Open Stack

Hi Eoghan,

> Can you explain the logic behind the deliberate shutdown of the
> cluster? Is it just intended for testing purposes? If so, what
> are the args generally passed to the topology in the non-test
> (i.e. long running) scenario?

Yes, it is just intended for testing. You can run strom topology in 
local mode[1] when you don't give topologyname arguments. Or 
run it on a cluster[2] with topologyname argument as below. 

[
    storm jar synaps-storm/target/synaps-storm-2012.09.b3.jar  com.spcs.synaps.PutMetricTopology topologyname
]

Thank you,
June Yi

[1] https://github.com/nathanmarz/storm/wiki/Local-mode
[2] https://github.com/nathanmarz/storm/wiki/Running-topologies-on-a-production-cluster



------- Original Message -------
Sender : Eoghan Glynn<eglynn at redhat.com> 
Date   : 2012-11-06 07:23 (GMT+09:00)
Title  : Re: [Synaps] a more forgiving interpretation of INSUFFICIENT_DATA?



> Expect to be nagged about any issues I encounter as I get all the
> components up & running ;)

So this is taking me a bit longer than expected to get up and running.
I have a quick question about the storm topology that you may be able
to help me with.

Basically I see the storm cluster shut down after about 3 mins, in a
quiescent scenario (i.e. with no metrics flowing). Presumably I'm
tripping over:

  https://github.com/spcs/synaps/blob/master/synaps-storm/src/com/spcs/synaps/PutMetricTopology.java#L170

Can you explain the logic behind the deliberate shutdown of the
cluster? Is it just intended for testing purposes? If so, what
are the args generally passed to the topology in the non-test
(i.e. long running) scenario?

[Currently I'm sparking up the storm topology with:

   storm jar synaps-storm/target/synaps-storm-2012.09.b3.jar  com.spcs.synaps.PutMetricTopology
]

Cheers,
Eoghan