[SERVER-7702] Improve chunk size accuracy in splitVector Created: 17/Nov/12  Updated: 08/Mar/13  Resolved: 15/Dec/12

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

Type: Improvement Priority: Minor - P4
Reporter: Ben Becker Assignee: Unassigned
Resolution: Won't Fix Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Participants:

 Description   

The splitVector command uses NamespaceDetails::stats.datasize and nrecords to approximate the data size for a chunk. This is not likely to result in an issue, but may result in splitting chunks too frequently or not frequently enough.

This behavior can be observed when newer documents are either larger or smaller than older documents, and have a shard key with some correlation to insertion time.



 Comments   
Comment by Eliot Horowitz (Inactive) [ 15/Dec/12 ]

We're never going to do exactly this.
Eventually a mongod will track the stats for each chunk, which will make splitvector not needed.

Generated at Thu Feb 08 03:15:21 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.