[SERVER-14644] Retrieving duplicate records with the same _id from secondaries. Created: 21/Jul/14  Updated: 10/Dec/14  Resolved: 22/Jul/14

Status: Closed
Project: Core Server
Component/s: Querying
Affects Version/s: 2.4.8
Fix Version/s: None

Type: Bug Priority: Blocker - P1
Reporter: sam flint 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-5931 Secondary reads in sharded clusters n... Closed
Operating System: Linux
Participants:

 Description   

We are getting duplicates of the same records returned from a query. I suspect this is due to a failed chunk migration.

We are running 2.4.8 and there is no chunk cleanup option in this version.

We are not able to serve all queries with the primaries because we are bound by CPU currently.

Is there a fix for this issue in 2.4.8 or an option we can use until we are able to upgrade to 2.6?

I believe this is being addressed in 2.8 as the secondaries are aware of the ranges and will take duplicates out before returning the set?



 Comments   
Comment by sam flint [ 22/Jul/14 ]

Thank you...I tried to find a ticket that had already been created, didn't so i put this one in. Thanks for the clarification.

Comment by Asya Kamsky [ 22/Jul/14 ]

additional note: this will happen even without failed migrations due to same documents existing on two shards at the same time during migration process.

Comment by Asya Kamsky [ 22/Jul/14 ]

Sam,

You filed this as P1 server ticket, however this is tracked and understood behavior described by SERVER-5931 - if you want to make sure that you only get documents owned by the shard you have to EITHER query by shard key, or query only primaries.

It sounds like these are secondary queries which don't involve the shard key, so it's a duplicate of SERVER-5931.

Asya

Generated at Thu Feb 08 03:35:30 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.