Highload fwdays'25
2025-06-14
-- online --
as the organisers put it
Highload fwdays'25 - The conference is dedicated to the development of high-load systems
who it's for
backend engineers, platform engineers, site reliability engineers, devops engineers, database administrators, data engineers, observability engineers
topics
Redis to Scylla migration performance; Istio Ambient Mesh (sidecar-less); Kubernetes autoscaling with KEDA; Temporal workflow cluster scaling; Elasticsearch observability at scale; Logstash persistent queue and ILM; Postgres tuning and partitioning; Kafka at high event rates (80k/min); vector search with on-disk IVF and quantization; service-access pattern to replace direct DB connections
| title | author | length |
|---|---|---|
| Break Out of Redis | Anton Moldovan | 01h 02m |
A company decided to stop using Redis as a critical datastore after repeated production incidents with Redis Sentinel — full resyncs, slow failovers and… | ||
| Provisioning via DOT-Chain: From Power Supply to a Drone Marketplace | Volodymyr Chuhai | 44m |
Presentation describes building a DOT-Chain infrastructure from scratch to support provisioning across use cases from catering to a drone marketplace, focusing… | ||
| Istio Ambient Mesh in Production: Our Journey from Sidecar to Sidecar-less | Hlib Smoliakov | 27m |
A team describes migrating their Kubernetes-based Istio deployment from the traditional sidecar model to Istio Ambient Mesh (sidecar-less) to reduce resource… | ||
| Scaling in Space and Time with Temporal | Andrii Lupa | 50m |
Describes using Temporal.io to keep real-time reporting systems current under nonlinear, spiky load. Covers applying Temporal for both short-lived pipelines… | ||
| Building Observability with Elasticsearch: Practices for High Load | Anton Prykhodko | 40m |
Practical guidance for running Elasticsearch as an observability backend under high load, covering index and shard layout, ILM (index lifecycle management)… | ||
| Growing Resilience in a Hyperscaled Financial Platform | Dmytro Hnatiuk | 36m |
Describes scaling a financial platform from 66 million to over 25 billion records while preserving numeric accuracy and low latency. Covers pragmatic… | ||
| Surviving 1 Billion Vectors Without Selling a Kidney: Our Low-cost Cluster | Maksym Mova | 29m |
A team describes scaling a vector search project from under 2 million vectors to nearly 1 billion, facing performance, index size, and resource limits. They… | ||
| Database Isolation: How We Fight a Hundred Direct DB Connections | Mykhailo Kratiuk | 51m |
The talk describes Solidgate’s migration from many services making hundreds of direct database connections to a service-based data access model. It covers… | ||
| How to Survive Black Friday: Preparing e‑commerce for Peak Season | Yurii Panaiotov | 35m |
Describes preparing an e-commerce platform for Black Friday peak season, focusing on configuring autoscaling and load balancing and the key operational aspects… | ||
| nothing matches | ||