I have a Spark job written in Python that works well, but when porting it to Scala, ran into a memory pressure issue. I have a flatMap operation in my code
def make_combinations(x:(String, Iterable[String])) = {
x._2
.toSeq
.combinations(2)
.map({case Seq(a,b) => if (a<b) List(a,b,1) else List(b,a,1)})
.map(_.mkString(","))
}
In Python:
def make_combinations(x):
for a,b in itertools.combinations(x[1], 2):
yield ','.join(map(str, [a,b,1])) if a < b else ','.join(map(str, [b,a,1]))
I can do this exact function in Python using generators and it works with no problem. With Scala, however, I get memory issues since it can generate more data than the machine can hold in memory, but any thoughts why Spark can't spread the data around the nodes like it seems to in Python. Am I messing up laziness here in Scala? Thanks
EDIT:
I'm using spark 1.6 on ec2 with spark standalone. The iterable in question can be up to a hundred million in length.
Related
I'm reading "Learning spark", and noticed this kind of code:
val result = input.map(x => x * x)
result.persist(StorageLevel.DISK_ONLY)
println(result.count())
println(result.collect().mkString(","))
does this code really persist the result rdd? I thought in Spark everything was immutable, but in this case it looks like we are mutating the result rdd.
Shouldn't this piece of code be written like this? :
val result = input.map(x => x * x)
val persistedResult = result.persist(StorageLevel.DISK_ONLY)
println(persistedResult.count())
println(persistedResult.collect().mkString(","))
There are many more code samples like this in the book, so that got me wondering...
Unlike typed transformations, persist() is applied to this dataset. This is because persist actually only marks dataset as such. From Spark Programming Guide:
You can mark an RDD to be persisted using the persist() or cache()
methods on it. The first time it is computed in an action, it will be
kept in memory on the nodes.
I'm trying (unsuccessfully) to run a simple hello world type program in Apache Flink. The code takes a message from Apache Kafka adds a "." after each letter and prints the new string to stdout. The code correctly gets the message from Kafka, but the map function to add the "." fails. I've tried the function at the REPL prompt and the scala code works correctly there.
Scala code:
scala> input = "hello"
input: String = hello
scala> val output = input.flatMap(value => value + ".")
output: String = h.e.l.l.o
flink program:
flink code
cut off line reads
val messageStream = env.addSource(new FlinkKafkaConsumer09("CL", new SimpleStringSchema, properties))
I cant figure out where I'm going wrong, I've tried the apache documentation to no avail. any help you could give me would be well received.
First of all I would recommend studying some basics about functional programming and operations like map/flatMap/reduce etc.
All of those mentioned functions apply to collections. In your scala example as #pedrofuria pointed out you apply the flatMap function to String which is collection of chars
In the flink example messageStream can be abstracted as a collection of strings, so to perform the operation you described you ought to do sth like:
val stream = messageStream.map(str => str.mkString("."))
I used the mkString instead of flatMap from your example because the former rather than
h.e.l.l.o (as you wrote)
it produces
h.e.l.l.o.
But once more really start with basics of functional programming.
thanks for the help, i figured out the problem, sort of.
Although the flatMap function works at the scala prompt, it doesn't work in Flink proper, as Flink requires FlatMap be passed a new FlatMapFunction with an override. Still unsure why it works at the scala prompt in flink, but the code now compiles and runs as expected.
Because flatMap in scala prompt is not the same with flatMap in your flink program.
flatMap in your scala prompt is just a function in scala.
The flink flatMap can be applied while input likes below:
val input = benv.fromElements(
"To be, or not to be,--that is the question:--",
"Whether 'tis nobler in the mind to suffer",
"The slings and arrows of outrageous fortune",
"Or to take arms against a sea of troubles,")
val counts = input
.flatMap { _.toLowerCase.split("\\W+") }
.map { (_, 1) }.groupBy(0).sum(1)
See: scala-shell
I prefer Python over Scala. But, as Spark is natively written in Scala, I was expecting my code to run faster in the Scala than the Python version for obvious reasons.
With that assumption, I thought to learn & write the Scala version of some very common preprocessing code for some 1 GB of data. Data is picked from the SpringLeaf competition on Kaggle. Just to give an overview of the data (it contains 1936 dimensions and 145232 rows). Data is composed of various types e.g. int, float, string, boolean. I am using 6 cores out of 8 for Spark processing; that's why I used minPartitions=6 so that every core has something to process.
