Amazon's Exabyte-Scale Migration from Apache Spark to Ray on Amazon EC2

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by igmor on Wednesday July 31, 2024 @04:50PM (#41123311)

Can you share any data on how big of a cluster is running Ray jobs?

[ Parent | on HN ]

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by thedood on Wednesday July 31, 2024 @05:26PM (#41123634)

From the blog post, the largest individual Ray cluster that was observed running a production compaction job in Q1 had 26,846 vCPUs and ~210TiB of RAM. This is roughly equivalent to a Ray cluster composed of 839 r5.8xlarge EC2 nodes (w/ 32 vCPUs and 256GiB RAM per node).

[ Parent | on HN ]

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by layoric on Wednesday July 31, 2024 @07:57PM (#41124811)

For those interested, this would be at a cost of:

- ~$1691/hour on demand

- ~$1065/hour reserved

- ~$521/hour spot

Not including any related data transfer costs.

[ Parent | on HN ]

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by LarsDu88 on Thursday August 01, 2024 @05:13PM (#41133691)

Wow that is not NEARLY as expensive as I would have imagined considering the scale of the data involved.

[ Parent | on HN ]

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by artemisart on Thursday August 01, 2024 @05:50PM (#41134010)

That's also $41k / 26k / 13k per day or $1.2M / 767k / 375k per month!

[ Parent | on HN ]

Re:Amazon's Exabyte-Scale Migration from Apache Spark to Ray... by LarsDu88 on Thursday August 01, 2024 @06:16PM (#41134201)

I mean, for all of Amazon's business intelligence, running on servers operated by Amazon... tis a mere pittance.

[ Parent | on HN ]