DSE服务自动关闭

Question

你好，卡桑德拉专家，

我们观察到一种不寻常的情况，卡桑德拉服务正在自行关闭，我们已经建立了一个3节点的Cassandra集群。我们使用脚本自动启动Cassandra服务，为第一个节点启动Cassandra服务，然后休眠60秒，然后开始第二个节点的Cassandra服务，然后是第三个节点。然而，我们观察到所有节点显示UN状态2到3分钟，但经过一段时间后，按照Cassandra日志，DSE关闭启动，所有服务被优雅地关闭。请查看附加的cassandra日志，配置。文件，并推荐您在我们的配置中发现任何不寻常的地方，或者任何关于如何自动化cassandra服务启动行为的最佳实践。

上传的卡桑德拉日志：https://filetransfer.io/data-package/wRNkcTEk#link

INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:43,742  MessageServer.java:130 - internode-messaging message server finished shutting down.
INFO  [NodeSync-1] 2022-10-31 23:22:43,747  NodeSyncService.java:552 - Disabled NodeSync service
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:43,748  HintsService.java:224 - Paused hints dispatch
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:43,748  Gossiper.java:1975 - Announcing shutdown
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:45,728  MessagingService.java:695 - Waiting for messaging service to quiesce
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:45,739  MessagingService.java:695 - Waiting for messaging service to quiesce
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:45,750  MessagingService.java:695 - Waiting for messaging service to quiesce
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,275  HintsService.java:224 - Paused hints dispatch
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,277  HintsService.java:224 - Paused hints dispatch
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,282  HintsService.java:224 - Paused hints dispatch
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,361  StorageService.java:6441 - Using default non periodic tasks shutdown timeout 1 minutes
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,366  PluginManager.java:195 - Plugins are stopped.
INFO  [StorageServiceShutdownHook] 2022-10-31 23:22:50,366  DseDaemon.java:965 - DSE shutdown complete.

Answer 1

一个友好的通知，让你知道，要求别人审查日志不是你通常会要求在一个公共论坛。您需要自己查看日志以寻找线索，因为这不是付费支持服务。

我注意到，您发布的日志文件显示了同一件事的多个条目，这使我相信您的所有节点都共享相同的文件系统，这是不允许的。

例如，“没有找到主机ID”有多个条目：

WARN  [DSE main thread] 2022-11-01 00:38:05,596  StorageService.java:1212 - No host ID found, created ab3baad3-33a3-42f7-ab90-afb0976fcdd2 (Note: This should happen exactly once per node).
WARN  [DSE main thread] 2022-11-01 00:40:12,174  StorageService.java:1212 - No host ID found, created 136a81e3-0213-4575-a622-34b65545d882 (Note: This should happen exactly once per node).
WARN  [DSE main thread] 2022-11-01 00:42:18,139  StorageService.java:1212 - No host ID found, created 2ce602d5-4285-4299-8522-99b32b685f49 (Note: This should happen exactly once per node).

正如警告信息所述：

... (Note: This should happen exactly once per node).

因此，有这样的多个条目是没有意义的。

StartupChecks还报告节点使用NFS：

WARN  [DSE main thread] 2022-11-01 00:42:18,029  StartupChecks.java:1135 - saved caches directory /mnt/provision/dlpxstgcluster/dse-6.8.25_delphix_n_2/saved_caches on device 10.43.89.253:/domain0/group-115/appdata_container-118/appdata_timeflow-119/datafile (appears to be rotational with logical sector size 512) uses a not recommended file system type 'nfs', (mounted as /mnt/provision/dlpxstgcluster)
WARN  [DSE main thread] 2022-11-01 00:42:18,029  StartupChecks.java:1135 - commitlog directory /mnt/provision/dlpxstgcluster/dse-6.8.25_delphix_n_2/commitlog on device 10.43.89.253:/domain0/group-115/appdata_container-118/appdata_timeflow-119/datafile (appears to be rotational with logical sector size 512) uses a not recommended file system type 'nfs', (mounted as /mnt/provision/dlpxstgcluster)
WARN  [DSE main thread] 2022-11-01 00:42:18,029  StartupChecks.java:1135 - data directory /mnt/provision/dlpxstgcluster/dse-6.8.25_delphix_n_2/data on device 10.43.89.253:/domain0/group-115/appdata_container-118/appdata_timeflow-119/datafile (appears to be rotational with logical sector size 512) uses a not recommended file system type 'nfs', (mounted as /mnt/provision/dlpxstgcluster)

您需要整理您的集群的配置，然后再试一次。确保您没有使用共享存储。干杯!