:( I just realized this is impossible because of the read-only file
systems, which hurts my heart, heh.

I suppose I'll just find out if the return time is anything to worry
about before deciding to go with another solution.

I think it should be noted, though, that not *everyone* wants to use
an RDBMS. Is there any support for running your own data storage
mechanisms in the future? Or better support for background processes?

On Mar 4, 12:53 pm, BigLove <[email protected]> wrote:
> Hello all!
>
> I've got just a couple of questions before I can really harness the
> power of the awesomeness that is Heroku (as you can tell, I'm
> thoroughly impressed!).
>
> Perhttp://heroku.com/how/dyno_grid#1
>
> I'm wondering how I might be able to kick off a new dyno? And what
> exactly goes into making this happen? Will I get a chance to run any
> startup scripts for when this happens?
>
> Which brings me to my next question.
>
> I'm working on a site right now where we're using TokyoTyrant, a
> small, lightweight, and really fast database. We're doing this because
> it's so damn easy to setup replication for your app. You have a
> master, point a slave to the master, done. What's so great about our
> setup is that we very rarely write to the database. As in very rarely,
> I'm talking about once every few months, which makes our replication
> needs even simpler.
>
> All of which makes me really happy when I look at Heroku. If I can
> startup a local Tyrant slave on each launch of my ruby application on
> Heroku, with the slave pointing at my hosted Tyrant master, then that
> means that I can provide a local database for each and every one of my
> application instances, while still easily harnessing the cloud
> computing structure you've provided.
>
> At the moment, I plan on creating a rake task that kicks off a tyrant
> server, unless someone here more intelligent/experienced than I could
> provide a better way to do this on Heroku.
>
> Ok. Sorry for the barrage of questions :P
>
> How do dynos work? Does each dyno have it's own ports that are
> accessible? If I kick off two dynos, each of them pointing their
> database configs at the localhost for their Tyrant configuration and
> port 45006, will they run into each other? Or is the environment
> sectioned off enough to provide safeguards against this?
>
> Also, for each dyno, what does "localhost" point to?
>
> Is what I'm asking possible? Wise?
>
> I'm wanting a tyrant per application instance to avoid the return time
> required for a request to an external Tyrant master. So having a
> tyrant slave per application instance is pretty important to me.
>
> If that's not possible, then I'd really like to have all my
> application instances be pointed to "localhost", and have a tyrant
> slave database available. How would something like that be setup for
> Heroku?
>
> Finally, the heroku command lists some commands for "bundles". Is
> there any documentation on this?
>
> Thanks!
--~--~---------~--~----~------------~-------~--~----~
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
-~----------~----~----~----~------~----~------~--~---

Reply via email to