From: Sage Weil <sage@redhat.com>
Date: Mon, 19 Aug 2019 21:32:22 +0000 (-0500)
Subject: osd/PeeringState: do not complain about past_intervals constrained by oldest epoch
X-Git-Tag: v14.2.5~176^2
X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=67e20d86e80d88c44ec6a9d471df2e023fd9bbe9;p=ceph.git

osd/PeeringState: do not complain about past_intervals constrained by oldest epoch

The start of the required interval has a floor set by the oldest osdmap
epoch we have.  That can lead to an invalid/empty required interval
(because the start is >= the end), but the PG may still have past
intervals.  That can be cause by a slow PG deletion.

No need to complain about this harmless condition.

Fixes: https://tracker.ceph.com/issues/39546
Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit b1396774a40b7b3e54c9f52b6bb6f03d31c35401)

Conflicts:
        src/osd/PeeringState.cc
- this file does not exist in nautilus: manually cherry-picked the changes into
  PG.cc
---

diff --git a/src/osd/PG.cc b/src/osd/PG.cc
index 8161087bfb47..079e9a50e78d 100644
--- a/src/osd/PG.cc
+++ b/src/osd/PG.cc
@@ -912,11 +912,15 @@ bool PG::needs_backfill() const
 
 void PG::check_past_interval_bounds() const
 {
+  auto oldest_epoch = osd->get_superblock().oldest_map;
   auto rpib = get_required_past_interval_bounds(
     info,
-    osd->get_superblock().oldest_map);
+    oldest_epoch);
   if (rpib.first >= rpib.second) {
-    if (!past_intervals.empty()) {
+    // do not warn if the start bound is dictated by oldest_map; the
+    // past intervals are presumably appropriate given the pg info.
+    if (!past_intervals.empty() &&
+        rpib.first > oldest_epoch) {
       osd->clog->error() << info.pgid << " required past_interval bounds are"
 			 << " empty [" << rpib << ") but past_intervals is not: "
 			 << past_intervals;