Re: In Intellij, maven failed to build Catalyst project

2017-02-20 Thread Armin Braun
I think the reason you're seeing this (and it then disappearing in Sean's case) is likely that there was a change in another that required a recompile of a module dependency. Maven doesn't do this automatically by default. So it eventually goes away when you do a full build either with Maven or SBT

Re: In Intellij, maven failed to build Catalyst project

2017-02-20 Thread Sean Owen
I saw this too yesterday but not today. It may have been fixed by some recent commits. On Mon, Feb 20, 2017 at 6:52 PM ron8hu wrote: I am using Intellij IDEA 15.0.6. I used to use Maven to compile Spark project Catalyst inside Intellij without problem. A couple of days ago, I fetched latest S

In Intellij, maven failed to build Catalyst project

2017-02-20 Thread ron8hu
I am using Intellij IDEA 15.0.6. I used to use Maven to compile Spark project Catalyst inside Intellij without problem. A couple of days ago, I fetched latest Spark code from its master repository. There was a change in CreateJacksonParser.scala. So I used Maven to compile Catalyst project a

Re: Will .count() always trigger an evaluation of each row?

2017-02-20 Thread Ryan Blue
I think it is a great idea to have a way to force execution to build a cached dataset. The use case for this that we see the most is to build broadcast tables. Right now, there's a 5-minute timeout to build a broadcast table. That's plenty of time if the data is sitting in a table, but we see a lo

Re: Output Committers for S3

2017-02-20 Thread Ryan Blue
We just wrote a couple new committers for S3 that we're beginning to roll out to our Spark users. I've uploaded a repo with it if you'd like to take a look: https://github.com/rdblue/s3committer The main problem with the UUID approach is that data is live as soon as the S3 upload completes. Tha

Output Committers for S3

2017-02-20 Thread Matthew Schauer
I'm using Spark 1.5.2 and trying to append a data frame to partitioned Parquet directory in S3. It is known that the default `ParquetOutputCommitter` performs poorly in S3 because move is implemented as copy/delete, but the `DirectParquetOutputCommitter` is not safe to use for append operations in

Re: Re:compile about the code

2017-02-20 Thread ??????????
hi zhai, l redid it by your idea, it is ok now. thanks. ---Original--- From: "StanZhai" Date: 2017/2/20 20:40:24 To: "dev"; Subject: Re:compile about the code Your antlr4-maven-plugin looks like incomplete, you can try to delete ~/.m2 in your home directory, then re-compile spark. ---

Re:compile about the code

2017-02-20 Thread StanZhai
Your antlr4-maven-plugin looks like incomplete, you can try to delete ~/.m2 in your home directory, then re-compile spark. -- Original -- From: " 萝卜丝炒饭 [via Apache Spark Developers List]";; Date: Feb 20, 2017 To: "Stan Zhai"; Subject: compile about the co

compile about the code

2017-02-20 Thread ??????????
hi all, when i compile spark2.0.2, i meet an error about the antlr4. i paste the info in the attach file, wpuld you like help me pls? 0220_2.png Description: Binary data - To unsubscribe e-mail: dev-unsubscr...@spark.ap