长时间正常写入,突然报错:openChunks failed to find enough datanodes for write

我部署的是linux环境高可用集群,3个控制节点,2个数据节点,1.30.13版本。

之前一直能正常写入,刚才突然写入不了了,检查日志发现一直打印下列错误信息:

2021-12-19 02:07:50.493171 <INFO> :jobId=96bfc8fa-692f-1da2-3d46-8d4eb5c3be3c taskId=6262e554-f31f-d899-bf4a-15fd0a972e6b openChunks recv: 02:07:50.490997 wait: 82 us exec: 2072 us  controller03 "RemoteRun[controller03] openChunks failed to find enough datanodes for write, need 2 datanodes, found 0"
2021-12-19 02:07:50.493314 <ERROR> :StreamExecutor msgId=25151296511 length=81920 exception=RemoteRun[controller03] openChunks failed to find enough datanodes for write, need 2 datanodes, found 0
2021-12-19 02:07:50.698341 <INFO> :jobId=77f72401-2fca-bd9f-0f4b-1e444d360814 taskId=9b946fdd-56f6-518e-7141-d70b5ed3460f openChunks recv: 02:07:50.696297 wait: 82 us exec: 1945 us  controller03 "RemoteRun[controller03] openChunks failed to find enough datanodes for write, need 2 datanodes, found 0"
2021-12-19 02:07:50.699235 <ERROR> :StreamExecutor msgId=5702982143 length=81920 exception=RemoteRun[controller03] openChunks failed to find enough datanodes for write, need 2 datanodes, found 0

请教可能是什么原因?

请先 登录 后评论

1 个回答

jinzhi

这个一般是磁盘空间不足了,低于97%的存储空间(如果有多块盘,会计算总的)。或者如果有一个节点挂掉,在创建新分区时,也会有类似报错。

请先 登录 后评论