HighLoad++ 2014
2014-10-31..2014-11-01
Moscow, Russian Federation
[HighLoad++ 2012, HighLoad++ 2013, HighLoad++ 2014, HighLoad++ 2015, HighLoad++ 2016, HighLoad++ 2017, HighLoad++ 2019, HighLoad++ 2022, HighLoad++ 2023, HighLoad++ 2024, HighLoad++ 2025]
| title | author | length |
|---|---|---|
| A Story about HighLoad++ 2014 | 01m | |
A report-style piece presents coverage of the HighLoad++ 2014 conference, including reports and interviews about building and operating high-load systems. It… | ||
| Interview | Maksim Shaposhnikov | 01m |
An interview-style report from the HighLoad++ 2014 conference covering professional discussions for developers of high-load systems. It focuses on the… | ||
| Interview | Oleg Bunin | 01m |
A set of interviews and reports from the HighLoad++ 2014 conference about building and operating high-load systems. The material covers practitioners’… | ||
| Interview | Omer Shapira | 01m |
A set of interviews and reports from the HighLoad++ 2014 conference about building and operating high-load systems. The coverage highlights practitioner… | ||
| Interview | Andrey Sas | 01m |
Collection of interviews and reports from the HighLoad++ 2014 conference focusing on building and operating high‑load systems; includes conversations with… | ||
| Interview | Igor Myzgin | 01m |
A set of interviews and reports from the HighLoad++ 2014 conference about building and operating high-load systems. The content covers practitioner… | ||
| Interview | Leonid Yurev | 01m |
Collection of interviews and reports from the HighLoad++ 2014 conference about building and operating high-load systems. Includes conversations with… | ||
| !sync: asynchronous interaction | Vyacheslav Turchaninov | 43m |
Covers how operating system schedulers, system calls, and asynchronous communication relate; explains how asynchronous interaction fundamentally works and when… | ||
| Implementing the SPDY Protocol in the LinkedIn Social Network | Oleg Shapiro | 40m |
LinkedIn's team describes implementing Google's SPDY protocol (basis for HTTP/2) across their global web stack to speed delivery of static content. The talk… | ||
| 1Hippeus: zerocopy messaging by the Laws of Sparta | Leonid Yurev | 54m |
1Hippeus is an open-source messaging infrastructure designed for extreme throughput (handling billions/sec) using zero-copy and lock-free communication. It… | ||
| Coub: how we built an analytical platform | Mikhail Tabunov | 36m |
Coub built its own analytics platform to process and analyze about 1 billion events per month from players, an iOS app, and the website. The talk explains why… | ||
| HighLoad for Beginners | Dmitriy Obukhov | 36m |
An introductory overview of high-load systems explaining what “highload” means in practical terms. Covers basic concepts and definitions for developers new to… | ||
| Nutanix Acropolis - KVM-based Cloud Turnkey | Maksim Shaposhnikov | 49m |
Nutanix Acropolis presents a turnkey cloud platform built on KVM, arguing KVM is now a top choice as a free, reliable, open-source hypervisor. It addresses the… | ||
| NoBigData: streaming analytics for ClientSide performance | S.Ryzhikov, A.Serbul | 43m |
Describes NoBigData, a streaming analytics system for measuring client-side web performance using browser APIs like the Navigation Timing API. It covers… | ||
| SDCH or new approaches to increasing performance | Dmitriy Markovich | 57m |
Overview of SDCH (Shared Dictionary Compression over HTTP) as an alternative to gzip for improving web transfer performance. Describes LinkedIn’s experimental… | ||
| Sharding: patterns and antipatterns | K.Osipov, A.Rybak | 56m |
Крупнейшая профессиональная конференция для разработчиков высоконагруженных систем Saint HighLoad++ 2026
Подробнее: https://clck.ru/3QZHTb
Июнь… | ||
| Backend Architecture of sputnik.ru Maps | Maksim Dementev | 43m |
A talk about Sputnik.ru’s backend that generates raster map tiles from OpenStreetMap data using a horizontally scalable distributed rendering cluster. It… | ||
| Virtualized highload NFV | Aleksandr Shalimov | 47m |
The talk explains Network Functions Virtualization (NFV) and how moving network services like traffic analyzers, firewalls, and load balancers from specialized… | ||
| Infrastructure Is Part of Your Application | Aleksandr Azimov | 38m |
Shows how to treat infrastructure as part of an application by measuring and improving network-level reliability. Describes methods to evaluate connectivity at… | ||
| Panel with the HighLoad++ Program Committee on choosing technologies for a project | 53m | |
A round-table discussion on choosing technologies for a project, held by the HighLoad++ program committee. The session covers evaluation and selection of… | ||
| Escape Spam — how to increase newsletter CTR | Andrey Sas | 48m |
Techniques to recover email campaigns from being marked as spam by increasing click-through rate (CTR) without harming overall site user activity. Covers what… | ||
| Wireless access for HighLoad participants | Aleksandr Pavlov | 26m |
Describes building wireless networks for large events (exhibitions, stadiums) and why traditional enterprise controller-plus-many-light-AP setups don’t scale… | ||
| Thorny path to the Large Scale Graph Processing | Aleksey Zinovev | 39m |
Big graphs like the web hit sizes that break single servers — examples: Moscow road graph 250k nodes/460k edges, EU roads ~18M/42M, and web/facebook scales to… | ||
| Scalable nginx configuration | Igor Sysoev | 44m |
How to design nginx configurations that remain maintainable over many years, covering common mistakes and comparing correct and incorrect approaches. The talk… | ||
| Elbrus primer | Konstantin Trushkin | 55m |
An overview of the Elbrus project: history of MCST, milestones in its development, and current product lines. Covers Elbrus CPU architecture features… | ||
| Heterogeneous services for highload projects | Igor Myzgin | 35m |
The talk analyzes monetization constraints in the Russian web (Runet) that prevent valuing audience the same way as in Europe/USA, causing a widening gap… | ||
| Tempesta FW: FrameWork and FireWall for WAF and DDoS mitigation | Aleksandr Krizhanovskiy | 40m |
Tempesta FW is an open-source hybrid HTTP accelerator and firewall designed for application-layer DDoS protection and building high-performance Web Application… | ||
| Docker & Puppet: how to cross them and whether you need it | Anton Turetskiy | 44m |
The talk compares Docker and Puppet and explains why the team introduced containerization alongside their existing configuration management. It covers… | ||
| DDoS attacks in Russia in 2014: UDP packets do not smell of lard | Aleksandr Lyamin | 48m |
Analyzes DDoS trends in Russia through the first 10 months of 2014, correlating attack patterns with geopolitical and economic events. Reviews common… | ||
| Application-level protocols in the browser | Ilya Kutukov | 44m |
Explains how implementing custom application-layer protocols inside the browser can improve channel utilization, enable arbitrary encryption, manage session… | ||
| ObjectManager, or How to work with a large number of objects on a map | Marina Stepanova | 30m |
Describes ObjectManager, a client-side technique for drawing large numbers of map points efficiently and the limits of standard marker solutions. Covers… | ||
| Development management tooling at Mail.Ru Group | Aleksandr Gornyy | 45m |
Describes Mail.Ru Group’s development-management toolset and how the company uses specific Atlassian and related products to run development processes. Lists… | ||
| Continuous Delivery, or the magic button for on‑demand releases | Denis Yakovlev | 38m |
Describes implementing a one-click Continuous Delivery pipeline for a complex distributed product (2GIS Reference API) so a manager can trigger a full release… | ||
| People make the difference, or video streaming | Aleksandr Tobol | 24m |
Describes how a video service sped up start-of-play for movies and series by over 10× using adaptive streaming technologies MPEG‑DASH and HLS. Covers which… | ||
| Payment processing in Badoo — an inside view | Anatoliy Panov | 43m |
An overview of Badoo’s billing and payments infrastructure for a large international, high-load service, covering architecture, operational challenges that… | ||
| How to Accurately Store Money in a Database | Roman Druzyagin | 52m |
Discusses correct ways to store and operate on monetary values in application databases, highlighting common developer mistakes when implementing custom… | ||
| Asynchronous Replication Without Censorship | Oleg Tsarev | 41m |
A Percona Server developer and Mail.Ru Target engineer analyzes MySQL replication: explains its internal design, points out flawed choices made in MySQL… | ||
| CREATE INDEX… USING VODKA. VODKA CONNECTING INDEXES! | Oleg Bartunov, Aleksandr Korotkov | 40m |
Prototype index access method called Vodka appears, replacing GIN’s B-tree entries with an ART/digital tree and combining multiple index types (including… | ||
| How PostgreSQL Interacts with the Disk | Ilya Kosmodemyanskiy | 45m |
Explains why a database needs disk storage and how PostgreSQL interacts with disks, contrasting PostgreSQL’s behavior with other databases. Covers hardware… | ||
| PostgreSQL: Ups, DevOps... | Aleksey Lesovskiy | 36m |
Practical checklist for automating PostgreSQL operations: automate OS/package/config management, replication setup, load‑balancer config and scripted failover… | ||
| Challenges of Using MySQL Efficiently on Modern Hardware | Aleksey Kopytov | 44m |
Talk examines MySQL performance issues on modern server hardware and workloads, naming MySQL and Percona Server. It covers typical bottlenecks caused by legacy… | ||
| Domain‑Specific Databases and Aviasales Mailings | Boris Kaplunovskiy | 45m |
Describes building a domain-specific data store for Aviasales to support notification emails when matching tickets appear. Uses a custom non-SQL database to… | ||
| Web‑Scale Backups for MySQL | Aleksey Kopytov | 50m |
Compares backup tools for large-scale MySQL (mysqldump, mydumper, mylvmbackup, Percona XtraBackup and commercial options) and discusses how to choose among… | ||
| 5 Reasons MongoDB Is the Leading NoSQL DBMS | Henrik Ingo | 52m |
Explains five reasons why MongoDB became a leading NoSQL database: it addresses massive growth in data and users, handles more complex and heterogeneous data… | ||
| The New MongoDB Sharding Architecture | Leif Walsh | 32m |
Explains MongoDB sharding vs replication: replication increases reliability and read scaling by copying data across more machines, while sharding splits data… | ||
| On-Disk Data Structures: B-Trees, LSM-Trees and Fractal Trees | Leif Walsh | 01h 16m |
The talk explains disk-based data structures used in databases and file systems, contrasting traditional B-trees with write-optimized structures like LSM-trees… | ||
| Partitioning and Data Migrations Using PostgreSQL as an Example | Denis Ivanov | 34m |
Practical techniques for partitioning and migrating PostgreSQL data: automatic distribution of rows across partitions plus support for reads, updates and… | ||
| Corpus collapsum: Galera Resilience to Partitioning | Raghavendra Prabhu | 01h 00m |
Examines Galera cluster behavior under network partitions, showing that despite CAP’s Partition tolerance goal, real-world distributed systems (including… | ||
| Postgres at the Core of Your Data Center | Bruce Momjian | 01h 00m |
PostgreSQL can serve as a central data-aggregator for multiple datacenters by using its extensibility, foreign-data wrappers to access external data sources… | ||
| Telemetry Analysis for Scaling | Theo Schlossnagle | 36m |
Circonus describes building a scalable telemetry pipeline: collecting, storing and analyzing high-volume metrics. It covers their architectural evolution… | ||
| Testing the Performance of Distributed Storage Systems | Aleksandr Kirov | 46m |
Covers testing performance of Software Defined Storage (SDS) systems: explains SDS architectures, cache layers, why sync is expensive, common bottlenecks and… | ||
