An extraction of Capistrano's parallel SSH command execution, capiche?
Capissh executes commands and file transfers on remote servers in parallel.
Capissh will maintain open connections with servers it has seen when it is sensible to do so. When the batch size is not restricted, Capissh will maintain all connections, which greatly reduces connection overhead. Sets of commands are run in parallel on all servers within the batch.
The interface is intentionally simple. A Capissh::Configuration object will maintain the open sessions, which will be matched up with servers for each command invocation.
require 'capissh' servers = ['email@example.com', 'firstname.lastname@example.org'] capissh = Capissh.new capissh.run servers, 'date' capissh.run servers, 'uname -a' capissh.sudo servers, 'ls /root'
Huge thank you to Jamis Buck and all the other Capistrano contributors for creating and maintaining this gem. Without them, this library would not exist, and ruby deployment wouldn't be at the level it is today.
Most of this code is directly extracted from capistrano without modification.