I can take a look at it later today, np.

-david

On Oct 6, 2011, at 3:51 PM, Andrei Savu wrote:

> I know about that and I feel bad that we failed at reviewing and committing
> the patch.
> https://issues.apache.org/jira/browse/WHIRR-243
> 
> Adrian, David I know this is annoying but do you think we can refresh that
> patch for the trunk?
> 
> -- Andrei Savu
> 
> On Thu, Oct 6, 2011 at 8:20 PM, Adrian Cole <[email protected]> wrote:
> 
>> Hi, Andrei.
>> 
>> Once upon a time we had a stub computeservice jira.  I bet this could
>> unveil
>> any potential concurrency problems.
>> 
>> -A
>> 
>> On Thu, Oct 6, 2011 at 9:34 AM, Andrei Savu <[email protected]> wrote:
>> 
>>> Hey,
>>> 
>>> Today I believe that we've found a critical concurrency bug (me, Adrian
>> and
>>> Paul Baclace).
>>> 
>>> I think it is possible to overlap the install & configure phases at the
>>> server side because we are running the install phase in an async fashion
>> as
>>> an user script. We assume that the install script was already executed
>> when
>>> jclouds reports that all machines started.
>>> 
>>> I believe we can solve this by having some sort of synchronization
>>> mechanism
>>> on the server side.
>>> 
>>> Can someone else check this? Am I missing something?
>>> 
>>> -- Andrei Savu
>>> 
>> 

Reply via email to