| Turning OpenStack Swift into a High‑Load Storage | Nikolay Dvas | 46m |
Describes turning OpenStack Swift into a storage backend that handles heavy load by addressing its operational weaknesses. Covers using Swift in a provider… | ||
| How We Counted Traffic in Vertica | Nikolay Golov | 47m |
This talk describes how Avito implemented traffic analytics (unique visitors and sessions) on HP Vertica. It explains the challenge that unique counts are… | ||
| How We Store and Analyze a Large Social Graph | Maksim Bartenev | 44m |
Describes storing and analyzing a large social graph (100M+ nodes, 1B+ edges) while new data are continuously added. Covers using graph databases DEX and… | ||
| Ulmart: The Story of Its Creation | Dmitriy Zavalishin | 51m |
A retrospective on building Ulmart’s web platform: covers business requirements, project management and client collaboration methods, integration challenges… | ||
| Speeding Up and Offloading a Web Server by Transparently Caching on SSD | Stanislav Nikolov | 54m |
Describes developing and deploying an nginx module that transparently caches static files on SSDs to accelerate and offload a busy web server without changing… | ||
| Stylish Scaling | Valentine Gogichashvili | 01h 04m |
Zalando describes how PostgreSQL is deployed, tuned, monitored, and used as the primary data store and backup for real-time warehouse management and… | ||
| Advertising at the Speed of Light (DMP Platform) | Sergey Zhemzhitskiy | 44m |
A presentation about building a real-time user-profile enrichment service for RTB ad serving that must handle tens to hundreds of thousands of requests per… | ||
| Mail.Ru Editor, or Speed‑Reading in Half an Hour | Pavel Zinovkin | 44m |
Describes Mail.Ru Editor, a system for allowing users to edit Word/Excel/PowerPoint files online as part of Mail.Ru services. Covers the problem of users… | ||
| Building an Analytics System for a High‑Load Media Platform | O.Novikov, I.Saltanov | 38m |
Describes building an in-house analytics system for a high-traffic media/social site (Sports.ru) after it became a social network with 12M monthly uniques… | ||
| DIY Fault‑Tolerant Microcluster | Vitaliy Gavrilov | 40m |
Describes how to build a fault-tolerant microcluster using minimal hardware: L2 topology for component interconnects and L3 setup (HSRP, corosync) for external… | ||
| The "Cloud" at Badoo a Year Later: Fixing Our Mistakes | Yuriy Nasretdinov | 46m |
A retrospective on Badoo’s cloud-like platform covering its overall architecture (creation history, load distribution, fault tolerance), script log collection… | ||
| KVM Modifications for Clustered Operation | Andrey Shetukhin, Ivan Povsten | 39m |
Describes modifying KVM to support cluster-aware, fault-tolerant operation by enabling dynamic redistribution of virtual machines across physical servers… | ||
| How to Serve 60 Million Subscribers | Artem Rufanov | 01h 00m |
Design and implementation of a PCRF node (per 3GPP) to serve 60 million mobile subscribers. Covers building a real-time-capable decision engine that enforces… | ||
| Using Hadoop at Badoo | Valeriy Starynin | 52m |
Badoo describes using Hadoop to store and process the entire clickstream from the web site and mobile app servers—about 1 billion events per day—and an… | ||
| Running JS on the Server at Scale — What Can Go Wrong | Brian Geffon | 37m |
Describes problems encountered when moving rendering from server-side HTML to client-side JavaScript at LinkedIn, focusing on performance trade-offs for users… | ||
| Pushing PHP to Its Limits | Michael Armstrong | 50m |
LiteSpeed Web Server (LSWS) for PHP: explains three LSWS PHP modes and compares their trade-offs (memory use, execution speed, opcode caching/suEXEC behavior)… | ||
| Memcached Injections: They Exist and They Work | Ivan Novikov | 40m |
The talk examines injection vulnerabilities in the Memcached ecosystem: real bugs in client wrappers/drivers for languages including Ruby, Python, Java, PHP… | ||
