[SERVER-9997] balance is too slow Created: 24/Jun/13 Updated: 16/Dec/13 Resolved: 16/Dec/13 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Sharding |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Improvement | Priority: | Major - P3 |
| Reporter: | chensi | Assignee: | Unassigned |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Participants: |
| Description |
|
In our production enviroment, one chunk balance takes almost 300s. If we want to balance 100G data, it will take a very long time. Thanks. |
| Comments |
| Comment by Stennie Steneker (Inactive) [ 16/Dec/13 ] |
|
Hi, As at MongoDB 2.4, the balancer no longer waits for the current migration's delete phase to complete before starting the next chunk migration. Please refer to Chunk Migration Queuing in the manual for more information. Another approach to speeding up the initial chunk balancing for an empty sharded collection is to pre-split the chunk ranges. Regards, |
| Comment by chensi [ 27/Jun/13 ] |
|
actually, mongo really needs so much time to delete balanced data. I want to know whether balanced data can be stay in the source shard, and manully delete the data. |
| Comment by Eliot Horowitz (Inactive) [ 27/Jun/13 ] |
|
Do you have hardware stats and logs? |
| Comment by chensi [ 25/Jun/13 ] |
|
By the way, there are 900,00 records in our chunk, and Write querys come in as 50 req/s. Delete data in step6 is slow which we cannot accecpt. |
| Comment by chensi [ 25/Jun/13 ] |
|
hi, we don't use mms; From config.changelog, most moveChunk cost 200s. After reading code, step6: delete data cost more than 100s. |
| Comment by Eliot Horowitz (Inactive) [ 24/Jun/13 ] |
|
Would need a lot more info to give any advice.
|