-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: 3.6.14, 4.0.13, 3.4.23
-
Component/s: Replication
-
Labels:None
-
Fully Compatible
-
ALL
-
v4.2, v4.0
-
Repl 2019-11-04
Definitely affects 4.0.13, but on inspection the same logic appears to be present in 4.2.
When marking a node as too stale before moving into maintenance, an optimization occurs to only perform the maintenance mode transition once. However, if this races (an election starts), we will return out of this function without ever successfully setting maintenance mode, allowing the member to remain in secondary state until restart.
Unable to find source-code formatter for language: text. Available languages are: actionscript, ada, applescript, bash, c, c#, c++, cpp, css, erlang, go, groovy, haskell, html, java, javascript, js, json, lua, none, nyan, objc, perl, php, python, r, rainbow, ruby, scala, sh, sql, swift, visualbasic, xml, yaml
2019-08-28T12:28:51.035+0000 E REPL [rsBackgroundSync] too stale to catch up -- entering maintenance mode 2019-08-28T12:28:51.943+0000 W REPL [rsBackgroundSync] Failed to transition into maintenance mode: NotSecondary: currently running for election
https://github.com/mongodb/mongo/blob/r4.0.13/src/mongo/db/repl/bgsync.cpp#L359