From: Casey Bodley <cbodley@redhat.com>
Date: Mon, 15 Jun 2020 15:45:11 +0000 (-0400)
Subject: test/rgw: test_datalog_autotrim filters out new entries
X-Git-Tag: v15.2.13~11^2~5^2
X-Git-Url: http://git.apps.os.sepia.ceph.com/?a=commitdiff_plain;h=c04f66bd75524cdfbacc2b58bd4435b19052e02e;p=ceph.git

test/rgw: test_datalog_autotrim filters out new entries

if other sync activity is racing with test_datalog_autotrim, it can
create new datalog entries after the 'datalog autotrim' command runs

instead of asserting that the datalog is empty after trim, assert that
any entries have a marker larger than the max-marker reported by
'datalog status' before the trim

Fixes: https://tracker.ceph.com/issues/45626

Signed-off-by: Casey Bodley <cbodley@redhat.com>
(cherry picked from commit abd08f1843642e318d74dfadb0f9cf1f6b86d827)
---

diff --git a/src/test/rgw/rgw_multi/tests.py b/src/test/rgw/rgw_multi/tests.py
index 018aa4bd4ec48..6cca6dc060693 100644
--- a/src/test/rgw/rgw_multi/tests.py
+++ b/src/test/rgw/rgw_multi/tests.py
@@ -5,6 +5,7 @@ import sys
 import time
 import logging
 import errno
+import dateutil.parser
 
 try:
     from itertools import izip_longest as zip_longest  # type: ignore
@@ -83,8 +84,13 @@ def mdlog_list(zone, period = None):
 def mdlog_autotrim(zone):
     zone.cluster.admin(['mdlog', 'autotrim'])
 
-def datalog_list(zone, period = None):
-    cmd = ['datalog', 'list']
+def datalog_list(zone, args = None):
+    cmd = ['datalog', 'list'] + (args or [])
+    (datalog_json, _) = zone.cluster.admin(cmd, read_only=True)
+    return json.loads(datalog_json)
+
+def datalog_status(zone):
+    cmd = ['datalog', 'status']
     (datalog_json, _) = zone.cluster.admin(cmd, read_only=True)
     return json.loads(datalog_json)
 
@@ -962,9 +968,21 @@ def test_datalog_autotrim():
 
     # trim each datalog
     for zone, _ in zone_bucket:
+        # read max markers for each shard
+        status = datalog_status(zone.zone)
+
         datalog_autotrim(zone.zone)
-        datalog = datalog_list(zone.zone)
-        assert len(datalog) == 0
+
+        for shard_id, shard_status in enumerate(status):
+            try:
+                before_trim = dateutil.parser.isoparse(shard_status['last_update'])
+            except: # empty timestamps look like "0.000000" and will fail here
+                continue
+            entries = datalog_list(zone.zone, ['--shard-id', str(shard_id), '--max-entries', '1'])
+            if not len(entries):
+                continue
+            after_trim = dateutil.parser.isoparse(entries[0]['timestamp'])
+            assert before_trim < after_trim, "any datalog entries must be newer than trim"
 
 def test_multi_zone_redirect():
     zonegroup = realm.master_zonegroup()