[chef] Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Force service start at end of recipe.

["Jason J. W. Williams" < >]

It seems like the optimal solution would be a recipe command that
would instruct chef-client to restart itself at the end of the recipe
and then pick up where it left off.

-J

On Tue, May 17, 2011 at 4:04 PM, Jason J. W. Williams
<
 >
 wrote:
> Running 'exec "chef-client"' inside the ruby_block gets further,
> except what the original chef-client process was doing with apt
> doesn't seem to release /var/lib/dpkg/lock causing the first recipe
> that hits an debian package in the new process to fail.
>
> -J
>
> On Tue, May 17, 2011 at 3:53 PM, Jason J. W. Williams
> <
 >
>  wrote:
>>> Yeah, I expect one of those - what happens to the child that forks? Does
>>> it complete the run?
>>
>> Doesn't look like it:
>>
>> IOError: Cannot open or read
>> /var/cache/cookbooks/python-essentials/attributes.rb!
>> FATAL: Stacktrace dumped to /var/cache/chef/chef-stacktrace.out
>>
>