Tips on Writing Dockerfiles | Hacker News Reader

Tips on Writing Dockerfiles

techbar.me

54 pointstestmes9 comments

Could someone please register the domain www.isdockerproductionready.com and put a big NO(later Yes) on this so I know when to catch up?

I love the idea and concept behind it so far but I sadly don't have time to check the status and read the relevant blog articles about every cool new OS project i'd like to follow... especially if they are so potentially business critical.

Am I wrong in my prioritization or do others have similar feelings in this regard?

I think production ready is relative. People have different risk appetites.

I agree on this. If you need example, Spotify is running Docker in production: http://www.slideshare.net/rohanrsingh/docker-at-spotify

By the time something becomes popular its been around for awhile. As far as I'm concerned the time to start taking advantage or at least learning is well before everyone starts saying its "production ready". I agree that we can't learn every new awesome technology. But I think that people have the tendency to find reasons to excuse learning the new technologies, like saying its not production ready or something. Just because you don't have time to learn and carefully evaluate a new technology doesn't mean you need to dismiss it. Just say "looks like it might be great but I can't say for sure because I don't have time to learn how to use it and then evaluate it.". Its just impossible to try every new thing. You don't need an excuse.

If you wait until most people are saying its production ready, that's because most people are already using it in production. If you wait until that point to start learning it then you will be behind.

This is definitely correct, but if the developers themselves say explicitly on their blog that they don't consider the current builds production ready, i wouldn't feel comfortable recommending to build something important on it, as it is prone to changes, even if it succeeds.

Behind what?

Just keep checking for it to hit version 1.0, the universal bat-signal for production-ready.

Some things become de-facto production ready before that, but if you don't have time to follow it that closely, that's what 1.0 is for.

Right now the visionaries are adopting Docker, using it and working closely with us to iron out the kinks.

One of the ways we're working risk-averse businesses who have bought in to the idea, is to pick a pragmatic project that's low stakes and will ship to production sometime late Q1 or Q2.

This provides best of both worlds. Low risk project gets visibility inside of the company, and is a strawman for future implementation. Win/win for everyone.

Solomon (lead maintainer of Docker) has done a brilliant job reconciling all of these projects and community needs in to their essence and implementing just enough to satisfy a plethora of use cases.

I would be interested learning about peoples techniques for including database dumps into their dev environments. I've been tweaking a Dockerfile for the past few days and its pretty enjoyable to watch the screen go by as it's getting built. Ideally I'd like to replicate everything in terms of our production environment in docker containers so our devs and test changes, but right now I'm taking the small steps. Currently I have a few different bash scripts that are initiated outside the Dockerfile to handle stuff like MySQL, Apache, and Redis configs. But I'm open to a different approach since this isn't cached. I assume after building the base docker container I can just IMAGE it as the future starting point.

Reply on news.ycombinator.com