| Highway to Continuous Integration | Denis Trifonov | 45m |
Uses Jenkins plus OpenStack and Chef to deploy isolated environments, runs PHPUnit tests in parallel with Paratest, and ties build results into JIRA. Starts… | ||
| Optimizing Software for Modern CPUs and Linux | Aleksandr Krizhanovskiy | 44m |
Covers practical techniques for optimizing software on modern CPUs and Linux, including data alignment, cache optimization and NUMA-aware design, CPU/thread… | ||
| How MySQL Replication Works | Andrey Aksenov | 43m |
Explains replication types (synchronous, asynchronous, semi-synchronous; master-master vs master-slave) and how MySQL implements them across versions. Covers… | ||
| Tools for High‑Load Projects: Caching and Queues | Vyacheslav Moskalenko | 39m |
Overview of RabbitMQ as a message broker: core concepts and how it works. Explanation of exchange types (Direct, Fanout, Topic) and their routing behavior… | ||
| Choosing a Search Engine with Your Head | Andrey Aksenov | 47m |
Practical guide to building search: how to implement a simple custom search in one hour, one evening, or one week and when that effort is justified. Compares… | ||
| Application Architecture Using MySQL | Petr Zaytsev | 47m |
Covers designing and reworking application architectures around MySQL, focusing on performance, high availability, security and scalability. Explains practical… | ||
| Anatomy of a Web Service | Andrey Smirnov | 49m |
An overview of what a backend/application server actually does, limits on load, and ways to improve performance. Covers concurrency approaches (threads… | ||
| Statistical Methods and Tools for System Performance Analysis | Aleksey Lavrenyuk | 36m |
Describes a method for performance testing when only real, highly seasonal traffic can be applied: identify input and output metrics, track them across… | ||
| Modern DoS Attacks and How to Defend Against Them | Aleksandr Vlasov | 45m |
Overview of modern DoS attacks that emphasizes application-layer threats which often evade network- and transport-level protections. Describes how operators… | ||
| Data Processing in RTB: Fast, Cheap and 98% Accurate | Pavel Kalaydin | 30m |
Covers real-time analytics for RTB (real-time bidding) at RuTarget, which processes a billion ad-impression requests per day. Describes using streaming… | ||
| Machine Learning in the MAIL.RU Advertising System | Igor Kretinin | 48m |
MAIL.RU’s advertising system uses machine learning on clickstream data stored in HDFS (about 100K records/sec) to train classifiers that segment Internet… | ||
| Client Applications Under Load | Andrey Smirnov | 46m |
Discusses how successful high-load projects require cooperation between client apps and servers, and examines what client-side developers must do when servers… | ||
| BigData Clustering Illustrated by OK Gifts | Artur Kadurin | 32m |
Processing Odnoklassniki gift-giving logs stored in a Hadoop cluster using a MapReduce computation model to filter data and compute a similarity measure… | ||
| How to Find a DDoS Attacker and Make Them Apologize | Dmitriy Volkov | 45m |
Covers practical steps for responding to an active DDoS attack: how to collect and preserve forensic evidence, methods to estimate financial damage, and ways… | ||
| Virtualization Games in JavaScript, or How I Rewrote an Emulator | Evgeniy Potapov | 24m |
A talk about emulators implemented in JavaScript: how they work, practical use cases, and the internal principles. It specifically covers the PCE.JS… | ||
| Escaping the 'Spam' Folder: How to Boost Newsletter CTR | Andrey Sas | 48m |
How to increase email newsletter CTR to avoid mail providers classifying messages as spam by improving engagement without harming overall site activity. Covers… | ||
| Wireless Access for Highload Conference Attendees | Aleksandr Pavlov | 26m |
Describes building wireless access networks for large events (exhibitions, stadiums) and why typical enterprise controller-plus-light-AP architectures don’t… | ||
| Thorny path to the Large Scale Graph Processing | Aleksey Zinovev | 39m |
Shows concrete scales and tooling: a Moscow road graph of 250k nodes, 460k edges, an EU road graph on the order of 18M nodes, 42M edges, and web/social graphs… | ||
| Running JS on the Server at Scale — What Can Go Wrong? | Brian Geffon | 37m |
Describes integrating a JavaScript engine into an HTTP proxy to run JS on the server side, and reports practical experiences from that integration. Covers… | ||
| Using Hadoop at Badoo | Valeriy Starynin | 52m |
Badoo describes using Hadoop to store and process the full clickstream from the website and mobile app servers—about 1 billion events per day—and additionally… | ||
| How to Serve 60 Million Subscribers | Artem Rufanov | 01h 00m |
Describes implementing a 3GPP-compliant PCRF node to serve 60 million mobile subscribers, making real-time policy decisions about users’ service rates based on… | ||
| How to Count and Analyze Hundreds of Gigabits of Traffic per Second | Stanislav Nikolov | 53m |
Describes collecting and analyzing hundreds of gigabits per second of client traffic using minimal extra servers for static content; addresses challenges when… | ||
| Scaling Microservices in Go | Matt Heath | 42m |
Describes migrating Hailo from monolithic PHP/Java to a cloud-native microservices platform deployed across three continents, implemented almost entirely in… | ||
| KVM Modifications for Clustered Operation | Andrey Shetukhin, Ivan Povsten | 39m |
Describes modifying KVM to support cluster-aware, fault-tolerant operation so virtual machines can be dynamically redistributed across physical hosts. Covers… | ||
| "The Cloud" at Badoo a Year Later: Fixing Our Mistakes | Yuriy Nasretdinov | 46m |
A retrospective of Badoo’s cloud-like infrastructure one year later, covering its overall architecture for load distribution and fault tolerance, plus how… | ||
| DIY Fault-Tolerant Microcluster | Vitaliy Gavrilov | 40m |
Describes how to build a fault-tolerant micro-cluster using minimal hardware: L2 topology for internal component links and L3 setup for external service… | ||
| Pushing PHP to Its Limits | Michael Armstrong | 50m |
The talk explains how LiteSpeed Web Server (LSWS) speeds up PHP-based dynamic websites compared with other web servers. It covers LSWS's performance advantages… | ||
| Building a Distributed Data Collection System with RabbitMQ | Alvaro Videla | 01h 19m |
How to build a distributed data-collection system using RabbitMQ to gather events from geographically dispersed servers and devices and replicate them into a… | ||
| Developing an Analytics System for a High-Load Media Service | Oleg Novikov, Ilya Saltanov | 38m |
Description of building a custom analytics system for Sports.ru after it became a social network with 12 million monthly users. Covers collecting and… | ||
| Mail.Ru Editor, or Speed-Reading in Half an Hour | Pavel Zinovkin | 44m |
Mail.Ru built an online document editor to let users open and edit Word/Excel/PowerPoint files in the cloud instead of downloading them. The talk explains the… | ||
| Advertising at the Speed of Light: DMP Platform | Sergey Zhemzhitskiy | 44m |
Describes building a high-performance DMP/RTB advertising platform that serves incoming ad-requests at tens to hundreds of thousands QPS with sub-20 ms… | ||
| Stylish Scaling | Valentine Gogichashvili | 01h 04m |
Zalando’s use of PostgreSQL as the core datastore for everything from customer records to article data and real-time warehouse systems is described, covering… | ||
| Speeding Up and Offloading a Web Server by Transparently Caching on SSD | Stanislav Nikolov | 54m |
Describes developing and deploying an nginx module that provides transparent SSD-based caching to accelerate and offload I/O on overloaded web servers. The… | ||
| Ulmart: The Story of Its Creation | Dmitriy Zavalishin | 51m |
A retrospective on building Ulmart’s web platform for Russia’s largest online electronics retailer, covering business requirements, project management methods… | ||
