brokerload导入数据超时,请问是什么原因导致的?

【详述】brokerload任务超时,导致brokerload被canceled。信息如:type:LOAD_RUN_FAIL; msg:coordinator could not finished before job timeout。
该任务每天一次,平时3-5分钟完成。发生问题时一小时未完成。
【背景】提交brokerload从hive导入数据至starrocks
【StarRocks版本】2.2.8
【集群规模】例如:6fe+34be(fe与be混部)
【表模型】更新模型 明细模型
【导入或者导出方式】brokerload 导入hive数据
【联系方式】1252687261@qq.com
【附件】
be.log
04:04:25.899596 821398 txn_manager.cpp:212] Commit txn successfully. tablet: 53540187, txn_id: 469985487, rowsetid: 020000005749b5a95244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.900581 526807 txn_manager.cpp:212] Commit txn successfully. tablet: 53540231, txn_id: 469985487, rowsetid: 020000005749b5aa5244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.902377 526806 txn_manager.cpp:212] Commit txn successfully. tablet: 53540367, txn_id: 469985486, rowsetid: 020000005749b57f5244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.889071 526807 txn_manager.cpp:212] Commit txn successfully. tablet: 53550222, txn_id: 469985489, rowsetid: 020000005749b5c55244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 W0224 04:04:25.903563 527105 thrift_rpc_helper.cpp:65] retrying call frontend service after 100 ms, address=TNetworkAddress(hostname=10.163.138.12, port=9020), reason=THRIFT_EAGAIN (timed out) 10.163.139.16
04:04:26.202 I0224 04:04:25.884374 728196 txn_manager.cpp:212] Commit txn successfully. tablet: 47393973, txn_id: 469985482, rowsetid: 020000005749b5315244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.885331 526808 txn_manager.cpp:212] Commit txn successfully. tablet: 47393933, txn_id: 469985482, rowsetid: 020000005749b52f5244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.953707 526809 txn_manager.cpp:212] Commit txn successfully. tablet: 53540231, txn_id: 469985497, rowsetid: 020000005749b66f5244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.884816 814790 txn_manager.cpp:212] Commit txn successfully. tablet: 47393941, txn_id: 469985482, rowsetid: 020000005749b5305244020fdbae4b170878a754fc87aebe 10.163.139.16
04:04:26.202 I0224 04:04:25.955353 814791 txn_manager.cpp:212] Commit txn successfully. tablet: 53540187, txn_id: 469985497, rowsetid: 020000005749b66e5244020fdbae4b170878a754fc87aebe 10.163.139.16

日志如下,为什么会一个小时才ok?

I0224 04:05:57.773290 1086288 tablet_sink.cpp:845] Olap table sink statistics. load_id: 5fd8029d-ec8e-4072-9f9c-1f7cadfc525d, txn_id: 469983142, add chunk time(ms)/wait lock time(ms)/num: {29413109:(54228)(0)(116)} {29413108:(62464)(0)(115)} {12752826:(70092)(0)(116)} {29401588:(74239)(0)(144)} {51016694:(37435)(0)(116)} {12753235:(47840)(0)(115)} {29413111:(50203)(0)(116)} {51016690:(50761)(0)(144)} {51016693:(39048)(0)(144)} {29284733:(59598)(0)(115)} {51016695:(35598)(0)(114)} {12753232:(41169)(0)(115)} {29409994:(59570)(0)(115)} {11002:(108084)(0)(115)} {29401602:(65462)(0)(144)} {29413107:(41489)(0)(115)}

I0224 05:01:36.732954 1481781 storage_engine.cpp:569] Clearing transaction task txn_id: 469983142

发生问题时,看一下BE机器的负载情况。