[GODRIVER-1641] server selection timeout, client broken Created: 09/Jun/20  Updated: 27/Oct/23  Resolved: 10/Jun/20

Status: Closed
Project: Go Driver
Component/s: Connections
Affects Version/s: 1.3.4
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Pierre Durand Assignee: Unassigned
Resolution: Gone away Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Go 1.14.3
MongoDB 3.0.11
Linux



 Description   

I started to receive this error in several of my applications:

server selection error: server selection timeout, current topology: { Type: ReplicaSetNoPrimary, Servers: [{ Addr: server-3:27506, Type: RSPrimary, State: Connected, Average RTT: 4820465 }, { Addr: server-1:27506, Type: RSSecondary, State: Connected, Average RTT: 4725379 }, { Addr: server-2:27506, Type: RSSecondary, State: Connected, Average RTT: 4543115 }, { Addr: arbiter-2.sendinblue.com:27506, Type: RSArbiter, State: Connected, Average RTT: 589930 }, { Addr: arbiter-1.sendinblue.com:27506, Type: RSArbiter, State: Connected, Average RTT: 4579950 }, ] }

All MongoDB queries returned the same error.

In order to fix the issue, I had to restart my application (and create a new Client object).

I guess there is a "broken" connection in the pool, and it is still used even if it's not working anymore ?

This is a MongoDB 3.0.11 cluster.



 Comments   
Comment by Divjot Arora (Inactive) [ 10/Jun/20 ]

Thanks for responding! I'll go ahead and close the ticket.

Comment by Pierre Durand [ 10/Jun/20 ]

Hi,

I've asked for the server's logs, but actually we're starting to deprecate this cluster, and we're migrating data to a more recent MongoDB version.

We don't have this issue on our other (up to date) clusters.

So we don't plan to investigate more on our side.

You can close the ticket, thank you.

Comment by Divjot Arora (Inactive) [ 09/Jun/20 ]

Hi pierrre,

It is strange that the driver reports ReplicaSetNoPrimary even though there is a primary. As we mentioned on GODRIVER-1184, though, server logs would be really helpful here. It's possible the driver is mishandling some server state change. Is it possible for you to get them?

– Divjot

Comment by Pierre Durand [ 09/Jun/20 ]

I just noticed it's a duplicate of my old issue https://jira.mongodb.org/browse/GODRIVER-1184

Please close it.

Generated at Thu Feb 08 08:36:49 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.