Private pip and st2pack builder

We are starting to move toward K8 and need assistance with the following.
GitHub - StackStorm/st2packs-dockerfiles: Source for st2packs-builder and st2packs-runner images looks pretty neat but how do we add our private pip repo (artifactory is what we use) to the build process?

Hi @badchad

I’m also using artifactory for custom pack image and This is what I’ve setup:

  1. Cloned , modified base image as per my requirements(Needed apt packages) and pushed all images to artifactory.
  2. Then cloned and changed all the FROM to my artifactory URL. Also did some changes to st2-image Dockerfile to as per the requirements.(This is basically used for st2-cutom po which is created when job runs.)
  3. In mentioned repository URL to artifactory.

I can help you out if you get some issues while following it.

I have build a new st2packs image while running a workflow it is showing a error at one task which I didnt get with previous image
“stderr”: “/opt/stackstorm/virtualenvs/kafka/bin/python: /lib/x86_64-linux-gnu/ version `GLIBC_2.25’ not found (required by /opt/stackstorm/virtualenvs/kafka/bin/python)\n”,

I finally ironed out how… it was easier then I thought. One thing I don’t get is how large a single pack is. I looked at the virtual env and it’s HUGE and full of things I think are probably not needed.
Thanks for the help BTW

Yeah that’s the concern, if you’ve more packages, image size would be huge… In my case it was around 1.5Gb and image build, helm upgrade used to take around 30 mins even for a small change.
I’ve now removed custom image and using PVC instead and installing packs directly from browser or inside containers.