On Tue, Sep 14, 2010 at 6:00 PM, Luke Kanies <[email protected]> wrote:
> On Sep 14, 2010, at 4:32 PM, Nigel Kersten wrote:
>
>> On Tue, Sep 14, 2010 at 4:26 PM, Jesse Wolfe <[email protected]> wrote:
>>>
>>>>
>>>> I propose that instead of checking the run_mode in the indirector, we
>>>> combine these settings into one setting and have the location be
>>>> dependent on run_mode in defaults.rb.  I can't think of any reason why
>>>> this wouldn't work and be less confusing.  Anyone else?
>>>>
>>>
>>> Sounds good to me.
>>> We'd probably still need to support clientyamldir as a synonym for backward
>>> compatibility for a while.
>>
>> FWIW we're interested in this because we're seeing issues with
>> puppet-load (that issues multiple simultaneous requests for the same
>> client) causing yaml corruption server-side, which means it's not so
>> great for benchmarking.
>
> This sounds more like a file locking issue or parallelization problem than 
> something that would be affected here, right?


Yes, but at the same time if you want to measure load, you want to do
it closer to real world, which involves multiple different clients
with different facts and even environments.

Simultaneous requests of the same client doesn't really exercise the
system accurately.

I hit the same problem as the OP when hacking this up.

-- 
nigel

-- 
You received this message because you are subscribed to the Google Groups 
"Puppet Developers" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/puppet-dev?hl=en.

Reply via email to