Re: Breaking the previous large-scale sort record with Spark
Hi all, We are excited to announce that the benchmark entry has been reviewed by the Sort Benchmark committee and Spark has officially won the Daytona GraySort contest in sorting 100TB of data. Our entry tied with a UCSD research team building high performance systems and we jointly set a new world record. This is an important milestone for the project, as it validates the amount of engineering work put into Spark by the community. As Matei said, For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Updated blog post: http://databricks.com/blog/2014/11/05/spark-officially-sets-a-new-record-in-large-scale-sorting.html On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Congrats to everyone who helped make this happen. And if anyone has even more machines they'd like us to run on next year, let us know :). Matei On Nov 5, 2014, at 3:11 PM, Reynold Xin r...@databricks.com wrote: Hi all, We are excited to announce that the benchmark entry has been reviewed by the Sort Benchmark committee and Spark has officially won the Daytona GraySort contest in sorting 100TB of data. Our entry tied with a UCSD research team building high performance systems and we jointly set a new world record. This is an important milestone for the project, as it validates the amount of engineering work put into Spark by the community. As Matei said, For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Updated blog post: http://databricks.com/blog/2014/11/05/spark-officially-sets-a-new-record-in-large-scale-sorting.html On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
The biggest scaling issue was supporting a large number of reduce tasks efficiently, which the JIRAs in that post handle. In particular, our current default shuffle (the hash-based one) has each map task open a separate file output stream for each reduce task, which wastes a lot of memory (since each stream has its own buffer). A second thing that helped efficiency tremendously was Reynold's new network module (https://issues.apache.org/jira/browse/SPARK-2468). Doing I/O on 32 cores, 10 Gbps Ethernet and 8+ disks efficiently is not easy, as can be seen when you try to scale up other software. Finally, with 30,000 tasks even sending info about every map's output size to each reducer was a problem, so Reynold has a patch that avoids that if the number of tasks is large. Matei On Oct 10, 2014, at 10:09 PM, Ilya Ganelin ilgan...@gmail.com wrote: Hi Matei - I read your post with great interest. Could you possibly comment in more depth on some of the issues you guys saw when scaling up spark and how you resolved them? I am interested specifically in spark-related problems. I'm working on scaling up spark to very large datasets and have been running into a variety of issues. Thanks in advance! On Oct 10, 2014 10:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Thank you for the details! Would you mind speaking to what tools proved most useful as far as identifying bottlenecks or bugs? Thanks again. On Oct 13, 2014 5:36 PM, Matei Zaharia matei.zaha...@gmail.com wrote: The biggest scaling issue was supporting a large number of reduce tasks efficiently, which the JIRAs in that post handle. In particular, our current default shuffle (the hash-based one) has each map task open a separate file output stream for each reduce task, which wastes a lot of memory (since each stream has its own buffer). A second thing that helped efficiency tremendously was Reynold's new network module (https://issues.apache.org/jira/browse/SPARK-2468). Doing I/O on 32 cores, 10 Gbps Ethernet and 8+ disks efficiently is not easy, as can be seen when you try to scale up other software. Finally, with 30,000 tasks even sending info about every map's output size to each reducer was a problem, so Reynold has a patch that avoids that if the number of tasks is large. Matei On Oct 10, 2014, at 10:09 PM, Ilya Ganelin ilgan...@gmail.com wrote: Hi Matei - I read your post with great interest. Could you possibly comment in more depth on some of the issues you guys saw when scaling up spark and how you resolved them? I am interested specifically in spark-related problems. I'm working on scaling up spark to very large datasets and have been running into a variety of issues. Thanks in advance! On Oct 10, 2014 10:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Well done guys. MapReduce sort at that time was a good feat and Spark now has raised the bar with the ability to sort a PB. Like some of the folks in the list, a summary of what worked (and didn't) as well as the monitoring practices would be good. Cheers k/ P.S: What are you folks planning next ? On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Congrats to Reynold et al leading this effort! - Henry On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Awesome news Matei ! Congratulations to the databricks team and all the community members... On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Brilliant stuff ! Congrats all :-) This is indeed really heartening news ! Regards, Mridul On Fri, Oct 10, 2014 at 8:24 PM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: dev-unsubscr...