| Postgres at the Heart of Your Data Center | Bruce Momjian | 01h 00m |
PostgreSQL is presented as a data-aggregation engine for data center environments, demonstrating its extensibility, foreign-data wrapper access to external… | ||
| Telemetry Analysis at Scale | Theo Schlossnagle | 36m |
Circonus describes collecting, storing and analyzing large-scale telemetry and the architectural evolution they used to scale that pipeline. The talk covers… | ||
| 5 Reasons MongoDB Is the Leading NoSQL DBMS | Henrik Ingo | 52m |
The talk explains five reasons why MongoDB became a leading NoSQL database by addressing Big Data challenges: rapid growth in data and users, increasing… | ||
| Corpus collapsum: Galera's Resilience to Partitioning | Raghavendra Prabhu | 01h 00m |
Examines Galera cluster behavior under network partitions, discussing how real networks fail and how Galera implements (or fails to implement) partition… | ||
| CREATE INDEX … USING VODKA VODKA CONNECTING INDEXES | Oleg Bartunov, Aleksandr Korotkov | 40m |
Prototype of a new PostgreSQL access method called vodka access method implements composable index layers (radix trees plus existing methods) to index jsonb… | ||
| PostgreSQL: Ups, DevOps... | Aleksey Lesovskiy | 36m |
Explains practical limits of automating PostgreSQL ops: use Ansible or Chef/Puppet/Salt for OS/config distro and repetitive tasks, but keep human review for… | ||
| SSDs for Your Database | Petr Zaytsev | 49m |
Overview of Flash (SSD) options for databases: explains different Flash technologies available, their pros and cons, and selection criteria to avoid… | ||
| Web-Scale Backups for MySQL | Aleksey Kopytov | 50m |
Discussion of practical strategies for backing up high-scale MySQL installations, comparing tools such as mysqldump, mydumper, mylvmbackup, Percona XtraBackup… | ||
| Asynchronous Replication Without Censorship | Oleg Tsarev | 41m |
A Percona Server developer and Mail.Ru Target engineer examines MySQL’s asynchronous replication, identifies why past design choices failed, and explains… | ||
| Deep Dive into On-Disk Data Structures | Leif Walsh | 01h 16m |
The talk explains why traditional on-disk B-tree structures (used in DBMSs and file systems) are being replaced by write-optimized on-disk structures such as… | ||
| Domain-Specific Databases and Aviasales Mailings | Boris Kaplunovskiy | 45m |
Describes building a domain-specific datastore and delivery pipeline for Aviasales email/SMS notifications: users subscribe with multi-predicate filters… | ||
| How PostgreSQL Interacts with Disk | Ilya Kosmodemyanskiy | 45m |
Explains why databases need disk storage and how PostgreSQL interacts with disks, covering disk vs. memory vs. CPU trade-offs and why administrators prioritize… | ||
| How We Counted Traffic on Vertica | Nikolay Golov | 47m |
Describes how Avito measured web traffic and computed detailed analytics (unique visitors and sessions) from hundreds of millions of events per day using… | ||
| How We Store and Analyze a Large Social Graph | Maksim Bartenev | 44m |
Describes a system for storing and analyzing a very large social graph (over 100 million nodes and over 1 billion edges). Covers use cases from traversing a… | ||
| How to Account Money Correctly in a Database | Roman Druzyagin | 52m |
The talk examines how to store and process money correctly in application databases, addressing common developer mistakes when designing billing and accounting… | ||
| Turning Openstack Swift into Storage for High Loads | Nikolay Dvas | 46m |
Describes turning OpenStack Swift into a production-grade, high-load object store by addressing its performance and reliability issues. Covers Swift deployment… | ||
| A New Sharding Architecture for MongoDB | Leif Walsh | 32m |
An overview of MongoDB sharding and replication trade-offs: how replication improves read scalability and reliability by copying data, while sharding improves… | ||
| Partitioning and Data Migrations Using PostgreSQL as an Example | Denis Ivanov | 34m |
