使用monstache同步MongoDB与Elasticsearch时遇ChangeStreamHistoryLost错误求助

我尝试使用monstache同步MongoDB与Elasticsearch，但运行时持续出现如下错误：

Jan 26 01:04:59 ip-172-31-1-200.eu-west-3.compute.internal monstache[25689]: ERROR 2023/01/26 01:04:59 Error starting change stream. Will retry: (ChangeStreamHistoryLost) Resume of change stream was not possible, as the resume point may no longer be in the oplog.
Jan 26 01:04:59 ip-172-31-1-200.eu-west-3.compute.internal monstache[25689]: ERROR 2023/01/26 01:04:59 Error starting change stream. Will retry: (ChangeStreamHistoryLost) Resume of change stream was not possible, as the resume point may no longer be in the oplog.
（其余重复错误省略）

已执行MongoDB oplog检查命令，结果如下：

Atlas atlas-lmkye1-shard-0 [primary] test> rs.printReplicationInfo()
actual oplog size
'1782.485107421875 MB'
---
configured oplog size
'1782.485107421875 MB'
---
log length start to end
'513558.99999570847 secs (142.66 hrs)'
---
oplog first event time
'Fri Jan 20 2023 02:45:29 GMT+0000 (Coordinated Universal Time)'
---
oplog last event time
'Thu Jan 26 2023 01:24:48 GMT+0000 (Coordinated Universal Time)'
---
now
'Thu Jan 26 2023 01:24:57 GMT+0000 (Coordinated Universal Time)'
Atlas atlas-lmkye1-shard-0 [primary] test>

错误原因分析 #

• monstache中断后重启间隔过长，超过oplog保留时长。你的oplog保留了约142.66小时（近6天），如果monstache停止运行的时间超过这个周期，它之前记录的恢复点已被oplog清理，无法继续基于该点恢复同步。
• 恢复点元数据损坏。monstache默认会在MongoDB的monstache集合中保存同步进度（恢复点），如果该集合数据损坏或丢失，会导致它尝试从一个不存在的oplog位置恢复，触发错误。
• 短时间内MongoDB写入量突增。即便oplog配置的保留时长正常，若短时间内有大量写入操作，旧的oplog会被快速覆盖，原本在保留期内的恢复点可能被提前清理。

内容的提问来源于stack exchange，提问作者nixmind