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.

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

[GLUTEN-7641][VL] Add perf analysis scripts for TPCH workload #8065

Open
wants to merge 12 commits into
base: main
Choose a base branch
from

Conversation

marin-ma
Copy link
Contributor

No description provided.

Copy link

#7641

1. Copy the sample file [trace_result_tpch_q1.json](./sample/trace_result_tpch_q1.json) to the `$HOME/trace_result` directory
2. Open the following link in your browser to view the results: http://[your-host-ip]:1088/tracing_examples/trace_viewer.html#/tracing/test_data/trace_result_tpch_q1.json

This visualization helps to better understand performance metrics and optimize accordingly.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you add one more guide that the json file can be open by chrome about://tracing?

@FelixYBW
Copy link
Contributor

Thank you, Rong. Can you add sample of tpch_q1.nbconvert.ipynb and tpch_q1.html as example as well?

@FelixYBW
Copy link
Contributor

Let's use the perf stat tool to add PMU events as well.

@weiting-chen weiting-chen self-requested a review December 15, 2024 03:14
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Any description missing in the document?
I cannot find in which steps will use this file to install the requirements.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it in tools/workload/benchmark_velox/initialize.ipynb, please ignore this comments. Thanks.

" self.finished_nb = f\"http://{self.server}:8888/tree/{self.base_dir}/{output_nb_name}\"\n",
" if output_nb.startswith(cwd):\n",
" output_nb = os.path.relpath(output_nb, home)\n",
" self.finished_nb = f\"http://{localhost}:8888/tree/{output_nb}\"\n",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just make sure this is not a typo.
I can find there are two jupyter notebook will be launched(8888 and 8889) in the initialize.ipynb.
8888 for user usage, 8889 for perf. analysis result, right?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, port 8888 is used for accessing benchmark scripts and 8889 for perf analysis scripts. We need to start the jupyter server for perf analysis script under virtual environment, as described in initialized.ipynb, therefore they cannot share the same port.

@FelixYBW
Copy link
Contributor

Let me try soon

Copy link
Contributor

@weiting-chen weiting-chen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The code looks good to me.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants