Setup GitHub Actions #182

Totktonada · 2020-12-21T02:09:49Z

Instead of Travis CI (which is going to be paid service) and AppVeyor (because GitHub Actions has Windows runners).

Before this patch, tarantool/tarantool-python CI was based on Travis CI (Linux runs) and Appveyor (Windows runs). This PR introduces GitHub Actions workflow files for both Linux and Windows test runs. Pipelines run for different supported tarantool, Python and msgpack package versions to ensure compatibility. tarantool instance is started in each Windows pipeline under WSL to run tests instead of using an external server (like in Appveyor). Since we start a new tarantool instance for each run, clean() procedure was removed. (The main reason to remove it was failing with "ER_DROP_FUNCTION: Can't drop function 1: function is SQL built-in" error on newer 2.x on bootstrap.) Testing pipeline for tarantool artifacts was also updated. Travis CI and Appveyor badges were replaced with GitHub Actions badge. Closes #182

Ping test started to fail (in ~70% cases) after Windows CI migration from Appveyor to GitHub Actions. As a part of this migration, the test Tarantool instance was changed from an instance started on a remote Linux server to an instance started on the same Windows server under WSL. Thus, request time has shortened and it supposedly caused the ping test to fail. Python documentation [1] declares that ...though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second... Particularly, StackOverflow answer [2] argues that For Linux and Mac precision is +- 1 microsecond or 0.001 milliseconds. Python on Windows uses +- 16 milliseconds precision due to clock implementation problems due to process interrupts. based on Windows documentation [3]. Assuming that ping requests between the same server services can easily be under 1 ms, it caused the test to fail. This patch adds skip for the flaky test, refer to #214 for further development. 1. https://docs.python.org/3/library/time.html#time.time 2. https://stackoverflow.com/a/1938096/11646599 3. https://docs.microsoft.com/en-us/windows-hardware/drivers/kernel/high-resolution-timers Follows up #182, part of #214

Ping test started to fail (in ~70% cases) after Windows CI migration from Appveyor to GitHub Actions. As a part of this migration, the test tarantool instance was changed from an instance started on a remote Linux server to an instance started on the same Windows server under WSL. Thus, request time has shortened and it supposedly caused the ping test to fail. Python documentation [1] declares that ...though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second... Particularly, StackOverflow answer [2] argues that For Linux and Mac precision is +- 1 microsecond or 0.001 milliseconds. Python on Windows uses +- 16 milliseconds precision due to clock implementation problems due to process interrupts. based on Windows documentation [3]. Assuming that ping requests between the same server services can easily be under 1 ms, it caused the test to fail. This patch adds skip for the flaky test, refer to #214 for further development. 1. https://docs.python.org/3/library/time.html#time.time 2. https://stackoverflow.com/a/1938096/11646599 3. https://docs.microsoft.com/en-us/windows-hardware/drivers/kernel/high-resolution-timers Follows up #182, part of #214

Before this patch, tarantool/tarantool-python CI was based on Travis CI (Linux runs) and Appveyor (Windows runs). This PR introduces GitHub Actions workflow files for both Linux and Windows test runs. Pipelines run for different supported tarantool, Python and msgpack package versions to ensure compatibility. tarantool instance is started in each Windows pipeline under WSL to run tests instead of using an external server (like in Appveyor). Since we start a new tarantool instance for each run, clean() procedure was removed. (The main reason to remove it was failing with "ER_DROP_FUNCTION: Can't drop function 1: function is SQL built-in" error on newer 2.x on bootstrap.) Testing pipeline for tarantool artifacts was also updated. Travis CI and Appveyor badges were replaced with GitHub Actions badge. Closes #182

Ping test started to fail (in ~70% cases) after Windows CI migration from Appveyor to GitHub Actions. As a part of this migration, the test tarantool instance was changed from an instance started on a remote Linux server to an instance started on the same Windows server under WSL. Thus, request time has shortened and it supposedly caused the ping test to fail. Python documentation [1] declares that ...though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second... Particularly, StackOverflow answer [2] argues that For Linux and Mac precision is +- 1 microsecond or 0.001 milliseconds. Python on Windows uses +- 16 milliseconds precision due to clock implementation problems due to process interrupts. based on Windows documentation [3]. Assuming that ping requests between the same server services can easily be under 1 ms, it caused the test to fail. This patch adds skip for the flaky test, refer to #214 for further development. 1. https://docs.python.org/3/library/time.html#time.time 2. https://stackoverflow.com/a/1938096/11646599 3. https://docs.microsoft.com/en-us/windows-hardware/drivers/kernel/high-resolution-timers Follows up #182, part of #214

