Setup resources to start tracking sitespeed.io

Hi!

I'd like to request some resources to help the frontend team track some sitespeed.io runs

This would require 3 things

• A graphite time series database
• A sitespeed.io docker image (there's a custom one that I created docker pull dev.gitlab.org:5005/gitlab/gitlab-build-images:sitespeed-gitlab)
• Access to the grafana dashboards to create a couple of them using the graphite data source

So far I've made some experiments using kubernetes I have a full repo going on here

I have lots of files that I used to create experiments, but so far the ones that would be required in this scenario would be:

• To create the sitespeed cron job https://gitlab.com/jivanvl/sitespeed-gitlab-kubernetes/blob/master/jobs/sitespeed-cron.yml

• To create the graphite server (pods and service) https://gitlab.com/jivanvl/sitespeed-gitlab-kubernetes/blob/master/pods/sitespeed.yml https://gitlab.com/jivanvl/sitespeed-gitlab-kubernetes/blob/master/services/graphite.yml

As far the requirements go, I used the recommended 200GB+ persistent volume that Google recommends to prevent performance issues. Persistent volume claims would be better as we could scale depending on the runs

Sitespeed.io docker images recommend 2GB of RAM as a minimum, kubernetes will provide that with the args that I provided

Also we would need to setup some secrets files for:

• Graphite logins
• Access to our own container registry

Other than this it works! I have an example of a graphite server here Let me know if you'd like access as it has credentials

Let me know if there's any other information that I missed and how I could help, we had this issue before: https://gitlab.com/gitlab-com/infrastructure/issues/2440

Thanks!

cc @ahanselka