Hi Lasantha,
I have changed my storm topology. Now, all stream go to the same
siddhiManager instance and all queries works fine.
But now, I only have 1 instance of siddhiManager without distributed
processing ... How can I get scalability???
Thanks,
Andrés
-
--
Andres Gomez
Developer
Hi Andrés,
In your earlier query, it seems that you have a sequence with two
predicates running on different threads/nodes. In such a case, synching
between the two threads can be non-trivial AFAIK.
However, you can try to scale first for other usecases (other than
patterns, sequences).
Thanks for your help, Lasantha!
--
Andres Gomez
Developer at Eneo Tecnologia.
Website: http://redborder.net
LinkedIn: http://lnkd.in/sHnbJe
En 3 de octubre de 2014 en 10:25:06, Lasantha Fernando [via WSO2 Oxygen Tank]
(ml-node+s10903n104356...@n7.nabble.com) escrito:
Hi Andrés,
In your
Hi Lasantha,
I try the same set of queries without distributed processing.
Now this query work fine:
*- queryB:*'from streamFlow#window.timeBatch( 5 sec ) select max(bytes) as
maxBytes, min(bytes) as minBytes, avg(bytes) as avgBytes, sum(bytes) as
sumBytes, src group by src insert into
Hi Andres,
Does the same set of queries work without any issue if you use Siddhi
without distributed processing?
Also, can you explain how you are using Siddhi within the TridentTopology?
Based on how Siddhi is used, you might not need to enable distributed
processing for Siddhi instance.
Hii
I am using siddhi on distribute mode inside Apache Storm with Trident API. I
am trying two queries like this:
* - queryA:* 'from flow=streamFlow[bytes250]+,
monitor=streamMonitor[monitor==cpu and value25.00] within 1 select
flow[0].src as src, monitor.value as cpu insert into streamD'
* -