sr版本2.5.9
执行方式：shell
执行命令：cmd=$(mysql -h xxx -u xxx -p xxx -P9030 -Ne “select * from table******” > output.csv)
执行结果：正常是有结果的，state都是EOF,但是某些执行有异常
问题1：脚本中cmd怎么捕获这个异常？ (根据if [ $? -ne 0 ]; then…取上次的返回状态是正常的)
问题2：怎么排查这个异常时是什么报错？
进一步排查发现根本原因是bdbje问题。
在上一个查询大sql时集群监控BDBJE Write 99位延迟特别高，最后会锁库：slow db lock
主节点fe相关日志报错：
2024-05-29 00:09:09,462 WARN (thrift-server-pool-76040155|76736446) [Database.logSlowLockEventIfNeeded():143] slow db lock. type: tryWriteLock, db id: 10150, db name: ODS, wait time: 3374ms, former owner id: 30, owner name: PUBLISH_VERSION, owner stack: dump thread: PUBLISH_VERSION, id: 30
2024-05-29 00:16:42,309 ERROR (JournalWriter|84) [BDBJEJournal.batchWriteCommit():407] failed to commit journal after retried 1 times! txn[] db[CloseSafeDatabase{db=22115775}]
com.sleepycat.je.rep.InsufficientAcksException: (JE 7.3.7) Transaction: -30710259 VLSN: 46,796,634, initiated at: 00:16:31. Insufficient acks for policy:SIMPLE_MAJORITY. Need replica acks: 1. Missing replica acks: 1. Timeout: 10000ms. FeederState=10.0.1.111_9010_1698301103480(2)[MASTER]
Current feeds:
10.0.1.110_9010_1698301086470: feederVLSN=46,796,636 replicaTxnEndVLSN=46,796,612
10.0.1.112_9010_1698301106434: feederVLSN=46,796,636 replicaTxnEndVLSN=46,796,632