Right now I'm using Hudson to manage this. I have a devel, staging and production environment (all separate apps on heroku) Google "heroku multiple environments" for more information.
I then have hudson monitor different branches, so when I push to my master branch (devel) it runs its tests and, assuming all is good push to heroku. When we're ready for a release I merge to staging and push. Hudson pulls, tests, then deploys to Heroku staging. Run some manual checks to make sure all is good. Assuming it is, merge to production and push, Hudson pulls, tests and pushes to Production if all's good. I have my testing/staging envs using the same DB at the moment, might split them all into their own DB at some point. All of this is supported with Heroku's current features. On Jan 11, 9:22 am, Alex <[email protected]> wrote: > Hi everyone, This is just a thought but it would be great to get some > feedback, see how useful this could be. > > Is it feasible to allow individual dynos to be given different code to > run? > Read below for more information: It would awesome if we could supply a > couple of our dynos with new code and monitor them for problems (while > they run in production, with our normal database). > > From what I've seen, we have a lot of different ways to set up > different working environments, staging apps, testing apps etc. but > these all end up with separate apps using separate databases. We can't > get our users to actually use the staging app for instance. > > Continuous deployment > -http://www.startuplessonslearned.com/2009/06/why-continuous-deploymen... > - is a simple idea. > > Code is checked in to the master branch, or whatever the release > branch is. That code gets tested and pushed to a small chunk of your > server instances (this assumes multiple servers, or in heroku's case, > multiple dynos). > Those instances are monitored, using whatever monitoring system you > want and if no problems occur, the new code is pushed out even more. > More monitoring. Code is then pushed to the whole system. > If at any point something goes wrong, the new code is rolled back and > a developer has to take a look before you can deploy again. > > This isn't something I'd expect Heroku to produce, but with some > support tools, maybe we can manage it for ourselves?
-- You received this message because you are subscribed to the Google Groups "Heroku" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/heroku?hl=en.
