[SERVER-1290] Connection pool in mongos should be reset after mongod restarted Created: 23/Jun/10  Updated: 09/Nov/10  Resolved: 09/Nov/10

Status: Closed
Project: Core Server
Component/s: Sharding
Affects Version/s: 1.5.2
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Che-Ching Wu Assignee: Eliot Horowitz (Inactive)
Resolution: Won't Fix Votes: 2
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Participants:

 Description   

Is it possible to monitor or reset connection pool once any mongod restarted?



 Comments   
Comment by Eliot Horowitz (Inactive) [ 09/Nov/10 ]

This is not exactly the right model.
Will be addressing this with replica sets in a cleaner way.

Comment by Eliot Horowitz (Inactive) [ 28/Oct/10 ]

Still not clear on what to here.
We often detect a mongod issue when something fails.
We're not notified any other way...

For replica sets we might be able to do something

Comment by Che-Ching Wu [ 14/Jul/10 ]

Here are the log entries:

Thu Jul 15 09:57:38 WriteBackListener exception : St9exception
Thu Jul 15 09:57:43 ~ScopedDBConnection: _conn != null
Thu Jul 15 09:57:43 caught exception while doing balance: DBClientBase::findOne: transport error
...
Thu Jul 15 09:59:40 MessagingPort recv() errno:104 Connection reset by peer 172.16.3.216:35363

The reproducing steps are:
1. connect to mongos and do some operations to make some connections in the connection pool
2. restart mongod of shard
3. connect to mongos to do some operations and the error occurs

Comment by Eliot Horowitz (Inactive) [ 14/Jul/10 ]

what was the issue you saw prompting this ticket?
the exact details would be helpful

Generated at Thu Feb 08 02:56:37 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.