osd/scrub: 'starts' messages should name PGs, not shards

By mistake, the 'scrub starts' message included the shard ID
of the primary OSD, instead of just the PG ID.

Fixes: https://tracker.ceph.com/issues/71780
Signed-off-by: Ronen Friedman <rfriedma@redhat.com>
(cherry picked from commit e8cde5811f07f0847a1aac20279aa83f57e4562d)

diff --git a/src/osd/scrubber/scrub_machine.cc b/src/osd/scrubber/scrub_machine.cc
index d6cf2c03b9679a566753dff802a02f5c09664de6..2d21c8d25c84d7cdd879bf97bd14430503d4dcd5 100644 (file)
--- a/src/osd/scrubber/scrub_machine.cc
+++ b/src/osd/scrubber/scrub_machine.cc
@@ -319,7 +319,7 @@ ActiveScrubbing::ActiveScrubbing(my_context ctx)
 
   session.m_osd_counters->inc(session.m_counters_idx->active_started_cnt);
   scrbr->get_clog()->debug()
-      << fmt::format("{} {} starts", pg_id, scrbr->get_op_mode_text());
+      << fmt::format("{} {} starts", pg_id.pgid, scrbr->get_op_mode_text());
 
   scrbr->on_init();
 }