If I have a very large DataFrame on my PySpark cluster, does calling df.count() on it cause the entire DataFrame df to be brought into memory of a single node, or do all the individual nodes count their part of the structure and return it somewhere to be aggregated as a final result?
I don't see anything in the documentation to indicate this one way or the other. Basically I don't want to call count() on a DataFrame that's too big to fit in the memory of any individual node.
count is something that can be distributed across executors. So, for each executor, count their number of records. Then send the aggregated number of records to be counted together. Spark optimizations will take care of those simple details.
If you call collect() then, that's what causes driver to be flooded with complete dataframe and most likely resulting in failure.
The best practice on the spark is not to usee count and it's recommended to use isEmpty method instead of count method if it's possible. Also, all of the spark actions except collect method will run on the spark executors and the only result will return to the spark driver
Related
Currently, I am working on a single node Hadoop and I wrote a job to output a sorted dataframe with only one partition to one single csv file. And I discovered several outcomes when using repartition differently.
At first, I used orderBy to sort the data and then used repartition to output a CSV file, but the output was sorted in chunks instead of in an overall manner.
Then, I tried to discard repartition function, but the output was only a part of the records. I realized without using repartition spark will output 200 CSV files instead of 1, even though I am working on a one partition dataframe.
Thus, what I did next were placing repartition(1), repartition(1, "column of partition"), repartition(20) function before orderBy. Yet output remained the same with 200 CSV files.
So I used the coalesce(1) function before orderBy, and the problem was fixed.
I do not understand why working on a single partitioned dataframe has to use repartition and coalesce, and how the aforesaid processes affect the output. Grateful if someone can elaborate a little.
Spark has relevant parameters here:
spark.sql.shuffle.partitions and spark.default.parallelism.
When you perform operations like sort in your case, it triggers something called a shuffle operation
https://spark.apache.org/docs/latest/rdd-programming-guide.html#shuffle-operations
That will split your dataframe to spark.sql.shuffle.partitions partitions.
I also struggled with the same problem as you do and did not find any elegant solution.
Spark generally doesn’t have a great concept of ordered data, because all your data is split accross multiple partitions. And every time you call an operation that requires a shuffle your ordering will be changed.
For this reason, you’re better off only sorting your data in spark for the operations that really need it.
Forcing your data into a single file will break when the dataset gets larger
As Miroslav points out your data gets shuffled between partitions every time you trigger what’s called a shuffle stage (this is things like grouping or join or window operations)
You can set the number of shuffle partitions in the spark Config - the default is 200
Calling repartition before a group by operation is kind of pointless because spark needs to reparation your data again to execute the groupby
Coalesce operations sometimes get pushed into the shuffle stage by spark. So maybe that’s why it worked. Either that or because you called it after the groupby operation
A good way to understand what’s going on with your query is to start using the spark UI - it’s normally available at http://localhost:4040
More info here https://spark.apache.org/docs/3.0.0-preview/web-ui.html
I’m facing a particularly bizarre issue while firing filter queries on a spark dataframe. Here's a screenshot of the filter command I'm trying to run:
As you can see, I'm trying to run the same command multiple times. Each time, it's giving a different number of rows. It is actually meant to return 6 records, but it ends up showing a random number of records every time.
FYI, The underlying data source (from which I'm creating the dataframe) is an Avro file in a Hadoop data lake.
This query only gives me consistent results if I cache the dataframe. But this is not always possible for me because the dataframe might be very huge and hence would choke up memory resources if I cache it.
What might be the possible reasons for this random behavior? Any advice on how to fix it?
Many thanks :)
I have a large parquet dataset that I am reading with Spark. Once read, I filter for a subset of rows which are used in a number of functions that apply different transformations:
The following is similar but not exact logic to what I'm trying to accomplish:
df = spark.read.parquet(file)
special_rows = df.filter(col('special') > 0)
# Thinking about adding the following line
special_rows.cache()
def f1(df):
new_df_1 = df.withColumn('foo', lit(0))
return new_df_1
def f2(df):
new_df_2 = df.withColumn('foo', lit(1))
return new_df_2
new_df_1 = f1(special_rows)
new_df_2 = f2(special_rows)
output_df = new_df_1.union(new_df_2)
output_df.write.parquet(location)
Because a number of functions might be using this filtered subset of rows, I'd like to cache or persist it in order to potentially speed up execution speed / memory consumption. I understand that in the above example, there is no action called until my final write to parquet.
