我有一个 Service Fabric 集群,创建了 5 个应用程序。所有无状态服务和节点均处于“正常”健康状态。突然出现错误:
*“System.FM”报告属性“State”错误。 分区低于目标副本或实例计数。 事件存储服务 5 3 00000000-0000-0000-0000-000000009000
-N/P InBuild _nt1vm_0 133730403896161441*
“System.FM”报告属性“State”错误。 分区低于目标副本或实例计数。 故障分析服务 5 3 00000000-0000-0000-0000-000000005000 -N/P InBuild _nt1vm_1 133730403919308195
我尝试使用
Restart-ServiceFabricPartition
重新启动这 2 个分区,但似乎该命令已过时。然后我尝试使用 Start-ServiceFabricPartitionRestart
但它给出了超时。接下来,我尝试删除一个分区重新创建它 - 也没有成功。近一周以来效果很好。这是什么原因?我该如何解决这两个问题?
如有任何提示,我将不胜感激。 问候
设置一个与您遇到问题的集群类似的 Service Fabric 集群。确保它至少有 5 个节点并运行多个系统服务,例如 EventStoreService 和 FaultAnalysisService。
节点停用:
Start-ServiceFabricNodeTransition -NodeName <NodeName> -NodeInstanceId <InstanceId> -TransitionType Deactivate
删除节点:
Remove-ServiceFabricNodeState -NodeName <NodeName>
节点停用日志:
Node Name: _nt1vm_0
Node Instance: 133730403896161441
Health State: Warning
Node is Deactivated.
Event Time: <Timestamp>
Reason: Manual Deactivation (User action)
Service Name: fabric:/System/EventStoreService
Partition ID: 00000000-0000-0000-0000-000000009000
'System.FM' reported that partition is moving to a different node due to node deactivation.
Current Target Replica Count: 5, Available Replicas: 3.
Status: InBuild
Node: _nt1vm_0 (Deactivated)
节点重新激活后的EventStoreService:
EventStoreService
已恢复,现在具有预期的副本数量 (5)。
Service Name: fabric:/System/EventStoreService
Partition ID: 00000000-0000-0000-0000-000000009000
'System.FM' reported that partition is building replicas on new nodes.
Health State: OK
Partition reached target replica or instance count.
Expected Replica Set Size: 5, Actual Replica Set Size: 5.
Node: _nt1vm_0
Status: Ready
Event Time: <Timestamp>
重启后故障分析服务:
解决问题后,FaultAnalysisService
分区也已恢复并达到其完整副本计数。
Service Name: fabric:/System/FaultAnalysisService
Partition ID: 00000000-0000-0000-0000-000000005000
'System.FM' reported that partition is healthy.
Health State: OK
Partition is now at the target replica or instance count.
Expected Replica Set Size: 5, Actual Replica Set Size: 5.
Status: Ready
Event Time: <Timestamp>