版本 2.5.16
使用streamload提交数据时，经常出现程序卡住，直到六个半小时之后客户端才超时，但是看日志里面10分钟就已经失败了。
error_info.txt (30.8 KB)
I0520 12:15:03.766275 24368 plan_fragment_executor.cpp:475] Fragment c34f7152-9953-5b42-8c74-f66f2388e6a0:(Active: 2h26m, non-child: 0.00%)
- MemoryLimit: -1.00 B
- PeakMemoryUsage: 32.28 MB
- RowsProduced: 4.10K
OlapTableSink:(Active: 14.979ms, non-child: 0.00%)- TxnID: 1092963441
- IndexNum: 1
- ReplicatedStorage: false
- CloseWaitTime: 0.000ns
- OpenTime: 4.435ms
- PrepareDataTime: 570.806us
- ConvertChunkTime: 4.611us
- ValidateDataTime: 373.436us
- RowsFiltered: 0
- RowsRead: 4.10K
- RowsReturned: 4.10K
- RpcClientSideTime: 0.000ns
- RpcServerSideTime: 0.000ns
- SendDataTime: 9.680ms
- PackChunkTime: 4.550ms
- SendRpcTime: 0.000ns
- CompressTime: 0.000ns
- SerializeChunkTime: 0.000ns
- WaitResponseTime: 0.000ns
FILE_SCAN_NODE (id=0):(Active: 2h26m, non-child: 100.00%)
- BytesRead: 0
- FileScanner: 0
- CastChunkTime: 0.000ns
- CreateChunkTime: 0.000ns
- FileReadTime: 0.000ns
- FillTime: 0.000ns
- MaterializeTime: 0.000ns
- ReadTime: 0.000ns
- NumDiskAccess: 0
- PeakMemoryUsage: 0
- RowsRead: 0
- RowsReturned: 4.10K
- RowsReturnedRate: 0
- ScannerQueueCounter: 1
- ScannerQueueTime: 4.019us
- ScannerThreadsInvoluntaryContextSwitches: 0
- ScannerThreadsTotalWallClockTime: 0.000ns
- MaterializeTupleTime(*): 0.000ns
- ScannerThreadsSysTime: 0.000ns
- ScannerThreadsUserTime: 0.000ns
- ScannerThreadsVoluntaryContextSwitches: 0
- ScannerTotalTime: 0.000ns
- TotalRawReadTime(*): 0.000ns
- TotalReadThroughput: 0.00 /sec
过了一段时间时候be节点就提示出这么个错误
集群中现在stream load 导入并发高么
大佬，查到问题了，是我们网络有问题，使用stream load的过程中传输到starrocks一直丢包，看be的日志是已经取消了，但是datax客户端的连接就一直没有断开，持续了6个多小时