[SERVER-1064] The query planner tries full table scans even when running --notablescan Created: 28/Apr/10  Updated: 07/Mar/14  Resolved: 14/Dec/10

Status: Closed
Project: Core Server
Component/s: Querying
Affects Version/s: 1.4.0
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Camilo Lopez Assignee: Aaron Staple
Resolution: Done Votes: 1
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Linux fc8xen x86_64 GNU/Linux


Participants:

 Description   

I'm running 1.4 mongod with --notablescan, I keep seeing this in my logs

runQuery: production.documents{ query:

{ _id: "ef1f5cd0e378bfff528ce7", accessor_id: 26 }

, orderby:

{ created_at: 1 }

}
running multiple plans
User Exception table scans not allowed: production.documents

I'm not sure if this can be related to the database sometimes locking without any other explanation.



 Comments   
Comment by Aaron Staple [ 14/Dec/10 ]

The reported logging is an artifact of the way notablescan was implemented. In notablescan mode the table scan plan is "tried" but then forced to fail immediately resulting in the UserException that was reported. In our current version a UserException does not get logged at the time it is generated, so this message no longer appears.

Generated at Thu Feb 08 02:55:57 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.