hi, JB As co-creator of this project, I’d love to explain more about the positioning of lakehouse management system.
When discussing databases or traditional data warehouses, we often used the term DBMS (Database Management System) to describe them. Traditional databases, including MPP databases, are typically considered “out-of-box” solutions. Unlike big data systems, they don’t require various components like compute engines, data lake formats, or metadata stores. When we need a database management tool, lightweight options like Navicat are commonly used. If we further abstract the capabilities of a DBMS and map them to the modern data stack, we find that the data read/write part of a DBMS is now shared among different compute engines such as Spark, Flink, Trino, and cloud-native services like Athena. Another part of a DBMS deals with data files, index files, and metadata (also known as the information schema) maintenance. Currently, there are successful open-source and commercial projects dedicated to managing metadata, such as HiveMetastore, UnityCatalog, and more recently, Gravitino. In practice, developers often combine these projects with compute engines to optimize data files. For example, many commercial compute engines include an optimize command. Amoro, as a lakehouse management system, aims to encapsulate the maintenance and management of data lake files, index files, and metadata in a way that is transparent and easy-to-use for users. The richness of diverse computing engines is a distinctive feature of the modern data stack, opening up a multitude of possibilities for various application scenarios. Additionally, concerning the part analogous to DBMS, we aspire to have a mature system in place—one that seamlessly accommodates data written to the lakehouse by any engine, in any manner, ensuring high data availability across all other engines. For instance, when Flink writes to Iceberg, Amoro’s self-optimizing mechanism ensures efficient data analysis performance by Trino or other engines while controlling compacting costs. Additionally, Amoro handles historical data, snapshots, and orphan file cleanup in the background. By positioning Amoro in this way, we aim to provide an ‘out-of-box’ experience that feels as straightforward as traditional DBMS while keeping openness to various computing engines. At the same time, Amoro hopes to empower data product builders with a lightweight solution that integrates seamlessly into their modern data workflows. Thanks. Jin Ma On 2024/02/23 14:16:43 Jean-Baptiste Onofré wrote: > Hi Justin > > Even if it looks interesting, I'm not sure to understand exactly the > purpose of the proposal. > > What lakehouse management system means exactly ? Is it an abstraction > layer on top of Iceberg, Paimon + query engine powered by Flink, > Spark, Trino ? > > Please let me know if you want an additional mentor, I would be happy to help. > > Thanks ! > Regards > JB > > On Fri, Feb 23, 2024 at 9:44 AM Justin Mclean <ju...@classsoftware.com> wrote: > > > > Hi, > > > > I would like to propose a new project to the ASF incubator - Apache Amoro. I’m one of the mentors, but there are a lot of other people involved who have done all of the hard work. > > > > Amoro is a Lakehouse management system built on open data lake formats like Apache Iceberg and Apache Paimon (Incubating). Working with compute engines including Apache Flink, Apache Spark, and Trino, Amoro brings pluggable and self-managed features for Lakehouse to provide out-of-the-box data warehouse experience, and helps data platforms or products easily build infra-decoupled, stream-and-batch-fused and lake-native architecture. You can find the proposal here. [1] > > > > We are looking forward to anyone's feedback or questions. > > > > Thanks, > > Justin > > > > [1] https://cwiki.apache.org/confluence/display/INCUBATOR/AmoroProposal > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org > > For additional commands, e-mail: general-h...@incubator.apache.org > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org > For additional commands, e-mail: general-h...@incubator.apache.org > >
