Andrei Manakov

A nice and short article about bloom filter llimllib.github.io/bloomfilter-... . It explains it great and bunch of good examples

How We Replaced Node.js with Apache Flink for Real-Time Deduplication and Cut Costs by 7x ShareChat is one of the largest social media platforms in India, with over 180 million monthly active users. We had a high-throughput real-time stream (>200K RPS) processing using a Node.js + Redis-ba...

Recording is published here current.confluent.io/post-confere...

I just finished my talk at #current2025 . The recording will be available soon. But you can find my slides already here docs.google.com/presentation/d…

Thanks @confluent.io for the opportunity to talk at the Current!

I have got 5th node to my scylla home cluster in #current25 from @scylladb.com . Thanks, I love them very much!

🌊 I’m excited to share that I’m speaking at #Current25 London! Join me for my talk about Real-Time Deduplication, where I’ll dive into writing efficient pipelines at a scale of more than 3 billions events per day.
📅 May 20 - 21, 2025
📍 ExCel London
🔗 Register now ➡️ cnfl.io/3GFvj1b @confluent.io

New blog post! 🔥

A practical guide to get you started with Go performance tuning. Covers setting up metrics , profiling , benchmarking, and leveraging PGO. Check it out for faster, more efficient Go services!

obrepalski.com/posts/observ...

#go #performance #observability

My colleague Ivan had an a very deep talk last year, I highly recommend to check it out

I received the speaker swag and assembled whole Scylla family. We enjoyed the Monster Scale summit together 🙂 Thanks @scylladb.com for the amazing conference!
#scylladb #scyllaSummit

YouTube video by ScyllaDB Monster Scale Summit 2025 - Andrei Manakov, Senior Staff Software Engineer at ShareChat

Efficient deduplication is a crucial component for providing relevant personal recommendations. I will explain how we solved this problem and reduced server cost by 90% www.youtube.com/watch?v=J_tF...

Join the Monster Scale Summit to check it out www.scylladb.com/monster-scal...

Good article about RecSys architecture, I liked it docs.nvidia.com/deeplearning...

YouTube video by The Geek Narrator AWS Aurora Distributed SQL internals with Marc Brooker - ​ @amazonwebservices

Today I released the next episode on The GeekNarrator, with @marcbrooker.bsky.social on AWS Aurora DSQL internals. This is filled with amazing insights.

Don't miss this one! youtu.be/ONkfXvPSSpc

Yes, I definitely use the last scyllaDB version with tablets😂.

I have got p99 swag pack and I finally assembled my home scyllaDB cluster. Thanks @scylladb.com !

Database Fundamentals The fundamental problems a database engineer thinks about in the shower

It’s really decent post about database internals tontinton.com/posts/databa...

Getting the Most Out of ScyllaDB Monitoring: ShareChat's Tips - ScyllaDB ScyllaDB monitoring provides a lot of useful information. But sometimes it’s not easy to find the root of the problem if something is wrong or even estimate the remaining capacity by the load on the c...

By the way, you might find this talk from my colleague Andrey @andection.bsky.social useful: www.scylladb.com/tech-talk/ge...
It provides some really useful advanced tips on how to troubleshoot Scylla clusters.

ScyllaDB: How To Analyse Cluster Capacity? Using these recommendations you can reduce your server cost without any latency degradation.

I wrote a new blog post about analysing #ScyllaDB cluster capacity. ScyllaDB is a highly performant database that supports gradual degradation under load. The post might help in understanding its true capacity and reducing server costs to some extent. open.substack.com/pub/andectio...

ScyllaDB 2024.2 Introduces New Efficiency & Elasticity via “Tablets” - ScyllaDB Scaling operations with Tablets are up to 30X faster while reducing network costs by up to 50%

The tablets feature
of ScyllaDB is very impressive. It simplifies maintenance and looks like a solid foundation for possible autoscaling www.scylladb.com/2024/12/03/e...

YouTube video by GOTO Conferences Beyond the Hype: A Realistic Look at Large Language Models • Jodie Burchell • GOTO 2024

Interesting and simple explanation Large Language Models youtu.be/Pv0cfsastFs?...

I never understood it as well.. once you are really proficient at any modern language you can easily learn a new language and solving close problems. It will be harder to change technology domain (from distributed systems to mobile dev)

YouTube video by The Geek Narrator How would you design a database on Object Storage?

Many databases are moving their state to Object storage. What makes it feasible in 2024?

More importantly, how would you design a database on top of an Object Storage?

Just released next episode with @sirupsen.bsky.social on The GeekNarrator podcast. youtu.be/yfgD0Pa7o58

Please like, share etc

An unordered / non-exhaustive list of things that helped scale Bluesky's infra efficiently:

+ Exiting the cloud (colocation)
+ HAProxy w/many Node backends
+ Go w/clever code
+ ScyllaDB
+ SQLite w/per user databases
+ Redis w/many instances
+ AMD servers w/many cores
+ Purchasing bandwidth directly

It sounds like the first service mesh version at our company🙂. What did the control plane look like?