[tor-dev] Gitlab CI runners available for experimentation on gitlab.torproject.org

Hans-Christoph Steiner hans at guardianproject.info
Thu Jul 9 08:39:36 UTC 2020


Happy to help!  I'm a big fan of gitlab-ci since it is a collection of
standard tools like Docker, YAML, bash, etc. It takes a bit more to
learn than Travis-CI, but it pays off by being more flexible and a
simpler setup.  E.g. it is easy to start with a plain, base Debian
image, only install the requirements, and run tests from there.  No
intermediate layer. And using YAML templates, it is possible to reuse
chunks of code. In the go setup I'm using right now for snowflake, I
have a template for the install and test run.  Then its trivial to
add/change the base image between Debian-derivs of various releases.

I should also add: these are actually F-Droid runners, not Guardian
Project.  F-Droid a bare metal server (16-core/32-thread, 142GB RAM, 3TB
disk) that could be allocated to only gitlab runners that can be shared
with Tor Project cost-free.  We just need someone to admin it.  We have
an almost complete setup with ansible.  @uniqx and I would happily help
someone finish that setup if there was someone to make sure it stays
updated and running.  (I'm personally already admining more servers than
I should be).

Also, these runners have KVM and privileged mode enabled, so you can run
any KVM VM in the gitlab-ci jobs.  Docker too.

.hc

Alexander Færøy:
> Hello folks!
> 
> Hans from The Guardian Project added his CI runners to our Gitlab
> instance. It looks like some pretty fast machines that allows each team
> to experiment with Gitlab CI on our Gitlab instance.
> 
> Hans says that the runners have no uptime promises or anything like
> that, so if they are down they are down :-)
> 
> Here's some documentation for getting started:
> https://docs.gitlab.com/ee/ci/
> 
> Thanks to Hans for this!
> 
> All the best,
> Alex.
> 

-- 
PGP fingerprint: EE66 20C7 136B 0D2C 456C  0A4D E9E2 8DEA 00AA 5556
https://pgp.mit.edu/pks/lookup?op=vindex&search=0xE9E28DEA00AA5556


More information about the tor-dev mailing list