The following code in a capfile will result in the command being run in
["localhost","otherhost"]
role :db, "localhost"
task :role_test, :roles=>:db do
role :db, "otherhost"
run("ls")
end
I would like to be able to completely redefine the role programmatically
(instead of just adding servers to the role). This would be easy if we
changed the attr_reader to attr_accessor on line 14 of
capistrano/configuration/roles.rb (talking cap 2.1.0 here).
Or provide a Capistrano::Configuration::Roles::clear_role method
def clear_role(which)
roles[which]=[]
end
This is my solution X to my problem Y:
A small scenario:
There is an "application server package" and a "database package". We
can deploy these on 2 separate machines, so one machine fulfils the
application server role and the other the database role. Or, we can do
an all-in-one deployment where a single server fulfils both roles.
In this particular scenario it is not desirable to do the deployment on
both machines - we'd like to control which server subset is used for the
role through other means .
I can solve this problem by simply not calling role outside of the
tasks, but I would like a bit more control.
Would that be possible?
Cheers,
V.-
---
http://www.braveworld.net/riva
--~--~---------~--~----~------------~-------~--~----~
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/capistrano
-~----------~----~----~----~------~----~------~--~---
