Name: 2023 RocksDB In-Person Meetup at Meta HQ (Zoom Available)
Start: 2023-12-12T15:30:00-08:00
End: 2023-12-12T20:00:00-08:00
Location: 305 Constitution Dr

***A parking pass is required to be shown for access to garage. Pass is included in the photos for the event as well as [here](https://www.dropbox.com/scl/fi/1ugp7se1s4f4skt2p4kdy/RocksDB-Meetup-parking-pass.png?rlkey=kqlz2kkzzwy6f7n9zasyzz1at&dl=0). You don't need to print the pass.***

We’re excited for our 2023 RocksDB meetup!

Date: Tuesday, December 12th from 3:30-8 pm
Where: Building 22 (Check-in at Lobby 2)
Conference Room: Transbay Tube

For remote attendees, here is the [zoom link](https://fb.zoom.us/j/97136137497?pwd=NWhyY0hSQUNnRndzaEJzSldrU0pQZz09).

***Internal Talks:***

**Tuning Optimizations on RocksDB Scans**:
Company: **Meta**
Speaker: **Akanksha Mahajan**
Abstract: RocksDB uses the readahead_size to prefetch data to reduce both disk read and overall scan latency. Come learn how we optimized this capability to reduce wasted IO and improve the efficiency of our engine.

**RocksDB Vision Expansion**
Company: **Meta**
Speaker: **Danny Chen**
Abstract: This talk will focus on Meta's strategic efforts to continue to innovate on our open-source storage engine and how we hope to serve the community and industry in the years to come.

***External Talks:***

**An efficient RocksDB Checkpoint Diff Mechanism**

* Company: **Cloudera**
* Speaker: **Prashant Pogde**
* Abstract: *RocksDB today is used by many open source and enterprise grade products out there. Apache Ozone is one such open source, distributed and scalable object store that uses RocksDB heavily to store most of its metadata. While we have leveraged as much as is offered by the RocksDB community, we have also augmented it with newer capabilities. An efficient rocksDB checkpoint diff mechanism is one such capability. This talk would cover the RocksDB checkpoint diff mechanism and how it addresses various use cases. It would also include challenges we faced including the impact of rocksDB compactions.*

**Key Value Unification at Pinterest**

* Company: **Pinterest**
* Speaker: **Rajath Prasad**
* Abstract: *In 2020, Pinterest embarked on a mission to consolidate the diverse key-value storage systems within the company. In this talk, we will delve into the comprehensive process of selecting RocksDB as the go-to storage engine for key-value storage at Pinterest. From improved performance and scalability to simplified operations and reduced maintenance costs, we will share how RocksDB has been the ideal choice for building our next-gen key-value infrastructure. Additionally, we will discuss challenges faced and the lessons learned while undertaking this massive migration.*

**Our experience with using RocksDB in TiKV**

* Company: **PingCap**
* Speakers: **Sunny Bains/Li Shen**
* Abstract: *In this talk, we'll talk about why we developed TiKV--an open-source, distributed, and transactional key-value database built with RUST--and it's evolution over the years including why RocksDB is used as a storage engine. We'll also discuss our learnings from using RocksDB in TiKV and some of the challenges including compaction, disaggregated storage engine, etc.*

**Dremel: Adaptive Configuration Tuning of RocksDB KV-Store**

* School: **University of Washington**
* Speaker: **Chenxingyu(Frank) Zhao**
* Abstract: *RocksDB is widely used to support many applications. However, configuring a RocksDB instance is challenging for the following reasons: 1) RocksDB has a massive parameter space to configure; 2) there are inherent trade-offs and dependencies between parameters; 3) right configurations are dependent on workload and hardware; and 4) evaluating configurations is time-consuming. In this talk, we present a system, Dremel, to adaptively and quickly configure RocksDB with strategies based on the Multi-Armed Bandit model.*

Agenda:

3:30 - 4:10 pm - Check-in/social time
4:10 pm: Kick-off & Welcome
4:15 pm: Our experience with using RocksDB in TiKV by PingCap (Sunny Bain/Li Shen)
4:35 pm: Key Value Unification at Pinterest(Rajath Prasad)
4:55 pm: RocksDB Vision Expansion by Meta (Danny Chen)
5:15 - 5:25 pm: Break
5:25 pm: An efficient RocksDB Checkpoint Diff Mechanism by Cloudera (Prashant Pogde)
5:45 pm: Tuning Optimizations on RocksDB Scans by Meta (Akanksha Mahajan)
6:05 pm: Dremel - Adaptive Configuration Tuning of RocksDB KV-Store by University of Washington (Frank Zhao)
6:25: Closing Thoughts
6:30 - 8 pm: Social Hour

Dinner will be served, with swag for in-person attendees(limited to first come first serve)!

Other Logistics:
Parking may be difficult so please arrive on-time.

We have reserved stalls on the 6th and 7th floor in the garage just north of Building 22 (305 Constitution Dr). Take any spots available in the garage on any floor. There is a security gate to enter through and attendants will help direct you. Please enter through Building 22, Lobby 2.

For Lyft/Uber there is a drop-off near Building 23 and a short walk over.

Note: Please check back on this event as we will update with more details as the date approaches.

Danny Chen

RocksDB

Technology

Open Source

Database Professionals

Software Development

Database Development

2023 RocksDB In-Person Meetup at Meta HQ (Zoom Available)

305 Constitution Dr

Share

RocksDB

2023 RocksDB In-Person Meetup at Meta HQ (Zoom Available)

RocksDB

Details

Members are also interested in