[SERVER-33528] Read preference mode to allow intra-shard parallelism Created: 27/Feb/18  Updated: 06/Dec/22

Status: Backlog
Project: Core Server
Component/s: Replication, Sharding
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Matthew Kruse Assignee: Backlog - Query Execution
Resolution: Unresolved Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Assigned Teams:
Query Execution
Participants:

 Description   

Let's say that I have a replica set with 3 members.

On that replica set I have part of a sharded collection and on that replica set it has 9 data chunks. All of the current read preference modes will pick a winner from the replica set and query all 9 data chunks on that server for a particular read operation.

It would be nice if an additional read preference mode existed that could see that it had 3 nodes and tell each node to run my read query, but constrain that query on each replica set member by a subset of the data chunks on that member.

Run query but constrain by chunks:
member1: chunk1, chunk2, chunk3
member2: chunk4, chunk5, chunk6
member3: chunk7, chunk8, chunk9

For some read use cases this could greatly improve read round trip performance by using all replica set members concurrently.



 Comments   
Comment by Connie Chen [ 20/Jan/22 ]

Passing this to Query team as we think this should be considered as part of Global Query Planning.

Comment by Kaloian Manassiev [ 16/Mar/18 ]

Hi mkruse@adobe.com,

Thank you for your report and for the detailed description. This certainly sounds like a good idea. I edited the title a bit and put this ticket on our team's backlog to be considered along with other projects we have for sharding.

I can't promise when it will be looked at, but we will post any updates here.

Best regards,
-Kal.

Generated at Thu Feb 08 04:33:42 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.