Scala Code
val input = sc.textFile("train.csv", minPartitions=6)
val input2 = input.mapPartitionsWithIndex { (idx, iter) =>
if (idx == 0) iter.drop(1) else iter }
val delim1 = "\001"
def separateCols(line: String): Array[String] = {
val line2 = line.replaceAll("true", "1")
val line3 = line2.replaceAll("false", "0")
val vals: Array[String] = line3.split(",")
for((x,i) <- vals.view.zipWithIndex) {
vals(i) = "VAR_%04d".format(i) + delim1 + x
}
vals
}
val input3 = input2.flatMap(separateCols)
def toKeyVal(line: String): (String, String) = {
val vals = line.split(delim1)
(vals(0), vals(1))
}
val input4 = input3.map(toKeyVal)
def valsConcat(val1: String, val2: String): String = {
val1 + "," + val2
}
val input5 = input4.reduceByKey(valsConcat)
input5.saveAsTextFile("output")
Python Code
input = sc.textFile('train.csv', minPartitions=6)
DELIM_1 = '\001'
def drop_first_line(index, itr):
if index == 0:
return iter(list(itr)[1:])
else:
return itr
input2 = input.mapPartitionsWithIndex(drop_first_line)
def separate_cols(line):
line = line.replace('true', '1').replace('false', '0')
vals = line.split(',')
vals2 = ['VAR_%04d%s%s' %(e, DELIM_1, val.strip('\"'))
for e, val in enumerate(vals)]
return vals2
input3 = input2.flatMap(separate_cols)
def to_key_val(kv):
key, val = kv.split(DELIM_1)
return (key, val)
input4 = input3.map(to_key_val)
def vals_concat(v1, v2):
return v1 + ',' + v2
input5 = input4.reduceByKey(vals_concat)
input5.saveAsTextFile('output')
Scala Performance
Stage 0 (38 mins), Stage 1 (18 sec)
Python Performance
Stage 0 (11 mins), Stage 1 (7 sec)
Both produces different DAG visualization graphs (due to which both pictures show different stage 0 functions for Scala (map) and Python (reduceByKey))
But, essentially both code tries to transform data into (dimension_id, string of list of values) RDD and save to disk. The output will be used to compute various statistics for each dimension.
Performance wise, Scala code for this real data like this seems to run 4 times slower than the Python version.
Good news for me is that it gave me good motivation to stay with Python. Bad news is I didn't quite understand why?
The original answer discussing the code can be found below.
First of all, you have to distinguish between different types of API, each with its own performance considerations.
RDD API
(pure Python structures with JVM based orchestration)
This is the component which will be most affected by the performance of the Python code and the details of PySpark implementation. While Python performance is rather unlikely to be a problem, there at least few factors you have to consider:
Overhead of JVM communication. Practically all data that comes to and from Python executor has to be passed through a socket and a JVM worker. While this is a relatively efficient local communication it is still not free.
Process-based executors (Python) versus thread based (single JVM multiple threads) executors (Scala). Each Python executor runs in its own process. As a side effect, it provides stronger isolation than its JVM counterpart and some control over executor lifecycle but potentially significantly higher memory usage:
interpreter memory footprint
footprint of the loaded libraries
less efficient broadcasting (each process requires its own copy of a broadcast)
Performance of Python code itself. Generally speaking Scala is faster than Python but it will vary on task to task. Moreover you have multiple options including JITs like Numba, C extensions (Cython) or specialized libraries like Theano. Finally, if you don't use ML / MLlib (or simply NumPy stack), consider using PyPy as an alternative interpreter. See SPARK-3094.
PySpark configuration provides the spark.python.worker.reuse option which can be used to choose between forking Python process for each task and reusing existing process. The latter option seems to be useful to avoid expensive garbage collection (it is more an impression than a result of systematic tests), while the former one (default) is optimal for in case of expensive broadcasts and imports.
Reference counting, used as the first line garbage collection method in CPython, works pretty well with typical Spark workloads (stream-like processing, no reference cycles) and reduces the risk of long GC pauses.
MLlib
(mixed Python and JVM execution)
Basic considerations are pretty much the same as before with a few additional issues. While basic structures used with MLlib are plain Python RDD objects, all algorithms are executed directly using Scala.
It means an additional cost of converting Python objects to Scala objects and the other way around, increased memory usage and some additional limitations we'll cover later.
As of now (Spark 2.x), the RDD-based API is in a maintenance mode and is scheduled to be removed in Spark 3.0.
DataFrame API and Spark ML
(JVM execution with Python code limited to the driver)
These are probably the best choice for standard data processing tasks. Since Python code is mostly limited to high-level logical operations on the driver, there should be no performance difference between Python and Scala.
