Feature/num 1061 parallel test execution (pabot)

[wlad]

Introduces pabot as test dependency for parallel test execution

• parallelization happens on test case level and is limited to the amount of available CPU cores/threads
• example: on a machine with 24 CPU cores 24 tests a run simultaneously
• since CircleCI's default docker executors has 2 vCPU cores only I had to configure a higher resource_class (large) to take a notice of any parallelization effect (NOTE: resource_classes other than default require a paid plan on CircleCI)
• execution time saving depends highly on amount of CPU cores
• actual execution time saved using resource_class=large (4 x vCPU) is about 50% (~ 10 min vs. ~20 - 30 min), compare builds
• some tests fail randomly when using parallel execution. This issue is tracked here
• not sure if it's always the same tests that are impacted. We'll sync this branch with develop and observe behavior over time
• failing tests are tagged with not-ready_parallel_fail and marked as not-ready

[wlad]

maybe is better to remove not-ready tags from impacted tests , so we see them failing (when they fail)?

[wlad]

⛔ blocked by #247

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
No Duplication information

[wlad]

⛔ blocked by ehrbase/ehrbase#460

[wlad]

UPDATE:

synced with develop and ran tests locally. only three tests fail (one is marked as skip).

Locally impressive speed up with parallel execution via pabot (~3 m vs ~30 m on CI. OK that's without ehrbase + fhirbridge startup overhead, but even with startup overhead still would be less than 5 m)

The massive fail on CI may be caused by limited resources of Docker executer (and now probably also due to downgrade to CircleCI free plan)