启动修复服务后 opscenterd.log 中的错误
Errors in opscenterd.log after starting repair service
从 Opscenter 5.x > 6.0.8 > 6.1.0 > 6.1.1 升级后,我在 opscenterd.log 中看到很多错误和警告,如下所示。我正在使用 DSE 4.8.10。我打开了维修服务,它似乎按预期工作。但是我在日志中看到警告。这些有什么需要注意的吗?
2017-06-27 01:00:00,356 [local] ERROR: The best practice rule 'Tombstone count' has failed. (MainThread)
2017-06-27 01:00:00,358 [local] ERROR: The best practice rule 'Wide partitions' has failed. (MainThread)
2017-06-27 01:00:00,451 [local] ERROR: The best practice rule 'Secondary indexes cardinality' has failed. (MainThread)
2017-06-27 13:10:11,672 [opscenterd] WARN: Unknown request 54688d7f-7c5f-4bcb-bc4d-07b7a0a79c3c (running {'started': 1498569009, 'details': u'Repair session f260e7b0-5b39-11e7-87cf-612516369059 for range (1042910172352712044,1065269862139026652] finished', 'details-type': None}) (MainThread)
2017-06-27 13:12:40,885 [opscenterd] WARN: Unknown request 341c9bc9-1c00-4771-aa64-27206ad4152a (running {'started': 1498569160, 'details': u'Repair session 4c3d5c50-5b3a-11e7-87cf-612516369059 for range (-1555782662812296764,-1538702344225528661] finished', 'details-type': None}) (MainThread)
OpsCenter 警告它无法识别 Cassandra 节点上的修复会话。它们可能是挂起的会话;在 Cassandra 2.1.x 及更早的版本中有许多与此相关的票证:
- https://issues.apache.org/jira/browse/CASSANDRA-10012
- https://issues.apache.org/jira/browse/CASSANDRA-10992
- https://issues.apache.org/jira/browse/CASSANDRA-10961
- https://issues.apache.org/jira/browse/CASSANDRA-12901
您可以尝试通过 StorageSession mbean 上的 JMX 调用 forceTerminateAllRepairSessions
来清理这些会话。或者,在集群上滚动重启应该可以解决问题。
无论如何,OpsCenter 的维修服务将正常运行。
从 Opscenter 5.x > 6.0.8 > 6.1.0 > 6.1.1 升级后,我在 opscenterd.log 中看到很多错误和警告,如下所示。我正在使用 DSE 4.8.10。我打开了维修服务,它似乎按预期工作。但是我在日志中看到警告。这些有什么需要注意的吗?
2017-06-27 01:00:00,356 [local] ERROR: The best practice rule 'Tombstone count' has failed. (MainThread)
2017-06-27 01:00:00,358 [local] ERROR: The best practice rule 'Wide partitions' has failed. (MainThread)
2017-06-27 01:00:00,451 [local] ERROR: The best practice rule 'Secondary indexes cardinality' has failed. (MainThread)
2017-06-27 13:10:11,672 [opscenterd] WARN: Unknown request 54688d7f-7c5f-4bcb-bc4d-07b7a0a79c3c (running {'started': 1498569009, 'details': u'Repair session f260e7b0-5b39-11e7-87cf-612516369059 for range (1042910172352712044,1065269862139026652] finished', 'details-type': None}) (MainThread)
2017-06-27 13:12:40,885 [opscenterd] WARN: Unknown request 341c9bc9-1c00-4771-aa64-27206ad4152a (running {'started': 1498569160, 'details': u'Repair session 4c3d5c50-5b3a-11e7-87cf-612516369059 for range (-1555782662812296764,-1538702344225528661] finished', 'details-type': None}) (MainThread)
OpsCenter 警告它无法识别 Cassandra 节点上的修复会话。它们可能是挂起的会话;在 Cassandra 2.1.x 及更早的版本中有许多与此相关的票证:
- https://issues.apache.org/jira/browse/CASSANDRA-10012
- https://issues.apache.org/jira/browse/CASSANDRA-10992
- https://issues.apache.org/jira/browse/CASSANDRA-10961
- https://issues.apache.org/jira/browse/CASSANDRA-12901
您可以尝试通过 StorageSession mbean 上的 JMX 调用 forceTerminateAllRepairSessions
来清理这些会话。或者,在集群上滚动重启应该可以解决问题。
无论如何,OpsCenter 的维修服务将正常运行。