qa/suites/rados/verify/validater: increase heartbeat grace timeout

OSD_DOWN cluster log warning is raised on rare occasions due to
the osd_hearbeat_grace timeout getting exceeded. The warning is
soon cleared. Given the nature of the test (valgrind), the
grace timeout is increased to 160 secs to avoid generating the
warning.

Fixes: https://tracker.ceph.com/issues/65768
Signed-off-by: Sridhar Seshasayee <sseshasa@redhat.com>
(cherry picked from commit aae02b6af468b1a62303655707a40d358088b0c1)

 Conflicts:
        qa/suites/rados/verify/validater/valgrind.yaml
 - Removed the setting for osd_op_thread_timeout made as part of
   https://tracker.ceph.com/issues/62992 which is not backported
   to squid.

diff --git a/qa/suites/rados/verify/validater/valgrind.yaml b/qa/suites/rados/verify/validater/valgrind.yaml
index 03accceaff2591057694228f57f353149d1b3f62..080e4988334c568765b43665d75cdf22bbfa2830 100644 (file)
--- a/qa/suites/rados/verify/validater/valgrind.yaml
+++ b/qa/suites/rados/verify/validater/valgrind.yaml
@@ -8,7 +8,8 @@ overrides:
   ceph:
     conf:
       global:
-        osd heartbeat grace: 80
+        # see https://tracker.ceph.com/issues/65768
+        osd heartbeat grace: 160
       mon:
         mon osd crush smoke test: false
       osd: