[Wranglers] FoundationDB Meetup July 2025

[Ali, Saqib]

Register here:
https://www.meetup.com/foundationdb-meetup/events/308524378/

Agenda:

------------------------------------
Talk 1: RocksDB Storage Engine in FoundationDB (30 mins + Q&A)

Speaker: Neethu Haneesha Bingi, Software Engineer, Apple

Speaker Bio: Neethu is a software engineer with extensive experience in
designing and building reliable, scalable storage and distributed systems.
At Apple for over four years, she has been a contributor to FoundationDB
and led the initiative to integrate RocksDB as a supported storage engine.
Prior to Apple, Neethu held key engineering roles at Tintri, Oracle, and
Meta, with a strong focus on system reliability, performance optimization
in distributed systems, and storage infrastructure.

Talk Summary: FoundationDB now supports a new storage engine: RocksDB,
joining the existing lineup that includes SQLite, Redwood, and Memory
storage engines. In this talk, Neethu will introduce the motivation behind
integrating RocksDB into FoundationDB, share a high-level overview of
RocksDB itself, and explain the integration process. She will also dive
into tuning capabilities, operational considerations, and key challenges
encountered along the way. This session aims to raise awareness of this new
capability and what it means for the FoundationDB ecosystem.

------------------------------------
Talk 2: Self Healing FoundationDB From Gray Failures (30 mins + Q&A)

Speaker: Paymaan Raza, Software Engineer, Apple

Speaker Bio: Paymaan is a software engineer at Apple, where he works on
FoundationDB. With a deep interest in databases and infrastructure, he
enjoys tackling hard problems from first principles to build simple,
scalable, and useful systems. Prior to Apple, he developed a core streaming
platform for Meta's control plane and engineered key data infrastructure
components at LinkedIn, gaining extensive experience in crafting reliable
solutions for large-scale distributed environments.

Talk Summary: Gray failures in FoundationDB are subtle performance
degradations that, while not full crashes, impact system health and burden
operators with manual intervention. This talk presents a simple yet
effective self-healing mechanism designed to address such issues. We'll
demonstrate how we reused FoundationDB's core architectural building
blocks—such as unbundled roles, central Cluster Controller, transport layer
latency metrics, and fast recovery—to implement a "complaining" algorithm
where processes report slow peers, enabling automated detection and
remediation, thereby significantly improving FoundationDB's resilience and
reducing operational load.a
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://analyticsengineering.net/pipermail/wranglers/attachments/20250616/77b8a0d6/attachment.htm>