@spark.apache.org For additional commands, e-mail: dev-h...@spark.apache.org - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Great! Congratulations! -- Nan Zhu On Friday, October 10, 2014 at 11:19 AM, Mridul Muralidharan wrote: Brilliant stuff ! Congrats all :-) This is indeed really heartening news ! Regards, Mridul On Fri, Oct 10, 2014 at 8:24 PM, Matei Zaharia matei.zaha...@gmail.com (mailto:matei.zaha...@gmail.com) wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: dev-unsubscr...@spark.apache.org (mailto:dev-unsubscr...@spark.apache.org) For additional commands, e-mail: dev-h...@spark.apache.org (mailto:dev-h...@spark.apache.org) - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org (mailto:user-unsubscr...@spark.apache.org) For additional commands, e-mail: user-h...@spark.apache.org (mailto:user-h...@spark.apache.org)
Re: Breaking the previous large-scale sort record with Spark
Wonderful !! On 11 Oct, 2014, at 12:00 am, Nan Zhu zhunanmcg...@gmail.com wrote: Great! Congratulations! -- Nan Zhu On Friday, October 10, 2014 at 11:19 AM, Mridul Muralidharan wrote: Brilliant stuff ! Congrats all :-) This is indeed really heartening news ! Regards, Mridul On Fri, Oct 10, 2014 at 8:24 PM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: dev-unsubscr...@spark.apache.org For additional commands, e-mail: dev-h...@spark.apache.org - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Breaking the previous large-scale sort record with Spark
Great stuff. Wonderful to see such progress in so short a time. How about some links to code and instructions so that these benchmarks can be reproduced? Regards, - Steve From: Debasish Das debasish.da...@gmail.com Date: Friday, October 10, 2014 at 8:17 To: Matei Zaharia matei.zaha...@gmail.com Cc: user user@spark.apache.org, dev d...@spark.apache.org Subject: Re: Breaking the previous large-scale sort record with Spark Awesome news Matei ! Congratulations to the databricks team and all the community members... On Fri, Oct 10, 2014 at 7:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort- record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org -- CONFIDENTIALITY NOTICE NOTICE: This message is intended for the use of the individual or entity to which it is addressed and may contain information that is confidential, privileged and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient, you are hereby notified that any printing, copying, dissemination, distribution, disclosure or forwarding of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and delete it from your system. Thank You.
Re: Breaking the previous large-scale sort record with Spark
Hi Matei - I read your post with great interest. Could you possibly comment in more depth on some of the issues you guys saw when scaling up spark and how you resolved them? I am interested specifically in spark-related problems. I'm working on scaling up spark to very large datasets and have been running into a variety of issues. Thanks in advance! On Oct 10, 2014 10:54 AM, Matei Zaharia matei.zaha...@gmail.com wrote: Hi folks, I interrupt your regularly scheduled user / dev list to bring you some pretty cool news for the project, which is that we've been able to use Spark to break MapReduce's 100 TB and 1 PB sort records, sorting data 3x faster on 10x fewer nodes. There's a detailed writeup at http://databricks.com/blog/2014/10/10/spark-breaks-previous-large-scale-sort-record.html. Summary: while Hadoop MapReduce held last year's 100 TB world record by sorting 100 TB in 72 minutes on 2100 nodes, we sorted it in 23 minutes on 206 nodes; and we also scaled up to sort 1 PB in 234 minutes. I want to thank Reynold Xin for leading this effort over the past few weeks, along with Parviz Deyhim, Xiangrui Meng, Aaron Davidson and Ali Ghodsi. In addition, we'd really like to thank Amazon's EC2 team for providing the machines to make this possible. Finally, this result would of course not be possible without the many many other contributions, testing and feature requests from throughout the community. For an engine to scale from these multi-hour petabyte batch jobs down to 100-millisecond streaming and interactive queries is quite uncommon, and it's thanks to all of you folks that we are able to make this happen. Matei - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org