…examples, and updates to follow API changes

I gave the Streaming code examples, both Scala and Java, a test run today. I turned up a number of small errors, mostly compile errors in the Java examples. There were a few typos in the Scala too.

I also took the liberty of adding things like imports, since in several cases they are not obvious. Feel free to push back on some changes.

There's one thing I haven't quite addressed in the changes. `JavaPairDStream` uses the Java API version of `Function2` in almost all cases, as `JFunction2`. However it uses `scala.Function2` in:

```
  def reduceByKeyAndWindow(reduceFunc: Function2[V, V, V], windowDuration: Duration)
  :JavaPairDStream[K, V] = {
    dstream.reduceByKeyAndWindow(reduceFunc, windowDuration)
  }
```

Is that a typo?

Also, in Scala, I could not get this to compile:
```
val windowedWordCounts = pairs.reduceByKeyAndWindow(_ + _, Seconds(30), Seconds(10))
error: missing parameter type for expanded function ((x$1, x$2) => x$1.$plus(x$2))
```

You can see my fix below but am I missing something?

Otherwise I can say these all worked for me!

Author: Sean Owen <[email protected]>

Closes #589 from srowen/SPARK-1663 and squashes the following commits:

65a906b [Sean Owen] Corrections for several compile errors in streaming code examples, and updates to follow API changes
(cherry picked from commit 11d5494)

Signed-off-by: Patrick Wendell <[email protected]>

SPARK-1076: Convert Int to Long to avoid overflow

Patch for PR apache#578.

Author: Xiangrui Meng <[email protected]>

Closes apache#589 and squashes the following commits:

98c435e [Xiangrui Meng] cast Int to Long to avoid Int overflow

…examples, and updates to follow API changes

I gave the Streaming code examples, both Scala and Java, a test run today. I turned up a number of small errors, mostly compile errors in the Java examples. There were a few typos in the Scala too.

I also took the liberty of adding things like imports, since in several cases they are not obvious. Feel free to push back on some changes.

There's one thing I haven't quite addressed in the changes. `JavaPairDStream` uses the Java API version of `Function2` in almost all cases, as `JFunction2`. However it uses `scala.Function2` in:

```
  def reduceByKeyAndWindow(reduceFunc: Function2[V, V, V], windowDuration: Duration)
  :JavaPairDStream[K, V] = {
    dstream.reduceByKeyAndWindow(reduceFunc, windowDuration)
  }
```

Is that a typo?

Also, in Scala, I could not get this to compile:
```
val windowedWordCounts = pairs.reduceByKeyAndWindow(_ + _, Seconds(30), Seconds(10))
error: missing parameter type for expanded function ((x$1, x$2) => x$1.$plus(x$2))
```

You can see my fix below but am I missing something?

Otherwise I can say these all worked for me!

Author: Sean Owen <[email protected]>

Closes apache#589 from srowen/SPARK-1663 and squashes the following commits:

65a906b [Sean Owen] Corrections for several compile errors in streaming code examples, and updates to follow API changes

SPARK-1076: Convert Int to Long to avoid overflow

Patch for PR apache#578.

Author: Xiangrui Meng <[email protected]>

Closes apache#589 and squashes the following commits:

98c435e [Xiangrui Meng] cast Int to Long to avoid Int overflow

1. Change job metadata from aarch64 to arm64
2. Remove useless nodeset "ubuntu-bionic-vxh-arm64"

Related-Bug: theopenlab/openlab#318