issues Search Results · repo:apache/incubator-gluten language:Scala
Filter by
2k results
(55 ms)2k results
inapache/incubator-gluten (press backspace or delete to remove)Description
In some cases to avoid R2C overhead and unexpected changes in num jobs, the child must support columnar execution.
Support such a feature during planning phase
enhancement
ArnavBalyan
- Opened yesterday
- #9166
Description
Vanilla spark(at least for 3.3.2) and CH both support writing parquet files with bloom filters. And it is enabled by
default.
Recently, CH had also supported skipping rowgroups by bloom filters: ...
enhancement
taiyang-li
- Opened yesterday
- #9164
Backend
VL (Velox)
Bug description
We found that hash columnar shuffle performed poorly and generated too many disk files.
So we used sort columnar shuffle, but it still generated more shuffle write ...
bug
triage
j7nhai
- Opened yesterday
- #9163
Backend
VL (Velox)
Bug description
After union two huge table, Gluten report an OOM.
Image
Spark version
None
Spark configurations
spark.gluten.memory.offHeap.size.in.bytes=89.6 GiB spark.gluten.memory.task.offHeap.size.in.bytes=5.6 ...
bug
triage
weiting-chen
- Opened 2 days ago
- #9160
Backend
VL (Velox)
Bug description
2025-03-27T08:59:23.6470522Z - datetime-parsing-invalid.sql *** FAILED ***
2025-03-27T08:59:23.6471095Z datetime-parsing-invalid.sql
2025-03-27T08:59:23.6471691Z ...
bug
triage
zhouyuan
- Opened 2 days ago
- #9159
Description
replace array_except with a simpler implementation. refer to arraySymmetricDifference in CH
enhancement
- Opened 2 days ago
- #9154
Description
After upgrading from 1.2 to 1.3, I found that simple queries were several hundred milliseconds slower because of closure
serialization.
enhancement
zml1206
- Opened 2 days ago
- #9152
Description
Now,
Gluten reported register issue, when set spark.shuffle.service.enabled=true;
25/03/26 14:44:37 ERROR YarnClusterScheduler: Lost executor 590 on 10.5.132.235: Unable to create executor ...
enhancement
zhixingheyi-tian
- 2
- Opened 2 days ago
- #9151
Description
JniWorkspace should not depend on spark, it may be used for flink. This is part of work for
https://github.com/apache/incubator-gluten/issues/9046
enhancement
shuai-xu
- Opened 2 days ago
- #9149
Backend
VL (Velox)
Bug description
All the shuffle fetch fail if use org.apache.spark.shuffle.sort.ColumnarShuffleManager.
Task log: img width= 1068 alt= Image src=
https://github.com/user-attachments/assets/615f7851-6b6c-45cf-b619-0d1e5ba829a7 ...
bug
triage
wangyum
- Opened 2 days ago
- #9148
Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Press the /
key to activate the search input again and adjust your query.Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Press the /
key to activate the search input again and adjust your query.