There should be no data-locality awareness with Beam on Flink because there are 
no APIs in Beam that Flink could use to schedule tasks with awareness. It seems 
it just happens that the readers are distributed as they are.

Are the files roughly of equal size?


> On 12. Mar 2018, at 05:50, Reinier Kip <r...@bol.com> wrote:
> Relevant versions: Beam 2.1, Flink 1.3.
> From: Reinier Kip <r...@bol.com>
> Sent: 12 March 2018 13:46:24
> To: user@beam.apache.org
> Subject: HDFS data locality and distribution, Flink
> Hey all,
> I'm trying to batch-process 30-ish files from HDFS, but I see that data is 
> distributed very badly across slots. 4 out of 32 slots get 4/5ths of the 
> data, another 3 slots get about 1/5th and a last slot just a few records. 
> This probably triggers disk spillover on these slots and slows down the job 
> immensely. The data has many many unique keys and processing could be done in 
> a highly parallel manner. From what I understand, HDFS data locality governs 
> which splits are assigned to which subtask.
> I'm running a Beam on Flink on YARN pipeline.
> I'm reading 30-ish files, whose records are later grouped by their millions 
> of unique keys.
> For now, I have 8 task managers by 4 slots. Beam sets all subtasks to have 32 
> parallelism.
> Data seems to be localised to 9 out of the 32 slots, 3 out of the 8 task 
> managers.
> Does the statement of input split assignment ring true? Is the fact that data 
> isn't redistributed an effort from Flink to have high data locality, even if 
> this means disk spillover for a few slots/tms and idleness for others? Is 
> there any use for parallelism if work isn't distributed anyway?
> Thanks for your time, Reinier

Reply via email to