HighLoad++ Siberia 2018
2018-06-25..2018-06-26
Novosibirsk, Russian Federation
| title | author | length |
|---|---|---|
| Video Report from HighLoad++ Siberia 2018 | 03m | |
Video report covering the HighLoad++ Siberia 2018 professional conference for developers of high-load systems, held in Novosibirsk on June 25–26. The footage… | ||
| Reliable Interaction in Distributed Systems | Mikhail Kurmaev | 40m |
Covers reliability issues in interactions between components of high-load distributed systems, focusing on connecting to and making requests to databases and… | ||
| PG Saga: Coordinating Dependent Data Changes Across Multiple Services | Konstantin Evteev | 48m |
Describes Avito’s approach to implementing cross-service business transactions and maintaining data consistency in a microservices environment that uses the… | ||
| How VK Inserts Data into ClickHouse from Tens of Thousands of Servers | Yuriy Nasretdinov | 48m |
Describes VK's production deployment of ClickHouse: why they adopted it, how much data they store, and the methods they use to write/insert data from tens of… | ||
| Replicating Data from MySQL to ClickHouse | Vladislav Klimenko, Valeriy Panov | 30m |
Describes moving analytical workloads from MySQL to ClickHouse by replicating data from MySQL into ClickHouse. Covers using ClickHouse’s scalable… | ||
| Machine learning @ booking.com | Viktor Bilyk | 44m |
Built a lightweight Scala prediction service that logs input vectors and predictions into Hadoop and aggregates results into databases, enabling a self‑service… | ||
| Multi-Armed Bandits in Recommendations | Mikhail Kamenshchikov | 32m |
Description of using multi-armed bandit algorithms in Avito’s Recommendations team as an alternative to A/B testing: technologies include Data Science and… | ||
| Speech Recognition: How to Build Speech-to-Text Yourself | Ivan Bondarenko | 56m |
Overview of building a Speech-to-Text system for tasks like call-center automation and smart-home voice control, comparing available market options from large… | ||
| Monitoring Cloud Infrastructure | Mikhail Prokopchuk | 34m |
Describes monitoring a self-hosted Kubernetes-based cloud using Prometheus, including operational experience, effective configuration, and the resource cost of… | ||
| The Evolution of Tools and Services at Badoo | Anton Turetskiy | 41m |
Overview of the operational tooling and services used by Badoo’s operations team, including service discovery with Consul, container deployment choices around… | ||
| A Million Video Calls a Day, or Call Your Mom | Aleksandr Tobol | 53m |
Covers building a large-scale video-calling service handling ~1,000,000 calls/day, focusing on signaling protocol details (multi-device calls), connection… | ||
| A System for Collecting Detailed Statistics on CDN Node Performance | Evgeniy Rossinskiy | 44m |
Describes building a detailed CDN node statistics-collection system to improve video playback quality by defining better quality metrics. It explains… | ||
| High-Performance Graph Database Built on Couchbase | Dmitriy Levanov | 50m |
A high-performance graph database built on Couchbase is presented to store and query large, branching user graphs for ad targeting. It addresses… | ||
| An Approach to Continuous Deployment in a Microservices Architecture | Aleksey Baitov | 01h 00m |
Describes how 2GIS implemented Continuous Deployment for a production microservices architecture, including the tooling and process used to deliver services… | ||
| Learning to Rank, a Little | Andrey Aksenov | 55m |
Shows how to build a custom ranking formula for search results, train a simple machine‑learning model to produce ranking scores, and evaluate whether the… | ||
| Frontend Architecture and Performance | Pavel Pavelko | 45m |
Examines frontend architecture and performance for e-commerce sites, focusing on how slow page loads hurt business metrics (citing Akamai research that a… | ||
| The history of IBSO migrations. Working with Oracle from 7 to 18 | Ivan Sharov | 53m |
The talk recounts the migration history of the IBSO core banking product while detailing how the team worked with Oracle across major versions from Oracle 7 up… | ||
| Event-driven architecture for distributed systems based on NServiceBus | Boris Tveritnev | 49m |
Presentation explains building event-driven distributed systems using NServiceBus. It covers challenges of exchanging information between interconnected… | ||
| Speeding up LZ4 decompression | Aleksey Milovidov | 46m |
The talk examines methods to speed up LZ4 decompression. It covers low-level CPU-specific optimizations ("black-magic") and data-science-driven techniques… | ||
| Applied mathematics for high loads | Aleksey Ragozin | 47m |
Survey of mathematical tools and theories that can be applied to building high-load information systems, arguing which techniques from mathematics and related… | ||
| njs — native JavaScript scripting in nginx | Dmitriy Volyntsev | 49m |
Presentation explains njs, the native JavaScript scripting engine embedded in nginx: its origin and architecture, what njs is, and how it works. It describes… | ||
| PostgreSQL monitoring fundamentals | Aleksey Lesovskiy | 45m |
Covers fundamentals of monitoring PostgreSQL databases in modern projects, explaining which database metrics and components matter for ensuring reliable… | ||
| New query execution features in Postgres 10, 11 & PostgresPro | Ivan Frolkov | 48m |
Covers new query execution features in PostgreSQL 10 and 11 and in PostgresPro: parallel query execution improvements introduced in PostgreSQL 10, multi-column… | ||
| Odyssey — a scalable connection pooler for PostgreSQL | D. Simonenko, D. Sarafannikov | 46m |
Odyssey is a scalable connection pooler for PostgreSQL that addresses the high overhead of establishing and maintaining database connections. It describes how… | ||
| How we integrated payment systems without using our own billing | Anton Rusakov | 29m |
A Badoo project needed to start issuing payouts to users but their existing billing system could only create invoices and accept payments, not send money. The… | ||
| On-demand test stands while splitting the monolith | Dmitriy Skokov | 43m |
Describes how the team implemented on-demand deployment of test environments while they were splitting a monolith into many services. Covers the iterative… | ||
| Performance testing from scratch | Ruslan Ziganshin | 45m |
Covers how to set up and run performance testing from scratch for systems that must meet SLA requirements, explaining why performance tests are necessary and… | ||
| How we reduced backend downtime during updates | A. Kablukov, R. Ziganshin | 49m |
Describes how a backend team reduced downtime during software updates by improving their deployment process. Covers concrete techniques and practices for… | ||
| Proactive performance optimization of Oracle databases | Aleksandr Makarov | 46m |
Presents a simple method for proactively finding backend bottlenecks using Oracle Database. The method is applied regularly (twice a year) and, in the… | ||
| Hardware won't fail me. How I prepare dozens of servers for battle every day | Artem Artemev | 27m |
Talk discusses preparing and provisioning dozens of physical servers daily for high-load production, focusing on hardware reliability and how redundancy in… | ||
| nothing matches | ||