Before this patch, tarantool/tarantool-python CI was based on Travis CI (Linux runs) and Appveyor (Windows runs). This PR introduces GitHub Actions workflow files for both Linux and Windows test runs. Pipelines run for different supported tarantool, Python and msgpack package versions to ensure compatibility. tarantool instance is started in each Windows pipeline under WSL to run tests instead of using an external server (like in Appveyor). Since we start a new tarantool instance for each run, clean() procedure was removed. (The main reason to remove it was failing with "ER_DROP_FUNCTION: Can't drop function 1: function is SQL built-in" error on newer 2.x on bootstrap.) Testing pipeline for tarantool artifacts was also updated. Travis CI and Appveyor badges were replaced with GitHub Actions badge. Closes #182

Ping test started to fail (in ~70% cases) after Windows CI migration from Appveyor to GitHub Actions. As a part of this migration, the test tarantool instance was changed from an instance started on a remote Linux server to an instance started on the same Windows server under WSL. Thus, request time has shortened and it supposedly caused the ping test to fail. Python documentation [1] declares that ...though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second... Particularly, StackOverflow answer [2] argues that For Linux and Mac precision is +- 1 microsecond or 0.001 milliseconds. Python on Windows uses +- 16 milliseconds precision due to clock implementation problems due to process interrupts. based on Windows documentation [3]. Assuming that ping requests between the same server services can easily be under 1 ms, it caused the test to fail. This patch adds skip for the flaky test, refer to #214 for further development. 1. https://docs.python.org/3/library/time.html#time.time 2. https://stackoverflow.com/a/1938096/11646599 3. https://docs.microsoft.com/en-us/windows-hardware/drivers/kernel/high-resolution-timers Follows up #182, part of #214

Before this patch, tarantool/tarantool-python CI was based on Travis CI (Linux runs) and Appveyor (Windows runs). This PR introduces GitHub Actions workflow files for both Linux and Windows test runs. Pipelines run for different supported tarantool, Python and msgpack package versions to ensure compatibility. tarantool instance is started in each Windows pipeline under WSL to run tests instead of using an external server (like in Appveyor). Since we start a new tarantool instance for each run, clean() procedure was removed. (The main reason to remove it was failing with "ER_DROP_FUNCTION: Can't drop function 1: function is SQL built-in" error on newer 2.x on bootstrap.) Testing pipeline for tarantool artifacts was also updated. Travis CI and Appveyor badges were replaced with GitHub Actions badge. Closes #182

Ping test started to fail (in ~70% cases) after Windows CI migration from Appveyor to GitHub Actions. As a part of this migration, the test tarantool instance was changed from an instance started on a remote Linux server to an instance started on the same Windows server under WSL. Thus, request time has shortened and it supposedly caused the ping test to fail. Python documentation [1] declares that ...though the time is always returned as a floating point number, not all systems provide time with a better precision than 1 second... Particularly, StackOverflow answer [2] argues that For Linux and Mac precision is +- 1 microsecond or 0.001 milliseconds. Python on Windows uses +- 16 milliseconds precision due to clock implementation problems due to process interrupts. based on Windows documentation [3]. Assuming that ping requests between the same server services can easily be under 1 ms, it caused the test to fail. This patch adds skip for the flaky test, refer to #214 for further development. 1. https://docs.python.org/3/library/time.html#time.time 2. https://stackoverflow.com/a/1938096/11646599 3. https://docs.microsoft.com/en-us/windows-hardware/drivers/kernel/high-resolution-timers Follows up #182, part of #214

SQL_BUILTIN was introduced in [1] (tarantool 2.2.1) and dropped in [2] (tarantool 2.10.0-beta2). It is not permitted to drop SQL_BUILTIN functions, thus, before this patch, tarantool_python_ci.lua was unable to run with versions between 2.2.1 and 2.10.0-beta2, failing with `ER_DROP_FUNCTION: Can't drop function 1: function is SQL built-in`. This patch fixes this. tarantool_python_ci.lua is used to start a tarantool instance for Windows tests. 1. tarantool/tarantool@200a492 2. tarantool/tarantool@c49eab9 Part of #182