by: Nathan Cutler <ncutler@suse.com>

diff --git a/src/mon/OSDMonitor.cc b/src/mon/OSDMonitor.cc
index 9d482adeec6f5c904c42a2b82834f40054c084a3..da9c7d992464dacf239215e3187a34e3e87207c3 100644 (file)
--- a/src/mon/OSDMonitor.cc
+++ b/src/mon/OSDMonitor.cc
@@ -6798,9 +6798,9 @@ int OSDMonitor::prepare_command_pool_set(const cmdmap_t& cmdmap,
     int64_t new_pgs = n - p.get_pg_num();
     if (new_pgs > g_conf->mon_osd_max_split_count * expected_osds) {
       ss << "specified pg_num " << n << " is too large (creating "
-        << new_pgs << " new PGs on ~" << expected_osds
-        << " OSDs exceeds per-OSD max with mon_osd_max_split_count of "
-         << g_conf->mon_osd_max_split_count << ')';
+         << new_pgs << " new PGs on ~" << expected_osds
+         << " OSDs would exceed the per-OSD max of " << g_conf->mon_osd_max_split_count
+         << " given by mon_osd_max_split_count); please increase the pg_num in smaller steps";
       return -E2BIG;
     }
     p.set_pg_num(n);