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

[draft] Workflow docs #380

Closed
wants to merge 17 commits into from
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions docs/1-getting-started.rst
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@ To install this, you can use::

pip install -r requirements.txt

Now you are ready to look at the benchmarks and systems available in Benchpark,
as described in :doc:`2-benchpark-list`.
The executable is in ``benchpark/bin``, to check the version you can run::

./bin/benchpark --v

Now you are ready to look at the benchmarks and systems available in Benchpark,
and determine your workflow as described in :doc:`2-benchpark-list`.
10 changes: 6 additions & 4 deletions docs/2-benchpark-list.rst
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,10 @@
Searching Benchpark
===================

The user can search for available system and experiment specifications in Benchpark.
The easiest way to get started is to run existing experiments one existing systems, or
to modify one that is similar. You can search through the existing experiments and benchmarks with the below commands.

Search for available system and experiment specifications in Benchpark.

.. list-table:: Searching for specifications in Benchpark
:widths: 25 25 50
Expand Down Expand Up @@ -39,7 +42,6 @@ The user can search for available system and experiment specifications in Benchp
-


Once you have decided on a ``system`` you will use, and the ``benchmark/ProgrammingModel``
to run, you can proceed to :doc:`4-benchpark-setup`.

For a complete list of options, see the help menu in :doc:`benchpark-help`.

Now that you know the existing benchmarks and systems, you can determine your necessary workflow in :doc:`3-benchpark-workflow`:
70 changes: 70 additions & 0 deletions docs/3-benchpark-workflow.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,70 @@
.. Copyright 2023 Lawrence Livermore National Security, LLC and other
Benchpark Project Developers. See the top-level COPYRIGHT file for details.

SPDX-License-Identifier: Apache-2.0

=================================
Benchpark Workflow
=================================

Follow the workflow diagram to determine how many new configurations are required.

.. image:: /_static/images/new-workflow.png

If you are running on an existing ``system``, and the ``benchmark`` and ``experiment`` are
already configured, you can proceed directly to :doc:`4-benchpark-setup`.

Otherwise, you will be editing one or more of the files below.

Benchpark configuration files are organized as follows::

$benchpark
├── var
| ├── exp_repo
| | └── experiments
| | └── ${BENCHMARK1}
| | └── experiment.py
| └── sys_repo
| └── systems
| └── ${SYSTEM1}
| ├── system.py
| ├── compilers
| └── externals
└── repo
├── ${BENCHMARK1}
│ ├── application.py
│ └── package.py
└── repo.yaml



System Specification
--------------------
Files under ``benchpark/var/sys_repo/systems/${SYSTEM}`` provide the specification
of the software stack on your system:

* Find a similar ``system``: :doc:`identifying-similar-system`

* Add or edit a ``system``: :doc:`add-a-system-config`


Benchmark Specification
-----------------------
If you would like to modify a specification of your benchmark,
you can do so by upstreaming changes to Spack and/or Ramble,
or working on your benchmark specification in ``benchpark/repo/${BENCHMARK}``:

* Add a ``benchmark``: :doc:`add-a-benchmark`


Experiment Specification
------------------------
Files under ``benchpark/var/exp_repo/experiments/${BENCHMARK}/${ProgrammingModel}``
provide the specifications for the experiments.
If you would like to make changes to your experiments, such as enabling
specific tools to measure the performance of your experiments,
you can manually edit the specifications in ``ramble.yaml``:

* Add/edit an ``experiment``: :doc:`add-an-experiment`


181 changes: 0 additions & 181 deletions docs/3-opt-edit-experiment.rst

This file was deleted.

2 changes: 2 additions & 0 deletions docs/5-build-experiment.rst
Original file line number Diff line number Diff line change
Expand Up @@ -33,3 +33,5 @@ Ramble will build the source code and set up the following workspace directory s

If you edit any of the files, see :doc:`FAQ-what-to-rerun` to determine
whether you need to re-do any of the previous steps.

Once that is complete, you are ready for :doc:`6-run-experiment`.
2 changes: 2 additions & 0 deletions docs/6-run-experiment.rst
Original file line number Diff line number Diff line change
Expand Up @@ -42,3 +42,5 @@ Further, if the benchmark has restart capability, existing output may alter the
benchpark would run in the second run. Generally, we would advise the user to remove the
``$workspace/experiments`` directory before re-running the experiments using
``ramble -P -D . on``.

Once you have run your experiment you can try :doc:`7-analyze-experiment`.
Binary file added docs/_static/images/new-workflow.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
11 changes: 10 additions & 1 deletion docs/add-a-benchmark.rst
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,10 @@
Adding a Benchmark
==================

The following system-independent specification is required for each ${Benchmark1}:
Add a New Benchmark
------------------------

The following system-independent specification is required for each ${Benchmark1} in ``benchpark/repo/${Benchmark1}``:

- ``package.py`` is a Spack specification that defines how to build and install ${Benchmark1}.
- ``application.py`` is a Ramble specification that defines the ${Benchmark1} input and parameters.
Expand All @@ -20,3 +23,9 @@ By default, Benchpark will use ${Benchmark1} specifications (``package.py`` and
provided in the Spack and Ramble repos.
It is possible to overwrite the benchmark specifications provided in the Spack and Ramble repos;
see :doc:`FAQ-benchpark-repo` for details.


Validate a Benchmark (Needed?)
------------------------

Now that the benchmark has been created/updated/validated you need to configure at least one experiment, see :doc:`add-an-experiment`.
Loading