A single exception is usage of row-wise Python UDFs which are significantly less efficient than their Scala equivalents. While there is some chance for improvements (there has been substantial development in Spark 2.0.0), the biggest limitation is full roundtrip between internal representation (JVM) and Python interpreter. If possible, you should favor a composition of built-in expressions (example. Python UDF behavior has been improved in Spark 2.0.0, but it is still suboptimal compared to native execution.
This may improved in the future has improved significantly with introduction of the vectorized UDFs (SPARK-21190 and further extensions), which uses Arrow Streaming for efficient data exchange with zero-copy deserialization. For most applications their secondary overheads can be just ignored.
Also be sure to avoid unnecessary passing data between DataFrames and RDDs. This requires expensive serialization and deserialization, not to mention data transfer to and from Python interpreter.
It is worth noting that Py4J calls have pretty high latency. This includes simple calls like:
from pyspark.sql.functions import col
col("foo")
Usually, it shouldn't matter (overhead is constant and doesn't depend on the amount of data) but in the case of soft real-time applications, you may consider caching/reusing Java wrappers.
GraphX and Spark DataSets
As for now (Spark 1.6 2.1) neither one provides PySpark API so you can say that PySpark is infinitely worse than Scala.
GraphX
In practice, GraphX development stopped almost completely and the project is currently in the maintenance mode with related JIRA tickets closed as won't fix. GraphFrames library provides an alternative graph processing library with Python bindings.
Dataset
Subjectively speaking there is not much place for statically typed Datasets in Python and even if there was the current Scala implementation is too simplistic and doesn't provide the same performance benefits as DataFrame.
Streaming
From what I've seen so far, I would strongly recommend using Scala over Python. It may change in the future if PySpark gets support for structured streams but right now Scala API seems to be much more robust, comprehensive and efficient. My experience is quite limited.
Structured streaming in Spark 2.x seem to reduce the gap between languages but for now it is still in its early days. Nevertheless, RDD based API is already referenced as "legacy streaming" in the Databricks Documentation (date of access 2017-03-03)) so it reasonable to expect further unification efforts.
Non-performance considerations
Feature parity
Not all Spark features are exposed through PySpark API. Be sure to check if the parts you need are already implemented and try to understand possible limitations.
It is particularly important when you use MLlib and similar mixed contexts (see Calling Java/Scala function from a task). To be fair some parts of the PySpark API, like mllib.linalg, provides a more comprehensive set of methods than Scala.
API design
The PySpark API closely reflects its Scala counterpart and as such is not exactly Pythonic. It means that it is pretty easy to map between languages but at the same time, Python code can be significantly harder to understand.
Complex architecture
PySpark data flow is relatively complex compared to pure JVM execution. It is much harder to reason about PySpark programs or debug. Moreover at least basic understanding of Scala and JVM in general is pretty much a must have.
Spark 2.x and beyond
Ongoing shift towards Dataset API, with frozen RDD API brings both opportunities and challenges for Python users. While high level parts of the API are much easier to expose in Python, the more advanced features are pretty much impossible to be used directly.
Moreover native Python functions continue to be second class citizen in the SQL world. Hopefully this will improve in the future with Apache Arrow serialization (current efforts target data collection but UDF serde is a long term goal).
For projects strongly depending on the Python codebase, pure Python alternatives (like Dask or Ray) could be an interesting alternative.
It doesn't have to be one vs. the other
The Spark DataFrame (SQL, Dataset) API provides an elegant way to integrate Scala/Java code in PySpark application. You can use DataFrames to expose data to a native JVM code and read back the results. I've explained some options somewhere else and you can find a working example of Python-Scala roundtrip in How to use a Scala class inside Pyspark.
It can be further augmented by introducing User Defined Types (see How to define schema for custom type in Spark SQL?).
What is wrong with code provided in the question
(Disclaimer: Pythonista point of view. Most likely I've missed some Scala tricks)
First of all, there is one part in your code which doesn't make sense at all. If you already have (key, value) pairs created using zipWithIndex or enumerate what is the point in creating string just to split it right afterwards? flatMap doesn't work recursively so you can simply yield tuples and skip following map whatsoever.
Another part I find problematic is reduceByKey. Generally speaking, reduceByKey is useful if applying aggregate function can reduce the amount of data that has to be shuffled. Since you simply concatenate strings there is nothing to gain here. Ignoring low-level stuff, like the number of references, the amount of data you have to transfer is exactly the same as for groupByKey.
Normally I wouldn't dwell on that, but as far as I can tell it is a bottleneck in your Scala code. Joining strings on JVM is a rather expensive operation (see for example: Is string concatenation in scala as costly as it is in Java?). It means that something like this _.reduceByKey((v1: String, v2: String) => v1 + ',' + v2) which is equivalent to input4.reduceByKey(valsConcat) in your code is not a good idea.
