I want to use file stream as input. And I look at SparkStreaming document again, it's saying file stream doesn't need a receiver at all.

So I'm wondering if I can control a specific DStream instance.

 


From: Evo Eftimov [mailto:evo.eftimov@isecc.com]
Sent: Monday, May 18, 2015 12:39 AM
To: 'Akhil Das'; Haopu Wang
Cc: 'user'
Subject: RE: [SparkStreaming] Is it possible to delay the start of some DStream in the application?

 

You can make ANY standard receiver sleep by implementing a custom Message Deserializer class with sleep method inside it.

 

From: Akhil Das [mailto:akhil@sigmoidanalytics.com]
Sent: Sunday, May 17, 2015 4:29 PM
To: Haopu Wang
Cc: user
Subject: Re: [SparkStreaming] Is it possible to delay the start of some DStream in the application?

 

Why not just trigger your batch job with that event?

 

If you really need streaming, then you can create a custom receiver and make the receiver sleep till the event has happened. That will obviously run your streaming pipelines without having any data to process.


Thanks

Best Regards

 

On Fri, May 15, 2015 at 4:39 AM, Haopu Wang <HWang@qilinsoft.com> wrote:

In my application, I want to start a DStream computation only after an
special event has happened (for example, I want to start the receiver
only after the reference data has been properly initialized).

My question is: it looks like the DStream will be started right after
the StreaminContext has been started. Is it possible to delay the start
of specific DStream?

Thank you very much!

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@spark.apache.org
For additional commands, e-mail: user-help@spark.apache.org