qa/tasks/ceph: osd_scrub_pgs: tolerate down osd at initial scrub time

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

diff --git a/qa/tasks/ceph.py b/qa/tasks/ceph.py
index 62af20537c4e6da4fb2a46ac6b68b64f8c8dccc9..a64d9aae602648a0bb452edad3dca1d07668b1d5 100644 (file)
--- a/qa/tasks/ceph.py
+++ b/qa/tasks/ceph.py
@@ -1052,7 +1052,12 @@ def osd_scrub_pgs(ctx, config):
     for role in teuthology.cluster_roles_of_type(all_roles, 'osd', cluster_name):
         log.info("Scrubbing {osd}".format(osd=role))
         _, _, id_ = teuthology.split_role(role)
-        manager.raw_cluster_cmd('osd', 'deep-scrub', id_)
+        # allow this to fail; in certain cases the OSD might not be up
+        # at this point.  we will catch all pgs below.
+        try:
+            manager.raw_cluster_cmd('osd', 'deep-scrub', id_)
+        except run.CommandFailedError:
+            pass
     prev_good = 0
     gap_cnt = 0
     loop = True