If you want to avoid groupByKey you can try to use aggregateByKey with StringBuilder. Something similar to this should do the trick:
rdd.aggregateByKey(new StringBuilder)(
(acc, e) => {
if(!acc.isEmpty) acc.append(",").append(e)
else acc.append(e)
},
(acc1, acc2) => {
if(acc1.isEmpty | acc2.isEmpty) acc1.addString(acc2)
else acc1.append(",").addString(acc2)
}
)
but I doubt it is worth all the fuss.
Keeping the above in mind, I've rewritten your code as follows:
Scala:
val input = sc.textFile("train.csv", 6).mapPartitionsWithIndex{
(idx, iter) => if (idx == 0) iter.drop(1) else iter
}
val pairs = input.flatMap(line => line.split(",").zipWithIndex.map{
case ("true", i) => (i, "1")
case ("false", i) => (i, "0")
case p => p.swap
})
val result = pairs.groupByKey.map{
case (k, vals) => {
val valsString = vals.mkString(",")
s"$k,$valsString"
}
}
result.saveAsTextFile("scalaout")
Python:
def drop_first_line(index, itr):
if index == 0:
return iter(list(itr)[1:])
else:
return itr
def separate_cols(line):
line = line.replace('true', '1').replace('false', '0')
vals = line.split(',')
for (i, x) in enumerate(vals):
yield (i, x)
input = (sc
.textFile('train.csv', minPartitions=6)
.mapPartitionsWithIndex(drop_first_line))
pairs = input.flatMap(separate_cols)
result = (pairs
.groupByKey()
.map(lambda kv: "{0},{1}".format(kv[0], ",".join(kv[1]))))
result.saveAsTextFile("pythonout")
Results
In local[6] mode (Intel(R) Xeon(R) CPU E3-1245 V2 # 3.40GHz) with 4GB memory per executor it takes (n = 3):
Scala - mean: 250.00s, stdev: 12.49
Python - mean: 246.66s, stdev: 1.15
I am pretty sure that most of that time is spent on shuffling, serializing, deserializing and other secondary tasks. Just for fun, here's naive single-threaded code in Python that performs the same task on this machine in less than a minute:
def go():
with open("train.csv") as fr:
lines = [
line.replace('true', '1').replace('false', '0').split(",")
for line in fr]
return zip(*lines[1:])
I want to use graphx. For now I just launchs it locally.
I've got NullPointerException in these few lines. First println works well, and second one fails.
..........
val graph: Graph[Int, Int] = Graph(users, relationships)
println("graph.inDegrees = " + graph.inDegrees.count) // this line works well
graph.mapVertices((id, v) => {
println("graph.inDegrees = " + graph.inDegrees.count) // but this one fails
42 // doesn't mean anything
}).vertices.collect
And it does not matter which method of 'graph' object I call. But 'graph' is not null inside 'mapVertices'.
Exception failure in TID 2 on host localhost:
java.lang.NullPointerException
org.apache.spark.graphx.impl.GraphImpl.mapReduceTriplets(GraphImpl.scala:168)
org.apache.spark.graphx.GraphOps.degreesRDD(GraphOps.scala:72)
org.apache.spark.graphx.GraphOps.inDegrees$lzycompute(GraphOps.scala:49)
org.apache.spark.graphx.GraphOps.inDegrees(GraphOps.scala:48)
ololo.MyOwnObject$$anonfun$main$1.apply$mcIJI$sp(Twitter.scala:42)
Reproduced using GraphX 2.10 on Spark 1.0.2. I'll give you a workaround and then explain what I think is happening. This works for me:
val c = graph.inDegrees.count
graph.mapVertices((id, v) => {
println("graph.inDegrees = " + c)
}).vertices.collect
In general, Spark gets prickly when you try to access an entire RDD or other distributed object (like a Graph) in code that's intended to execute in parallel on a single partition, like the function you're passing into mapVertices. But it's also usually a bad idea even when you can get it to work. (As a separate matter, as you've seen, when it doesn't work it tends to result in really unhelpful behavior.)
The vertices of a Graph are represented as an RDD, and the function you pass into mapVertices runs locally in the appropriate partitions, where it is given access to local vertex data: id and v. You really don't want the entire graph to be copied to each partition. In this case you just need to broadcast a scalar to each partition, so pulling it out solved the problem and the broadcast is really cheap.
There are tricks in the Spark APIs for accessing more complex objects in such a situation, but if you use them carelessly they will destroy your performance because they'll tend to introduce lots of communication. Often people are tempted to use them because they don't understand the computation model, rather than because they really need to, although that does happen too.
