You could always create a custom (different) message for deploy:web:disable. Or you could attempt to hide the problem by displaying something more like "Please wait..." and setting an auto-refresh on that page -- that way, users won't have to do anything special.
You could reduce it by reducing the time your app takes to start. The only way I know of to eliminate it is to do a rolling deploy (think "rolling blackouts") -- to restart one Mongrel at a time, or one server at a time, and configure Apache to failover to the others. This might be dangerous, though, depending on what changes you've made -- and I don't know of any recipes to help with that. Might be safe if you haven't made any schema changes -- maybe. A similar, safer solution might be to stop them all at once, do your migrations, then start one Mongrel at a time. Especially on a single-core machine, this should reduce the time needed to get at least one Mongrel running. But again, it depends on Apache being able to failover quickly, and it depends on you writing custom recipes -- and this one only reduces the time users would see that message, it doesn't eliminate it. On Thu, Jul 17, 2008 at 2:11 PM, chris johnson <[EMAIL PROTECTED]> wrote: > > Hi all. > > When we deploy our application using cap2's "cap deploy" command, > while the mongrel cluster restarts our site says "Service temporarily > unavailable". I know this is because apache can't talk to the > mongrels, but is there a way to avoid this message altogether (I > suppose "web:disable" works but that is something we typically reserve > for long deploys)? > > Thanks for your time. > -Chris > > > --~--~---------~--~----~------------~-------~--~----~ To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/capistrano -~----------~----~----~----~------~----~------~--~---
