Retrying a pipeline

[jwor...@sportsrisq.com]

I have a client pipeline which needs to post a request to an external calculation server, then retrieve the results. The calculation takes some time, so the external server processes it in a task queue. The client pipeline isn't exactly sure when the results will be ready; instead it needs to poll the server (get_status(id)), then if status==Completed retrieve the results (get_results(id))

Problem is that occasionally the server calculation fails, in which case I need the client pipeline to retry. I have a pipeline pattern based on this post:

http://stackoverflow.com/questions/11365172/google-appengine-pipelines-api

which looks as follows:

class ReducePipeline(pipeline.Pipeline):

    def run(self, *args):

        results=[result for result in list(args)

                 if result]

        if results==[]:

            return None

        return results[0]

 

class CallbackPipeline(pipeline.Pipeline):

    def run(self, id):

        status=get_status(id) # call external server status

        results, future_results = None, None

        if status==Error:

            pass

        elif status==Completed:

            results=get_results(id) # get results from external server

        elif status!=Completed:

            with pipeline.InOrder():

                yield Delay(seconds=CallbackWait) # NB

                future_results=yield CallbackPipeline(id)

        yield ReducePipeline(results, future_results)

class StartPipeline(pipeline.Pipeline):

    def run(self, request):

        id=start(request) # post request to external server; get job id in return

        yield CallbackPipeline(id)

    """

    is this really the best way to retry the pipeline ? dev_appserver.py results don't look promising :-(

    """

    def finalized(self):

        if not self.outputs.default:

            raise pipeline.Retry()

but this doesn't seem to work on dev_appserver.py, simply causing repeated errors at the StartPipeline finalisation stage. I was hoping to get the entire StartPipeline to retry instead. 

Can anyone advise me on a sensible pattern for retrying the StartPipeline on receipt of None results ?

Thanks.

[Arie Ozarov]

You should follow the advice given in your SO thread http://stackoverflow.com/questions/25883652/retrying-a-python-app-engine-pipeline to either

1) Use asynchronous pipeline instead of polling (would be simple to do if the calculation server can call you back upon completion)

or

2) raise the Retry from a run (I think that could be done by ReducePipeline).

Arie.

--
You received this message because you are subscribed to the Google Groups "Google App Engine Pipeline API" group.
To unsubscribe from this group and stop receiving emails from it, send an email to app-engine-pipeli...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.