Status update for Kallithea based workflow PEPs
As of earlier today, I've arranged to spend around 1 day of week of paid time on CPython infrastructure work, focusing on container based developer workflow improvements for the web services we run or are considering running. My proposals for Kallithea based workflow changes are going to be the starting point for that work, so I've updated both PEP 462 and 474 accordingly: https://hg.python.org/peps/rev/12d44a2d7c59 As you can see, this also means I've returned to this list, as contributing to productive discussions here is now part of my job, rather than a purely volunteer activity. Regards, Nick. -- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia
On Mon Feb 09 2015 at 6:15:22 AM Nick Coghlan
As of earlier today, I've arranged to spend around 1 day of week of paid time on CPython infrastructure work, focusing on container based developer workflow improvements for the web services we run or are considering running.
Fantastic! And what exactly are you aiming for with "container based developer workflow improvements for the web services we run or are considering running"? If you're taking ideas then I vote for stuff like getting Buildbots up using containers so that we can download the same containers as developers and run them under, e.g. Valgrant to do local testing on the OS the container is targeting? IOW less volunteers and more structured, high uptime stuff for OS coverage? -Brett
My proposals for Kallithea based workflow changes are going to be the starting point for that work, so I've updated both PEP 462 and 474 accordingly: https://hg.python.org/peps/rev/12d44a2d7c59
As you can see, this also means I've returned to this list, as contributing to productive discussions here is now part of my job, rather than a purely volunteer activity.
Regards, Nick.
-- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia _______________________________________________ core-workflow mailing list core-workflow@python.org https://mail.python.org/mailman/listinfo/core-workflow This list is governed by the PSF Code of Conduct: https://www.python.org/psf/codeofconduct
On 10 Feb 2015 01:34, "Brett Cannon"
On Mon Feb 09 2015 at 6:15:22 AM Nick Coghlan
wrote: As of earlier today, I've arranged to spend around 1 day of week of paid time on CPython infrastructure work, focusing on container based developer workflow improvements for the web services we run or are considering running.
Fantastic!
And what exactly are you aiming for with "container based developer workflow improvements for the web services we run or are considering running"? If you're taking ideas then I vote for stuff like
getting Buildbots up using containers so that we can download the same containers as developers and run them under, e.g. Valgrant to do local testing on the OS the container is targeting? IOW less volunteers and more structured, high uptime stuff for OS coverage? Yep, that's the general idea. I'd mostly been thinking in terms of Kallithea's development workflow, but you're right, we could also apply it to enabling cloud based testing in the buildbot fleet. Cheers, Nick.
-Brett
My proposals for Kallithea based workflow changes are going to be the starting point for that work, so I've updated both PEP 462 and 474 accordingly: https://hg.python.org/peps/rev/12d44a2d7c59
As you can see, this also means I've returned to this list, as contributing to productive discussions here is now part of my job, rather than a purely volunteer activity.
Regards, Nick.
-- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia _______________________________________________ core-workflow mailing list core-workflow@python.org https://mail.python.org/mailman/listinfo/core-workflow This list is governed by the PSF Code of Conduct:
On Mon, 09 Feb 2015 15:34:02 +0000
Brett Cannon
If you're taking ideas then I vote for stuff like getting Buildbots up using containers so that we can download the same containers as developers and run them under, e.g. Valgrant to do local testing on the OS the container is targeting?
As long as the tested platform can run in a container... (or perhaps you meant a VM?). Regards Antoine.
On Mon Feb 09 2015 at 3:30:32 PM Antoine Pitrou
On Mon, 09 Feb 2015 15:34:02 +0000 Brett Cannon
wrote: If you're taking ideas then I vote for stuff like getting Buildbots up using containers so that we can download the same containers as developers and run them under, e.g. Valgrant to do local testing on the OS the container is targeting?
As long as the tested platform can run in a container... (or perhaps you meant a VM?).
I did mean container but VMs also work. I realize this won't work for all OSs, but it does work for enough to warrant looking into since containers have better workflows for local testing when something does fail. -Brett
Regards
Antoine.
_______________________________________________ core-workflow mailing list core-workflow@python.org https://mail.python.org/mailman/listinfo/core-workflow This list is governed by the PSF Code of Conduct: https://www.python.org/psf/codeofconduct
On 10 Feb 2015 06:39, "Brett Cannon"
On Mon Feb 09 2015 at 3:30:32 PM Antoine Pitrou
wrote:
On Mon, 09 Feb 2015 15:34:02 +0000 Brett Cannon
wrote: If you're taking ideas then I vote for stuff like getting Buildbots up using containers so that we can download the same containers as developers and run them under, e.g. Valgrant to do local testing on the OS the container is targeting?
As long as the tested platform can run in a container... (or perhaps you meant a VM?).
I did mean container but VMs also work. I realize this won't work for all OSs, but it does work for enough to warrant looking into since containers have better workflows for local testing when something does fail.
The main overhead is actually maintaining your stable of base images, and the infrastructure for doing that with containers independently of DockerHub is currently still fairly immature. At this point in history, we'd probably be better off hooking up the BuildBot master's VM worker support up to the PSF's Rackspace account - our throughput is low enough that the VM vs container overhead shouldn't matter. There may still be value in enabling local container based testing for CPython itself, though. At the very least, between Vagrant and Docker we should be able to automate *local* testing across various POSIX systems without affecting the host system (beyond installing Vagrant and potentially VirtualBox), and Windows folks may even be able to figure out how to automate that side of things as well (at least with an active MSDN subscription). Cheers, Nick.
-Brett
Regards
Antoine.
_______________________________________________ core-workflow mailing list core-workflow@python.org https://mail.python.org/mailman/listinfo/core-workflow This list is governed by the PSF Code of Conduct:
https://www.python.org/psf/codeofconduct
_______________________________________________ core-workflow mailing list core-workflow@python.org https://mail.python.org/mailman/listinfo/core-workflow This list is governed by the PSF Code of Conduct:
participants (3)
-
Antoine Pitrou
-
Brett Cannon
-
Nick Coghlan