Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[BUG] Some SQL metrics are not reset between runs (GpuInsertIntoHadoopFsRelationCommand) #40

Closed
revans2 opened this issue May 29, 2020 · 1 comment · Fixed by #197
Closed

[BUG] Some SQL metrics are not reset between runs (GpuInsertIntoHadoopFsRelationCommand) #40

revans2 opened this issue May 29, 2020 · 1 comment · Fixed by #197
Assignees
@revans2
Labels
bug Something isn't working SQL part of the SQL/Dataframe plugin

Comments

@revans2
Copy link
Collaborator

revans2 commented May 29, 2020

Describe the bug
I am noticing that the SQL metrics appear to stick around between runs. The first time the query is executed, "number of files written" is 32, but then when I run the same query again it becomes 64. This seems like an issue in the metrics, but it could be something else.

Steps/Code to reproduce bug

spark.time(Q22Like(spark).write.mode("overwrite").parquet("./target/tmp"))
spark.time(Q22Like(spark).write.mode("overwrite").parquet("./target/tmp"))

And look at the metrics in between.

Expected behavior
It should behave the same as spark does.
@revans2 revans2 added bug Something isn't working ? - Needs Triage Need team to review and classify SQL part of the SQL/Dataframe plugin labels May 29, 2020
@revans2 revans2 changed the title [BUG] Some SQL metrics are not reset between runs [BUG] Some SQL metrics are not reset between runs (GpuInsertIntoHadoopFsRelationCommand) May 29, 2020
@revans2 revans2 self-assigned this Jun 17, 2020
@revans2 revans2 removed the ? - Needs Triage Need team to review and classify label Jun 17, 2020
@revans2
Copy link
Collaborator Author

revans2 commented Jun 17, 2020

So the issue is that the metrics a a singleton in GpuWriteJobStatsTracker. It looks like this was done because they wanted to get at those same metrics from multiple ways.

This was referenced Jun 17, 2020
Merged
Open
wjxiz1992 pushed a commit to wjxiz1992/spark-rapids that referenced this issue Oct 29, 2020
@tgravescs @revans2
Update instructions for Databricks 5.5 (NVIDIA#40)
3ca6a32 
* Update instructions for Databricks 5.5

* fix extra space and remove hardware
liurenjie1024 added a commit to liurenjie1024/spark-rapids that referenced this issue Jul 8, 2024
@liurenjie1024
Introduce LORE framework. (NVIDIA#11084) (NVIDIA#40)
dc02510 
* Introduce lore id

* Introduce lore id

* Fix type

* Fix type

* Conf

* style

* part

* Dump

* Introduce lore framework

* Add tests.

* Rename test case



* Fix AQE test

* Fix style

* Use args to display lore info.

* Fix build break

* Fix path in loreinfo

* Remove path

* Fix comments

* Update configs

* Fix comments

* Fix config

---------

Signed-off-by: liurenjie1024 <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@revans2 revans2

Labels
bug Something isn't working SQL part of the SQL/Dataframe plugin
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Fix metrics for writes
1 participant
@revans2