Sai Srirampur is the co-founder of PeerDB and a veteran Postgres Solutions Engineer with experience at Citus Data and Microsoft. He has been at the forefront of optimizing and scaling Postgres for large data workloads and is now spearheading innovation in data movement and replication with PeerDB. In this episode, we'll discuss the challenges of tuning massive Postgres systems, real-time data streaming solutions, and PeerDB’s vision for the future.

In this episode we explore:

• Leaving Microsoft and starting Sai’s own company, PeerDB
• Developing PeerDB's data movement and replication tool
• Why Postgres might not be suitable for everything at a large scale
• Having team members be database experts first before being software developers
• The growing customer demand for Clickhouse and Snowflake

Links mentioned:

PeerDB

Sai Srirampur on X (@saisrirampur)

Sai Srirampur on LinkedIn

In this episode, we talk Postgres committer Melanie Plagemean about her involvement with @PGConfdev, improvements to vacuum scheduling and auto vacuum configuration, the logical replication of DDL, and platforms such as CNCF and Kubernetes integrating with Postgres.

Links mentioned:

PGConf.dev

YouTube – CMU Database Group

Path To Citus Con Podcast

@melanieplageman@mastodon.social

Melanie on X (@melanieplageman)

Bharath Rupireddy has carved a niche for himself in the Postgres community since he began using the database system back in 2020. From his start at EnterpriseDB to making strides at Microsoft, and now contributing to the AWS open-source project, Bharath is entrenched in the inner workings of Postgres development. He has worked in many areas of Postgres such as WAL, Replication, pg_walinspect extension, bug fixes, performance improvements, new SQL functions etc.

In this episode we explore:

• Starting with Postgres in 2020 and learning the basics while contributing to Parallel COPY feature
• Postgres recent features such as pg_walinspect extension, WAL source switch from archive to streaming, WAL insertion lock improvements, WAL read from buffers, more replication slot invalidation mechanisms like XID age based and inactive_timeout based ones, new table access methods for multi inserts.
• The future of native active-active replication
• The complexities of conflict resolution
• Using PostgreSQL 16 in production scenarios for active-active replication
• The role of databases evolving within different industries

Links mentioned:

¡Databases! – A Database Seminar Series

Hacking Postgres, Ep. 9: Bertrand Drouvot

PGConf India

Hyderabad PostgreSQL User Group

PGConf.dev

Bharath on LinkedIn

Bharath on X (@BRupireddy)

Bharath on Github