thrift-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mario Emmenlauer <>
Subject Re: does not stop when asked to...
Date Wed, 19 Jul 2017 12:03:03 GMT

Maybe my question is more a feature request or bug. So I've opened
a ticket for the problem here (in case other people want to follow
this up later):

All the best,


On 13.07.2017 20:10, Mario Emmenlauer wrote:
> Ok, I found a bit more info myself, but still would need some help.
> Below more:
> On 13.07.2017 15:04, Mario Emmenlauer wrote:
>> I have a Java TSimpleServer runnable in a thread running. The
>> main thread eventually asks the server to stop(). But they seem
>> to ignore the request. I checked the code of
>> and I'm under the impression that the innermost loop the server
>> does not poll the variable stopped_ or does it?
>> [...] from line 76:
>>           while (true) {
>>             if (eventHandler_ != null) {
>>               eventHandler_.processContext(connectionContext, inputTransport, outputTransport);
>>             }
>>             if(!processor.process(inputProtocol, outputProtocol)) {
>>               break;
>>             }
>>           }
> I found that this loop is only interrupted if the client disconnects.
> I tried changing 'while(true)' for 'while(!stopped_)' but that did
> not help. I guess that one of the methods in the loop must be blocking.
> Furthermore I found
> 'Cannot shutdown TThreadedServer when clients are still connected'
> that seems to address exactly my issue! But it addresses only C++
> code, not the Java servers, is that correct?
> Should I open a new issue for this problem? Or am I on the wrong
> track here?
>> Do I understand correctly that the TSimpleServer will only check
>> stopped_ when there is no client connected? Personally I would
>> prefer if the client can not "force" my server to continue running.
>> If required, clients should be forcibly disconnected on stop().
>> Is there a way to achieve that? I found no methods "forceStop()" or
>> "disconnectClients()" or something like it. Am I even on the right
>> track?

View raw message