No worries, it's really my fault for the abysmal lack of
documentation. Currently, the source code is the documentation,
meaning that if you want to know what options are available, you have
to go spelunking through the code. :(
Regarding a wait time for max_hosts, do you mean, a delay before the
next set of hosts is used? There is currently no delay between runs,
but that might make a good patch. The code in question is in
capistrano/configuration/connections.rb, just search for :max_hosts.
- Jamis
On Jun 10, 2008, at 1:24 PM, JasonSmithy wrote:
>
> Thank you Jamis!
>
> Please forgive my ignorance, but how can I know what other options
> like :max_hosts are available? And is there a way to configure the
> time that max_hosts waits? I have not had any luck finding it in the
> source yet.
>
>
> Thanks again!
>
>
> On Jun 9, 9:16 pm, Jamis Buck <[EMAIL PROTECTED]> wrote:
>> Actually, capistrano has that functionality baked in, via
>> the :max_hosts
>> option. You could define your role with a block, so it gets evaluated
>> late (from a temp file, for example), and then just set
>> the :max_hosts
>> option on the task in question. E.g.:
>>
>> role :dynamic_servers do
>> IO.readlines(server_list).map { |line| line.chomp }
>> end
>>
>> desc "Test cycling through servers"
>> task :test_each, :roles => :dynamic_servers, :max_hosts => 5 do
>> run "uptime"
>> end
>>
>> Then, when capistrano is asked to execute the "uptime" command on all
>> servers in the dynamic_servers role, it will first evaluate the block
>> for the dynamic_servers role, and will then see that it is only to
>> use 5
>> hosts at a time.
>>
>> - Jamis
>>
>> JasonSmithy wrote:
>>> Hi,
>>
>>> Is there anyway to reset a role?
>>
>>> What I want to do is read a list of IPs dynamically from a temporary
>>> file, then run a task on a portion of the list at a time, not have
>>> the
>>> task run on all the IP's directly in parallel. What I am trying to
>>> do
>>> so far is to get the list, run the task on the first IP in the list,
>>> sleep for 5 seconds and then run the task again on the next IP in
>>> the
>>> list. Ideally, this would take 5 IPs at a time until all IPs are
>>> executed, but I need to learn to crawl before I can walk ;)
>>
>>> task :test_cycle do
>>> transaction do
>>> set :j, 4
>>> set :k, 0
>>> $servers = IO.readlines(server_list) # Get the list
>>> of servers from
>>> temp file.
>>> while k < j
>>> # This is a method to increment by one.
>>> Capistrano
>>> # complains that "+" is an unknown
>>> method without this.
>>> def add(value)
>>> value = value + 1
>>> return value
>>> end
>>
>>> $ip = $servers[k].chomp
>>
>>> role :servers_each, $ip
>>> # Set the
>>> server_each role to the new IP
>>
>>> test_each # Run
>>> the task
>>> set :k, add(k) # Increment k
>>> by 1.
>>> role :servers_each, servers_each.clear
>>> end
>>> end
>>> end
>>
>>> desc "Test cycling through servers"
>>> task :test_each, :roles => :servers_each do
>>> run "uptime"
>>> puts "Sleeping for 5 seconds"
>>> sleep 5
>>> end
>>
>>> This works except for each time the loop is run, the next IP gets
>>> added to the server_each role. This has the effect of re-running the
>>> command on each server each time. So the first time it runs from 1
>>> server. The next time, it runs on 2 servers, and so on.
>>
>>> How can I set server_each back to empty after running test_each
>>> and is
>>> this even the a good way to do this?
>>
>>> Thanks in advance for you help!
> >
--~--~---------~--~----~------------~-------~--~----~
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/capistrano
-~----------~----~----~----~------~----~------~--~---
