ABSTRACT: F1 Query is a federated query processing engine at Google that handles billions of SQL queries across diverse data sources like Bigtable and Spanner on a daily basis. It seamlessly supports OLTP-style point queries, low-latency OLAP querying, and large ETL pipelines while integrating declarative queries with custom business logic to streamline development. This capability is particularly relevant in the context of HTAP (Hybrid Transactional and Analytical Processing) systems, where the demand for simultaneously running transactional and analytical workloads over the same dataset is on the rise. F1 Lightning, an extension service to F1 Query, addresses the challenge of integrating existing transactional data from multiple systems, supporting both new and legacy applications. This talk presents the design and production experience of F1 Query and F1 Lightning, catering to some of Google's most demanding applications.
BIO: Ahmed Mahmood is a Senior Software Engineer at Google, specializing in distributed geospatial processing, query optimization, and distributed execution as part of the F1 Query Team. He earned his Ph.D. in Computer Science from Purdue University, focusing on distributed spatial-keyword stream processing. Ahmed is also the author of the book "Scalable Processing of Spatial-Keyword Queries." His research has been published in prestigious research venues, including ACM SIGSPATIAL, VLDB, ACM SIGMOD, ACM TSAS, and IEEE ICDE.
ABSTRACT: F1 Query is a federated query processing engine at Google that handles billions of SQL queries across diverse data sources like Bigtable and Spanner on a daily basis. It seamlessly supports OLTP-style point queries, low-latency OLAP querying, and large ETL pipelines while integrating declarative queries with custom business logic to streamline development. This capability is particularly relevant in the context of HTAP (Hybrid Transactional and Analytical Processing) systems, where the demand for simultaneously running transactional and analytical workloads over the same dataset is on the rise. F1 Lightning, an extension service to F1 Query, addresses the challenge of integrating existing transactional data from multiple systems, supporting both new and legacy applications. This talk presents the design and production experience of F1 Query and F1 Lightning, catering to some of Google's most demanding applications.
BIO: Ahmed Mahmood is a Senior Software Engineer at Google, specializing in distributed geospatial processing, query optimization, and distributed execution as part of the F1 Query Team. He earned his Ph.D. in Computer Science from Purdue University, focusing on distributed spatial-keyword stream processing. Ahmed is also the author of the book "Scalable Processing of Spatial-Keyword Queries." His research has been published in prestigious research venues, including ACM SIGSPATIAL, VLDB, ACM SIGMOD, ACM TSAS, and IEEE ICDE.