enriquegh / sauce-log-parser

Sauce Log Parser

Small python script that analyzes the log.json file provided by a Sauce Labs test.

It gets the time commands took to run and the time in between Sauce Labs sent a response and received another request. With those two metrics it gets the average, min, max and total of them.

This can help determine if a "slow" test is because of Selenium commands/server issue or if there is latency in the connection

NOTE: There is a discrepancy between the total time posted on the Sauce Labs test details page and the total time of between_commands and duration.

This is because the log.json file (which is used by this script) only takes into account test time but does not take into account start up and processing time.

python sel_log_parser.py -h
usage: sel_log_parser.py [-h] [-a ADMIN] [-k ACCESS_KEY] [-u USER] [-s]
                         [-r REGION] [-v] [--csv] [--jobid JOBID [JOBID ...]]
                         [-b BUILDID]

optional arguments:
  -h, --help            show this help message and exit
  -a ADMIN, --admin ADMIN
                        Sauce Admin username. For Saucers only.
  -k ACCESS_KEY, --access_key ACCESS_KEY
                        Sauce Admin access key. For Saucers only.
  -u USER, --user USER  Sauce username. Account Username of the Test Owner
                        that ran the session.
  -s, --save            Save the output as a .log file in cwd. Schema is
                        log_session-id.log.
  -r REGION, --region REGION
                        Sauce region where test was performed(us-west-1, us-
                        east-1, eu-central-1, apac-southeast-1)
  -v, --verbose         Verbose flag to print at debug level
  --csv                 Save the output of all tests as csv in cwd. Schema is
                        date_job-ids.csv
  --jobid JOBID [JOBID ...]
                        Sauce Labs Session ID to be examined.
  -b BUILDID, --buildid BUILDID
                        Sauce Labs Build ID to be examined.

Usage

Run against one session!

python sel_log_parser.py -a $SAUCE_USERNAME -k $SAUCE_ACCESS_KEY -u some-username --some-session-id It would look like this: python sel_log_parser.py -a $SAUCE_USERNAME -k $SAUCE_ACCESS_KEY -u awesome.userName --jobid 6e709ff746664fy896b467675b886e40 or if you are using Python3 python3 sel_log_parser.py -a $SAUCE_USERNAME -k $SAUCE_ACCESS_KEY -u awesome.userName --jobid 6e709ff746664fy896b467675b886e40

Run against multiple sessions!

python sel_log_parser.py -a $SAUCE_USERNAME -k $SAUCE_ACCESS_KEY -u some-username some-session-id another-session-id third-session-id ... Nth-session

**NOTE: You only need to provide --jobid once when running analysis for multiple sessions in one command:

python sel_log_parser.py -a $SAUCE_USERNAME -k $SAUCE_ACCESS_KEY -u awesome.userName --jobid 6e709ff746664fy896b467675b886e40 a3332d53yead44899962354drr789747

Output:

---
test_id: your-session-id
duration:
  mean: 0.33958974618178145
  max: 480.00422978401184
  min: 0.0022919178009033203
  total: 551.8333375453949
between_commands:
  mean: 0.1021887963922153
  max: 180.5098659992218
  min: -477.2646918296814
  total: 165.95460534095764

Setup

Manual Setup via git master branch:

1. cd to a directory where you want the python tools/files.
2. git clone [e-mail]:enriquegh/sauce-log-parser.git
3. cd sauce-log-parser
4. pip install -r requirements.txt or pip3 install -r requirements.txt
5. Start analyzing tests!

Contributing and Testing

Fork the remote master branch and make pull requests. Open issues if something is wrong and copy-paste your input & output.

Launch tests with py.test. Pytest is used and the /tests directory should get you started on seeing what is covered and what tests should look like.