My questions is, do I need to insert some sort of call to count(), for example, in order to trigger the caching, or if Spark during that final write to parquet call will be able to see that this dataframe is being used in f1 and f2 and will cache the dataframe itself.
If yes, is this an idiomatic approach? Does this mean in production and large scale Spark jobs that rely on caching, random operations that force an action on the dataframe pre-emptively are frequently used, such as a call to count?
there is no action called until my final write to parquet.
and
Spark during that final write to parquet call will be able to see that this dataframe is being used in f1 and f2 and will cache the dataframe itself.
are correct. If you do output_df.explain(), you will see the query plan, which will show that what you said is correct.
Thus, there is no need to do special_rows.cache(). Generally, cache is only necessary if you intend to reuse the dataframe after forcing Spark to calculate something, e.g. after write or show. If you see yourself intentionally calling count(), you're probably doing something wrong.
You might want to repartition after running special_rows = df.filter(col('special') > 0). There can be a large number of empty partitions after running a filtering operation, as explained here.
The new_df_1 will make cache special_rows which will be reused by new_df_2 here new_df_1.union(new_df_2). That's not necessarily a performance optimization. Caching is expensive. I've seen caching slow down a lot of computations, even when it's being used in a textbook manner (i.e. caching a DataFrame that gets reused several times downstream).
Counting does not necessarily make sure the data is cached. Counts avoid scanning rows whenever possible. They'll use the Parquet metadata when they can, which means they don't cache all the data like you might expect.
You can also "cache" data by writing it to disk. Something like this:
df.filter(col('special') > 0).repartition(500).write.parquet("some_path")
special_rows = spark.read.parquet("some_path")
To summarize, yes, the DataFrame will be cached in this example, but it's not necessarily going to make your computation run any faster. It might be better to have no cache or to "cache" by writing data to disk.
I am curious to know when i need to persist my dataframe in spark and when not. Cases:-
If i need data from file ( Do i need to persist it? if i apply repetitive count like:-
val df=spark.read.json("file://root/Download/file.json")
df.count
df.count
Do i need to persist df?? because according to me it should store df in memory after first count and use same df in second count. Record in file is 4 , Because when i practically check it , it read file again and again, So why spark doesn't store it in memory
Second question is in spark read is an action or transformation?
DataFrames by design are immutable, so every transformation done on them would create a new data frame altogether. A spark pipeline generally involves multiple transformations leading to multiple data frames being created. If spark stores all of these data frames, the memory requirement would be huge. So spark leaves the responsibility of persisting data frames to the user. Whichever data frame you are planning on re using, you can persist them and later unpersist them when done.
I don't think we can define spark read as an action or a transformation. Action/Transformation is applied over a data frame. To identify the difference, you should remember that the transformation operation will return a new dataframe while action will return some value/s.
Is there a general explanation, why spark needs so much more time to calculate the maximum value of a column?
I imported the Kaggle Quora training set (over 400.000 rows) and I like what spark is doing when it comes to rowwise feature extraction. But now I want to scale a column 'manually': find the maximum value of a column and divide by that value.
I tried the solutions from Best way to get the max value in a Spark dataframe column and https://databricks.com/blog/2015/06/02/statistical-and-mathematical-functions-with-dataframes-in-spark.html
I also tried df.toPandas() and then calculate the max in pandas (you guessed it, df.toPandas took a long time.)
The only thing I did ot try yet is the RDD way.
Before I provide some test code (I have to find out how to generate dummy data in spark), I'd like to know
can you give me a pointer to an article discussing this difference?
is spark more sensitive to memory constraints on my computer than pandas?
As #MattR has already said in the comment - you should use Pandas unless there's a specific reason to use Spark.
Usually you don't need Apache Spark unless you encounter MemoryError with Pandas. But if one server's RAM is not enough, then Apache Spark is the right tool for you. Apache Spark has an overhead, because it needs to split your data set first, then process those distributed chunks, then process and join "processed" data, collect it on one node and return it back to you.
#MaxU, #MattR, I found an intermediate solution that also makes me reassess Sparks laziness and understand the problem better.
sc.accumulator helps me define a global variable, and with a separate AccumulatorParam object I can calculate the maximum of the column on the fly.
In testing this I noticed that Spark is even lazier then expected, so this part of my original post ' I like what spark is doing when it comes to rowwise feature extraction' boils down to 'I like that Spark is doing nothing quite fast'.
On the other hand a lot of the time spent on calculating the maximum of the column has most presumably been the calculation of the intermediate values.
Thanks for yourinput and this topic really got me much further in understanding Spark.