Covers partitioning and data migration techniques for PostgreSQL, including automatic data distribution across partitions and performing reads/updates/deletes… | ||
| Payment Processing at Badoo — An Inside Look | Anatoliy Panov | 43m |
Describes Badoo’s billing and payment systems for a high-traffic international service, focusing on architecture, operational problems encountered as volumes… | ||
| Testing Performance of Distributed Storage Systems | Aleksandr Kirov | 46m |
Covers performance testing of Software Defined Storage (SDS) systems by examining SDS architectures, cache layers, the high cost of sync operations, common… | ||
| The Design of a Modern Distributed Object Storage | Aleksandr Chistyakov | 31m |
A description of how to move from vertical to horizontal scaling for large amounts of static user content by adopting distributed Object Storage with an… | ||
| People Make the Difference or Video Streaming at Odnoklassniki | Aleksandr Tobol | 24m |
Describes how the Odnoklassniki team sped up video start times by more than 10× using adaptive streaming technologies MPEG‑DASH and HLS. Explains which… | ||
| Continuous Delivery, or the Magic Button for On‑Demand Releases | Denis Yakovlev | 38m |
Describes building a one-button Continuous Delivery workflow for 2GIS Reference API so a manager can trigger a full production release on demand. Covers using… | ||
| ObjectManager, or How to Work with a Large Number of Objects on a Map | Marina Stepanova | 30m |
Describes ObjectManager, a system for efficiently handling and rendering very large numbers of geographic points on a map. Covers client-side rendering… | ||
| Application‑Level Protocols in the Browser | Ilya Kutukov | 44m |
Examines implementing custom application-layer protocols inside web browsers to better use bandwidth, enable arbitrary encryption, manage session state, and… | ||
| DDoS Attacks in Russia in 2014: UDP Packets Don’t Smell of Lard | Aleksandr Lyamin | 48m |
An analysis of DDoS attacks in Russia during 2014, using ten months of telemetry to describe current attack trends and correlate them with geopolitical and… | ||
| Docker & Puppet: How to Crossbreed Them and Whether You Need It | Anton Turetskiy | 44m |
The talk explains how Docker containers and Puppet configuration management can be integrated: it gives a brief overview of Docker and Puppet, reasons for… | ||
| Tempesta FW: FrameWork and FireWall for WAF and DDoS Mitigation | Aleksandr Krizhanovskiy | 40m |
Tempesta FW is an open-source hybrid HTTP accelerator and firewall designed for application-layer DDoS protection and building high-performance Web Application… | ||
| Heterogeneous Services for Highload Projects | Igor Myzgin | 35m |
Discusses optimizing code and infrastructure for high-load web projects in Russia where ad-based monetization and company valuation differ from Western models… | ||
| Scalable nginx Configuration | Igor Sysoev | 44m |
Covers techniques for building long-lived, maintainable nginx configurations, showing common mistakes and correct vs incorrect approaches. Discusses when to… | ||
| Development Management Tooling at Mail.Ru Group | Aleksandr Gornyy | 45m |
Describes Mail.Ru Group’s development toolset and how the company manages engineering work using Atlassian and related products: JIRA for issue tracking… | ||
| Challenges of Efficiently Using MySQL on Modern Hardware | Aleksey Kopytov | 44m |
Covers performance issues when running MySQL on modern server hardware and workloads, naming MySQL and Percona Server as the technologies. Explains typical… | ||
| Sharding in MongoDB | Henrik Ingo | 47m |
Explains how MongoDB’s built-in sharding provides horizontal scalability by distributing disk, RAM and CPU load across a cluster, with automatic sharding and… | ||
| Working with Indexes — Best Practices for MySQL 5.6 | Petr Zaytsev | 01h 05m |
Practical guidance on using indexes in MySQL 5.6: how the MySQL 5.6 optimizer uses indexes to execute queries, how to design an optimal indexing strategy, when… | ||
| nothing matches | ||