be节点在指定时间掉线

💬 StarRocks 用户问答 日常运维
#1

【详述】be节点多次在凌晨3点16分掉线
【StarRocks版本】例如：2.5.4
【集群规模】例如：3fe+3be（fe与be混部）
【be.out异常信息】

tracker:consistency consumption: 0
*** Aborted at 1688498161 (unix time) try “date -d @1688498161” if you are using GNU date ***
PC: @ 0x2c7e58c starrocks::vectorized::BinaryColumnBase<>::append()
*** SIGSEGV (@0x2530) received by PID 37549 (TID 0x7f736af43700) from PID 9520; stack trace: ***
@ 0x5824342 google::(anonymous namespace)::FailureSignalHandler()
@ 0x7f73ff88f630 (unknown)
@ 0x2c7e58c starrocks::vectorized::BinaryColumnBase<>::append()
@ 0x4de36ea starrocks::vectorized::NullableColumn::append()
@ 0x4dc747b starrocks::vectorized::Chunk::append_safe()
@ 0x2df9102 starrocks::vectorized::ChunksSorterHeapSort::do_done()
@ 0x2df57b5 starrocks::vectorized::ChunksSorter::done()
@ 0x2f75e50 starrocks::pipeline::PartitionSortSinkOperator::set_finishing()
@ 0x2ca6bb9 starrocks::pipeline::PipelineDriver::_mark_operator_finishing()
@ 0x2ca7ee8 starrocks::pipeline::PipelineDriver::process()
@ 0x4ec2213 starrocks::pipeline::GlobalDriverExecutor::_worker_thread()
@ 0x48c3b92 starrocks::ThreadPool::dispatch_thread()
@ 0x48be68a starrocks::thread::supervise_thread()
@ 0x7f73ff887ea5 start_thread
@ 0x7f73feea2b0d __clone
@ 0x0 (unknown)
start time: Wed Jul 5 09:55:14 CST 2023

#2

query_id:3670857c-1a9f-11ee-a0f9-0242dfbe15a3, fragment_instance:3670857c-1a9f-11ee-a0f9-0242dfbe15a5
tracker:process consumption: 857917248
tracker:query_pool consumption: 607816
tracker:load consumption: 0
tracker:metadata consumption: 4337242
tracker:tablet_metadata consumption: 1890898
tracker:rowset_metadata consumption: 772008
tracker:segment_metadata consumption: 195968
tracker:column_metadata consumption: 1478368
tracker:tablet_schema consumption: 766666
tracker:segment_zonemap consumption: 109003
tracker:short_key_index consumption: 0
tracker:column_zonemap_index consumption: 74400
tracker:ordinal_index consumption: 654736
tracker:bitmap_index consumption: 0
tracker:bloom_filter_index consumption: 0
tracker:compaction consumption: 0
tracker:schema_change consumption: 0
tracker:column_pool consumption: 172046730
tracker:page_cache consumption: 364840576
tracker:update consumption: 0
tracker:chunk_allocator consumption: 95485448
tracker:clone consumption: 0
tracker:consistency consumption: 0
*** Aborted at 1688498161 (unix time) try “date -d @1688498161” if you are using GNU date ***

#3

是不是这个query_id:3670857c-1a9f-11ee-a0f9-0242dfbe15a3, 查一下看是不是有定时调度把be crash了

#4

已经找到问题了，就是这个查询复杂度太高了。
SELECT
temp10.formName,
temp10.uname,
temp10.uid,
temp10.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp9.formName,
temp9.uname,
temp9.uid,
temp9.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp8.formName,
temp8.uname,
temp8.uid,
temp8.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp7.formName,
temp7.uname,
temp7.uid,
temp7.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp6.formName,
temp6.uname,
temp6.uid,
temp6.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp5.formName,
temp5.uname,
temp5.uid,
temp5.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp4.formName,
temp4.uname,
temp4.uid,
temp4.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp3.formName,
temp3.uname,
temp3.uid,
temp3.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp2.formName,
temp2.uname,
temp2.uid,
temp2.zpmc
FROM
(
(
SELECT
formName,
uname,
uid,
zpmc
FROM
(
SELECT
temp0.formName,
temp0.uname,
temp0.uid,
temp0.zpmc
FROM
t_form_data_144400_594082 AS temp0
)AS temp1
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595435
)
)AS temp2
)AS temp2
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595457
)
)AS temp3
)AS temp3
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595460
)
)AS temp4
)AS temp4
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595461
)
)AS temp5
)AS temp5
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595465
)
)AS temp6
)AS temp6
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595470
)
)AS temp7
)AS temp7
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595471
)
)AS temp8
)AS temp8
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595473
)
)AS temp9
)AS temp9
)
UNION ALL
(
SELECT
formName,
uname,
uid,
zpmc
FROM
t_form_data_144400_595474
)
)AS temp10
ORDER BY
formName
LIMIT 500,
500

#5

升级到2.5.8试试，

#6

:joy:扛不住

京ICP备2022026421号-1