MBU products perf&scale library

OPL library

Contains various helpers for not only RHCloud performance testing.

Repository Structure

This repository uses a split-directory workspace layout to cleanly separate lightweight core tools from those requiring heavy dependencies (like kafka-python, psycopg2-binary, locust). Python's implicit namespace packages (PEP 420) are used so that imports like import opl.some_module work seamlessly across both directories.

• core/opl/ - Contains the lightweight, minimal-dependency core tools.
• extras/opl/ - Contains tools and generators that require heavier dependencies.

Generic command-line tools

There is couple of tools used mostly to integrate with CPT and comp.:

• cluster_read.py - Library (and command-line tool) to get data in a ways defined in a config file. It can execute command, read from Prometheus or Grafana. It is used by status_data.py to enritch status data files with facts (as defined in per-test config file). Usually it is info about the cluster (using oc ... | jq ... commands) or monitoring data from Prometheus. E.g. see core/opl/cluster_read_example.yaml.
• pass_or_fail.py - Tool which uses simple statistic in an attempt to decide if latest test result is PASS or FAIL when compared to historical results. Example of the config file is in core/opl/investigator/sample_config.yaml.
• junit_cli.py - Script to manipulate JUnit XML file from CLI. It is used by Satellite CPT.
• status_data.py - Library (and command-line tool) to manipulate status data files.
• status_data_report.py - Tool to render Jinja2 template with data from status data file. It is used by tests to provide final summary without need to write per-test tool which would do it - Jinja2 template seems to provide enough logick for what is needed in the output. Each test have its own template - e.g. see core/opl/status_data_report.txt.
• status_data_diff.py - Convenience tool to show differences in two status data files.

These tools are only used by tests:

• script-skip-to-end.py - Script to skip to the end of Kafka queues for given consumer group (it should be trivial given python-kafka docs, but simplest way did not worked for me)
• script-manage-db.py - Helper script to manipulate with utility DB. Consumes config with list of tables relevant for given test and SQL needed to create them. E.g. see extras/opl/sample-tables.yaml

Installation

You can install the full library or just the lightweight core tools.

Full Installation:

Installs all tools, including those with heavy dependencies.

python -m venv venv
source venv/bin/activate
python -m pip install git+https://github.com/redhat-performance/opl.git

Core Installation:

Installs only the lightweight core tools from core/opl/ to minimize dependency footprint.

python -m venv venv
source venv/bin/activate
python3 -m pip install --no-cache-dir -e "git+https://github.com/redhat-performance/opl.git#egg=opl-rhcloud-perf-team-core&subdirectory=core"

Extras Installation:

Note: Do you really need to do this? If you do full installation, it does exactly this.

If you choose to install only the extras package manually, it is your responsibility to also install the core package, as extras modules depend on core functionality (because setting that dependency for this package is not straight forward):

python3 -m pip install "git+https://github.com/redhat-performance/opl.git#egg=opl-rhcloud-perf-team-core&subdirectory=core"
python3 -m pip install "git+https://github.com/redhat-performance/opl.git#egg=opl-rhcloud-perf-team-extras&subdirectory=extras"

Local Development:

If you have cloned the git repository and want to develop locally, install the full package in editable mode:

python -m pip install -e .[dev]

Running unit tests

source venv/bin/activate
python -m pytest tests/

This will also generate a test coverage report. The terminal will show a summary, and a detailed HTML report can be found in htmlcov/index.html.

Notes

• Our Jinja2 by default loads templates from the folder where the script generic.py is located.
  • If no template with a given name is found, the second option is used: the /home/ folder. That's more useful for temporary changes. For example, if I have a template in iperf in /home/compliance/mycustomtemplate.j2, I'll omit the /home to use this: compliance/mycustomtemplate.j2.