xinyiZzz opened a new pull request, #11944:
URL: https://github.com/apache/doris/pull/11944
# Proposed changes
Issue Number: close #xxx
## Problem summary
Vectorized query ssb 3.1, the memory is mainly composed of 4 parts,
1. The memory ratio of each part in the first query: scanner_thread (57%),
hash_table_build_thread (32%), open + get_next (8.6%), prepare (5.4%)
2. The memory ratio of each part in the second query: scanner_thread (12%),
hash_table_build_thread (61.9%), open + get_next (15.3%), prepare (10.8%)
3. Change the hash table to 75% fill, which can reduce memory usage by 30%.
| Growing physical memory peak | query tracker peak | scanner_threadpeak |
_hash_table_build_thread peak | _exec_actual peak |
PlanFragmentExecutor::prepare peak
-- | -- | -- | -- | -- | -- | --
first run | 10702M | 2406M | 1368M | 770M | 207M | 132M
second run | 1291M | 1218M(少70?) | 147M | 754M | 187M | 132M
| | | | | |
75% max_fillfirst run | 10721M | 2046M | 1364M | 406M | 207M | 132M
75% max_fillsecond run | 924M | 856M(少70?) | 141M | 405M | 190M | 132M
## Checklist(Required)
1. Does it affect the original behavior:
- [ ] Yes
- [ ] No
- [ ] I don't know
2. Has unit tests been added:
- [ ] Yes
- [ ] No
- [ ] No Need
3. Has document been added or modified:
- [ ] Yes
- [ ] No
- [ ] No Need
5. Does it need to update dependencies:
- [ ] Yes
- [ ] No
6. Are there any changes that cannot be rolled back:
- [ ] Yes (If Yes, please explain WHY)
- [ ] No
## Further comments
If this is a relatively large or complex change, kick off the discussion at
[[email protected]](mailto:[email protected]) by explaining why you
chose the solution you did and what alternatives you considered, etc...
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
