I replied on the thread. I believe the steps in the description can cause the 
problem as described.

-JZ

On Aug 3, 2013, at 6:06 AM, chao chu <[email protected]> wrote:

> By checking the code again, I thought the issue mentioned in the description 
> of CURATOR-3 should have already been fixed.
> 
> in the callback of the 'create' node operation in 'reset':
> 
>                 if ( event.getResultCode() == 
> KeeperException.Code.OK.intValue() )
>                 {
>                     setNode(event.getName());
>                     if ( state.get() == State.CLOSED )
>                     {
>                         setNode(null);
>                     }
>                     else
>                     {
>                         getChildren();
>                     }
>                 }
>                 else
>                 {
>                     log.error("getChildren() failed. rc = " + 
> event.getResultCode());
>                 }
> on creating the node successfully, in that callback, it will check the state 
> again and if it's already 'CLOSED', it will call setNode(null), thus the node 
> will be created. The earlier code doesn't have this check, I checked the 
> latest version (1.3.4) back to the netflix age, the code is like below:
> 
>                 if ( event.getResultCode() == 
> KeeperException.Code.OK.intValue() )
>                 {
>                     setNode(event.getName());
>                     getChildren();
>                 }
> 
> Could you please take a look at this and maybe you can close the issue if 
> it's the case. Thanks
> 
> 
> On Sat, Aug 3, 2013 at 1:23 AM, Jordan Zimmerman <[email protected]> 
> wrote:
> How do you plan to address this? It seems infinitely recursive ;)
> -JZ
> 
> 
> On Aug 2, 2013, at 7:42 AM, chao chu <[email protected]> wrote:
> 
>> Ok, thanks for all your replies. I will try to see if I can re-produce and 
>> fix the issue https://issues.apache.org/jira/browse/CURATOR-3 then.
>> 
>> Cheers,
>> 
>> 
>> On Fri, Aug 2, 2013 at 9:35 PM, Michael Morello <[email protected]> 
>> wrote:
>> Both have a high priority (at least in JIRA), let's go for 
>> https://issues.apache.org/jira/browse/CURATOR-45
>> 
>> 
>> 2013/8/1 Luciano Resende <[email protected]>
>> 
>> On Thu, Aug 1, 2013 at 12:44 PM, Michael Morello <[email protected]> 
>> wrote:
>> I volunteer to (try to) help on one of these two problems, Jordan, which one 
>> is best suited for a first contribution ?
>> 
>> 
>> Whichever you feel comfortable with. But one could assume the highest 
>> priority one would be the most complex.
>> 
>> 
>> 
>> 
>> 
>> -- 
>> ChuChao
> 
> 
> 
> 
> -- 
> ChuChao

Reply via email to