The build file is actually project/SparkBuild.scala. You can add your Hadoop 
version there.

Running sbt assembly/assembly is apparently a lot faster than assembly, try 
that.

Matei

On Dec 11, 2013, at 2:28 PM, Nan Zhu <[email protected]> wrote:

> Hi, all
> 
> maybe it's a stupid question
> 
> I want to compile spark with a customized hdfs jar, what should I do? (it 
> seems that I didn't find build.sbt)
> 
> the other question is, is there any method to accelerate the running of sbt 
> assembly?
> 
> Thank you very much
> 
> Best,
> 
> Nan

Reply via email to