[SERVER-50441] Hash-sharding an empty collection should not place initial chunks into a draining shard Created: 21/Aug/20  Updated: 17/Jan/24  Resolved: 17/Jan/24

Status: Closed
Project: Core Server
Component/s: Sharding
Affects Version/s: 4.0.20, 3.6.19, 4.4.0, 4.2.9
Fix Version/s: None

Type: Improvement Priority: Minor - P4
Reporter: Josef Ahmad Assignee: Backlog - Catalog and Routing
Resolution: Duplicate Votes: 0
Labels: oldshardingemea, sharding-common-backlog
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
duplicates SERVER-65666 Do not create chunks on draining shar... Closed
Related
is related to SERVER-65666 Do not create chunks on draining shar... Closed
Assigned Teams:
Catalog and Routing
Participants:
Case:

 Description   

Avoid unnecessary extra draining work as the collection gets populated.



 Comments   
Comment by Kaloian Manassiev [ 25/Aug/20 ]

The problem here is that the initial split policy policy only "kind of" acts like the balancer, thus omitting some of the more important BalancerPolicy checks.

Comment by Dinesh Chander [ 21/Aug/20 ]

Just to clarify, this is not just for hashed sharding. This also happens with normal shard key if we explicitly specify numInitialChunks.
Taking this back. This is only applicable to hashed shard key. Josef pointed this out and docs also mention that numInitialChunks is only available for hashed shard keys.

Generated at Thu Feb 08 05:22:40 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.