-
Type: Bug
-
Resolution: Duplicate
-
Priority: Critical - P2
-
None
-
Affects Version/s: 2.2.1, 2.2.3, 2.4.1
-
Component/s: Concurrency
-
None
-
Environment:Ubuntu Linux, OSX
-
ALL
-
In production we have noticed that we cannot connect to the DB while it is busy doing a long query. I tried to reproduce this locally and could not until I realized that we have auth turned on in production. When I enabled auth I also saw this behavior in my development environment. I CAN successfully connect to a DIFFERENT DB than the currently busy db.
I assume this is because auth requires a lock and the DB query is holding the lock.
I have not looked at the connection with wireshark to see if the connection completes and THEN the client blocks waiting to auth or whether there is never a connection established in the first place.
The upshot of what I'm saying is that the title of this ticket may be misleading.
In case it is important the long-running query in question is a Geo query.
- duplicates
-
SERVER-9986 Refactor $geoNear searches into new query framework
- Closed
- is related to
-
SERVER-9983 Authenticating as internal user shouldn't require a database lock
- Closed