Good question. What I have read about is that Spark is not a magician and
can't know how many tasks will be better for your input, so it can fail.
Spark set the default parallelism as twice the number of cores on the
cluster.
In my jobs, it seemed that using the parallelism inherited from input
I am following up on this question because I have a similar issue.
When is that we need to control the parallelism manually? Skewed partitions?
--
Sent from: http://apache-spark-user-list.1001560.n3.nabble.com/
-
To
* It is not getPartitions() but getNumPartitions().
El mar., 12 de feb. de 2019 a la(s) 13:08, Pedro Tuero (tuerope...@gmail.com)
escribió:
> And this is happening in every job I run. It is not just one case. If I
> add a forced repartitions it works fine, even better than before. But I run
>
And this is happening in every job I run. It is not just one case. If I add
a forced repartitions it works fine, even better than before. But I run the
same code for different inputs so the number to make repartitions must be
related to the input.
El mar., 12 de feb. de 2019 a la(s) 11:22, Pedro
Hi Jacek.
I 'm not using SparkSql, I'm using RDD API directly.
I can confirm that the jobs and stages are the same on both executions.
In the environment tab of the web UI, when using spark 2.4
spark.default.parallelism=128 is shown while in 2.3.1 is not.
But in 2.3.1 should be the same, because
Hi,
Can you show the plans with explain(extended=true) for both versions?
That's where I'd start to pinpoint the issue. Perhaps the underlying
execution engine change to affect keyBy? Dunno and guessing...
Pozdrawiam,
Jacek Laskowski
https://about.me/JacekLaskowski
Mastering Spark SQL
I did a repartition to 1 (hardcoded) before the keyBy and it ends in
1.2 minutes.
The questions remain open, because I don't want to harcode paralellism.
El vie., 8 de feb. de 2019 a la(s) 12:50, Pedro Tuero (tuerope...@gmail.com)
escribió:
> 128 is the default parallelism defined for the
128 is the default parallelism defined for the cluster.
The question now is why keyBy operation is using default parallelism
instead of the number of partition of the RDD created by the previous step
(5580).
Any clues?
El jue., 7 de feb. de 2019 a la(s) 15:30, Pedro Tuero (tuerope...@gmail.com)
Hi,
I am running a job in spark (using aws emr) and some stages are taking a
lot more using spark 2.4 instead of Spark 2.3.1:
Spark 2.4:
[image: image.png]
Spark 2.3.1:
[image: image.png]
With Spark 2.4, the keyBy operation take more than 10X what it took with
Spark 2.3.1
It seems to be