[SERVER-22486] query to the router never done when the ip of the config1 lxc is down Created: 05/Feb/16  Updated: 24/Feb/16  Resolved: 24/Feb/16

Status: Closed
Project: Core Server
Component/s: Admin, Sharding
Affects Version/s: 3.0.6, 3.0.9, 3.2.1
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Guillaume Assignee: Unassigned
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
duplicates SERVER-21531 Shard's connection blocks forever whe... Closed
Related
related to SERVER-17617 One config server being down can bloc... Closed
related to SERVER-21293 Loading config data when CSRS member ... Closed
Operating System: ALL
Steps To Reproduce:

shutdown config lxc (ip and services)
query mongos => not work

Participants:

 Description   

As SERVER-21496, eatch element of my shared cluster run into lxc, so when my first confg lxc is down (and its ip is down), the router don't work (seem's trying to connect to the config server forever).
I have upgrade my cluster to 3.2.1.



 Comments   
Comment by Kelsey Schubert [ 24/Feb/16 ]

Hi visiaume,

Thank you for confirming that this issue is only affects legacy config servers. This issue is tracked in SERVER-21531.

Replica set config servers use the standard replica set read and write protocols to replicate data across the config servers. We recommend that users upgrade to CSRS to take advantage of these improvements.

As part of SERVER-22318, we will be removing code that supports legacy config servers. Therefore, we will not be correcting this behavior.

Kind regards,
Thomas

Comment by Guillaume [ 08/Feb/16 ]

This is the same log than SERVER-21496,
there is 1 mongos
When the container which host the first config is down, instantatly, the cluster don't respond, when the container is up, instantatly the cluster is up
This problem append whith legacy config, i change it to replicatset and no more error (the replicatset work fine )

Comment by Kelsey Schubert [ 05/Feb/16 ]

Hi visiaume,

Thank you for the report. So we can get a better idea of what is going on here, we will need additional information.

  1. Can you provide the log of a mongos?
  2. How many mongoses are there?
  3. How long has the config1 been down? Is this expected maintenance? When are you planning to bring it back up?
  4. In 3.2.1, are you using legacy config servers or config servers as a replica set?

Thank you,
Thomas

Generated at Thu Feb 08 04:00:32 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.