Spark does not support nested RDDs or user-defined functions that refer to other RDDs, hence the NullPointerException; see this thread on the spark-users mailing list. In this case, you're attempting to call count() on a Graph (which performs an action on a Spark RDD) from inside of a mapVertices() transformation, leading to a NullPointerException when mapVertices() attempts to access data structures that are only callable by the Spark driver.
In a nutshell, only the Spark driver can launch new Spark jobs; you can't call actions on RDDs from inside of other RDD actions.
See https://stackoverflow.com/a/23793399/590203 for another example of this issue.
So I've been working with parallel collections in Scala for a graph project I'm working on, I've got the basics of the graph class defined, it is currently using a scala.collection.mutable.HashMap where the key is Int and the value is ListBuffer[Int] (adjacency list). (EDIT: This has since been change to ArrayBuffer[Int]
I had done a similar thing a few months ago in C++, with a std::vector<int, std::vector<int> >.
What I'm trying to do now is run a metric between all pairs of vertices in the graph, so in C++ I did something like this:
// myVec = std::vector<int> of vertices
for (std::vector<int>::iterator iter = myVec.begin(); iter != myVec.end(); ++iter) {
for (std::vector<int>::iterator iter2 = myVec.begin();
iter2 != myVec.end(); ++iter2) {
/* Run algorithm between *iter and *iter2 */
}
}
I did the same thing in Scala, parallelized, (or tried to) by doing this:
// vertexList is a List[Int] (NOW CHANGED TO Array[Int] - see below)
vertexList.par.foreach(u =>
vertexList.foreach(v =>
/* Run algorithm between u and v */
)
)
The C++ version is clearly single-threaded, the Scala version has .par so it's using parallel collections and is multi-threaded on 8 cores (same machine). However, the C++ version processed 305,570 pairs in a span of roughly 3 days, whereas the Scala version so far has only processed 23,573 in 17 hours.
Assuming I did my math correctly, the single-threaded C++ version is roughly 3x faster than the Scala version. Is Scala really that much slower than C++, or am I completely mis-using Scala (I only recently started - I'm about 300 pages into Programming in Scala)?
Thanks!
-kstruct
EDIT To use a while loop, do I do something like..
// Where vertexList is an Array[Int]
vertexList.par.foreach(u =>
while (i <- 0 until vertexList.length) {
/* Run algorithm between u and vertexList(i) */
}
}
If you guys mean use a while loop for the entire thing, is there an equivalent of .par.foreach for whiles?
EDIT2 Wait a second, that code isn't even right - my bad. How would I parallelize this using while loops? If I have some var i that keeps track of the iteration, then wouldn't all threads be sharing that i?
From your comments, I see that your updating a shared mutable HashMap at the end of each algorithm run. And if you're randomizing your walks, a shared Random is also a contention point.
I recommend two changes:
Use .map and .flatMap to return an immutable collection instead of modifying a shared collection.
Use a ThreadLocalRandom (from either Akka or Java 7) to reduce contention on the random number generator
Check the rest of your algorithm for further possible contention points.
You may try running the inner loop in parallel, too. But without knowing your algorithm, it's hard to know if that will help or hurt. Fortunately, running all combinations of parallel and sequential collections is very simple; just switch out pVertexList and vertexList in the code below.
Something like this:
val pVertexList = vertexList.par
val allResult = for {
u <- pVertexList
v <- pVertexList
} yield {
/* Run algorithm between u and v */
((u -> v) -> result)
}
The value allResult will be a ParVector[((Int, Int), Int)]. You may call .toMap on it to convert that into a Map.
Why mutable? I don't think there's a good parallel mutable map on Scala 2.9.x -- particularly because just such a data structure was added to the upcoming Scala 2.10.
On the other hand... you have a List[Int]? Don't use that, use a Vector[Int]. Also, are you sure you aren't wasting time elsewhere, doing the conversions from your mutable maps and buffers into immutable lists? Scala data structures are different than C++'s so you might well be incurring in complexity problems elsewhere in the code.
Finally, I think dave might be onto something when he asks about contention. If you have contention, parallelism might well make things slower. How faster/slower does it run if you do not make it parallel? If making it not parallel makes it faster, then you most likely do have contention issues.
I'm not completely sure about it, but I think foreach loops in foreach loops are rather slow, because lots of objects get created. See: http://scala-programming-language.1934581.n4.nabble.com/for-loop-vs-while-loop-performance-td1935856.html
Try rewriting it using a while loop.
Also Lists are only efficient for head access, Arrays are probably faster.