Dear all in OpenShift community,
I've decided to offer help with development of OpenShift Origin v3 as I see this product as a future of all smart application development life cycle.
Now I have production OpenShift Origin consisting of 8 nodes + one broker and one development OpenShift origin for our devel team. Currently there is arount 150 apps deployed but this is first phase - we count with 1000 apps deployed till the end of Q3. I've worked as as system administrator (11 years) and system designer / architect in telco industry. Now I am helping with startup company deploying RoR application around the globe.
I've managed to get up and running whole infrastructure based on RedHat community products and this is my point of view - I got different approach:
* OpenShift is designed to work from point of view of developer mainly - we take it in different way - as a central point of large scale app deployment maintained by very small team. This brings questions like having central git repository (now there is no way how to tell OpenShift application to update itself from one central repository using initial git url - as OpenShift has different approach and is itself master git repository - this is little pain for us)
Much of this is different in OpenShift 3, in that there's more of a focus on external Git integration. We agree that it should be a primary use case to run the way you describe.
* Smart backuping - as we wrote our own cartridge for Bacula where every app is running Bacula file daemon - but I don't see this as a great solution as we want to have around 2k apps deployed around the globe in next year and this for me seems as a overhead. My architecture is "as cheap as possible" - I just lease servers and run OpenNebula on top of them connect them via Floodlight / OpenVSwitch - manage nodes via Puppet / Foreman combo and then have central gateways as entry points into infrastructure running modified version of openshift-nginx-routing. Nginxes are combined with Varnish case and I use GeoIP provided by Amazon for DNS resolving of central domain names. Now I experiment with CEPH for HA of individual VMs.
As I see OpenShift as "final" PaaS solution for many years I would like to contribute on v3 development. I've never did anything with golang but I am good at system design, RoR, Shell and others.
Are there currently any topics I can help with - even with single line of code.
There are definitely a lot of areas to get involved with, although many of our Trello cards are not as detailed as they could be. I'd recommend looking through the backlog lists of the public boards here:
If you see anything that looks interesting, please ask questions on Trello or on this list. Since you have familiarity with cartridges, it might be useful to get familiar with Source-To-Images and offer feedback about the experience building an image that exposes a runtime - we definitely could benefit from the perspective you would bring:
Finally, the GitHub issues list has some bugs, if you just want to get familiar with the source.
Thanks you all for "absolute power" which OpenShift gives me.
Thanks for being a part of the community - we're always glad to hear about how OpenShift has helped folks and how it can be better!