Spark-optimized Shuffle (SOS) any update?

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Spark-optimized Shuffle (SOS) any update?

marek-simunek
Hi everyone,
    we are facing same problems as Facebook had, where shuffle service is a bottleneck. For now we solved that with large task size (2g) to reduce shuffle I/O.

I saw very nice presentation from Brian Cho on Optimizing shuffle I/O at large scale[1]. It is a implementation of white paper[2].
Brian Cho at the end of the lecture kindly mentioned about plans to contribute it back to Spark[3]. I checked mailing list and spark JIRA and didn't find any ticket on this topic.

Please, does anyone has a contact on someone from Facebook who could know more about this? Or are there some plans to bring similar optimization to Spark?

[1] https://databricks.com/session/sos-optimizing-shuffle-i-o
[2] https://haoyuzhang.org/publications/riffle-eurosys18.pdf
[3] https://image.slidesharecdn.com/5brianchoerginseyfe-180613004126/95/sos-optimizing-shuffle-io-with-brian-cho-and-ergin-seyfe-30-638.jpg?cb=1528850545
Reply | Threaded
Open this post in threaded view
|

Re: Spark-optimized Shuffle (SOS) any update?

ifilonenko
Recently, the community has actively been working on this. The JIRA to follow is: 
https://issues.apache.org/jira/browse/SPARK-25299. A group of various companies including Bloomberg and Palantir are in the works of a WIP solution that implements a varied version of Option #5 (which is elaborated upon in the google doc linked in the JIRA summary). 

On Wed, Dec 19, 2018 at 5:20 AM <[hidden email]> wrote:
Hi everyone,
    we are facing same problems as Facebook had, where shuffle service is a bottleneck. For now we solved that with large task size (2g) to reduce shuffle I/O.

I saw very nice presentation from Brian Cho on Optimizing shuffle I/O at large scale[1]. It is a implementation of white paper[2].
Brian Cho at the end of the lecture kindly mentioned about plans to contribute it back to Spark[3]. I checked mailing list and spark JIRA and didn't find any ticket on this topic.

Please, does anyone has a contact on someone from Facebook who could know more about this? Or are there some plans to bring similar optimization to Spark?

[1] https://databricks.com/session/sos-optimizing-shuffle-i-o
[2] https://haoyuzhang.org/publications/riffle-eurosys18.pdf
[3] https://image.slidesharecdn.com/5brianchoerginseyfe-180613004126/95/sos-optimizing-shuffle-io-with-brian-cho-and-ergin-seyfe-30-638.jpg?cb=1528850545
Reply | Threaded
Open this post in threaded view
|

Re: Spark-optimized Shuffle (SOS) any update?

marek-simunek
Hi, thanks for reply. I finally got time and glanced through the design doc.

It seems that it has nothing to do with the paper I mentioned. The paper is trying to solve the problem of I/O ops required for shuffle are growing quadratically with number of tasks (shuffle files), therefore we need to keep number of tasks low.
Am I missing something?


Recently, the community has actively been working on this. The JIRA to follow is: 
https://issues.apache.org/jira/browse/SPARK-25299. A group of various companies including Bloomberg and Palantir are in the works of a WIP solution that implements a varied version of Option #5 (which is elaborated upon in the google doc linked in the JIRA summary). 

On Wed, Dec 19, 2018 at 5:20 AM <[hidden email]> wrote:
Hi everyone,
    we are facing same problems as Facebook had, where shuffle service is a bottleneck. For now we solved that with large task size (2g) to reduce shuffle I/O.

I saw very nice presentation from Brian Cho on Optimizing shuffle I/O at large scale[1]. It is a implementation of white paper[2].
Brian Cho at the end of the lecture kindly mentioned about plans to contribute it back to Spark[3]. I checked mailing list and spark JIRA and didn't find any ticket on this topic.

Please, does anyone has a contact on someone from Facebook who could know more about this? Or are there some plans to bring similar optimization to Spark?

[1] https://databricks.com/session/sos-optimizing-shuffle-i-o
[2] https://haoyuzhang.org/publications/riffle-eurosys18.pdf
[3] https://image.slidesharecdn.com/5brianchoerginseyfe-180613004126/95/sos-optimizing-shuffle-io-with-brian-cho-and-ergin-seyfe-30-638.jpg?cb=1528850545