Percona Live 2017 Open Source Database Conference

April 24 - 27, 2017

Santa Clara, California

Slides from Percona Live

ProxySQL is a very powerful platform that allows us to manipulate and manage our connections and queries in a simple but effective way. Historically MySQL lacks in sharding capability. This significant missing part had often cause developer do implement sharding at application level, or DBA/SA to move on to another solution. ProxySQL comes with an elegant and simple solution that allow us to implement sharding capability with MySQL...
ProxySQL is a flexible, fast and highly configurable piece of software that provides dramatic improvement to your cluster and high availability (HA) architecture. The presentation will discuss how to deploy ProxySQL in support of architectures using standard master/slave, Percona XtraDB Cluster/MySQL-Galera or MHA solutions. The presentation discusses details on: - ProxySQL host group - Using balancing and failover on standard...
MariaDB Server 10.2 has brought two new important querying features, Common Table Expressions (CTEs) and Window Functions. Both features provide, greater expressibility to queries, which leads to more opportunities for the optimiser to provide speedups With a focus on analytics, this talk will go through a number of examples on how to transform long complicated queries into simple, maintainable ones. With smart use of window functions,...
MariaDB Server has had histograms since MariaDB 10.0, now, MySQL 8.0 will have them too. This session will present an overview of histogram implementations in MariaDB, MySQL 8.0, and we will also take PostgreSQL for comparison. We’ll cover everything about histograms: - Why do query optimizers need histograms - What are the costs of collecting and maintaining a histogram in each database - How the query optimizers use histogram data...
MyRocks is a new storage engine based on RocksDB. It is included in MariaDB Server, and together the combination is called MariaRocks. This talk will describe what one gets when MyRocks’ storage efficiency is combined with MariaDB’s features: - Optimistic parallel replication with MyRocks - Virtual and Dynamic Columns with MyRocks - MyRocks’ Read-Free Replication with MariaDB We will also cover household items like - How MariaDB...
We’ve all used or heard about various filesystems like ext3, ext4, or XFS but what about copy-on-write filesystems like btrfs or ZFS. With XFS being the de facto standard to host MySQL on, learn why Pardot chose ZoL (ZFS On Linux) instead to host their highly transactional MySQL database servers in production on. ZoL is a port of OpenZFS to Linux that was released in 2013. However, since then there has been a lot of FUD about its stability...
GitHub has been using gh-ost for online schema migrations since before it was open sourced in August of 2016. Once gh-ost was in place helping our migrations run more smoothly, the Database Engineers didn't want to have to be involved in running them anymore. In this session we'll discuss the steps we took to remove humans from most of the steps of running a migration. This includes implementing chatops with Hubot, leveraging...
With MongoDB 3.2 WiredTiger is now the default storage engine. This presentation will go through the configuration parameters when using MongoDB with WiredTiger and demonstrate performance benchmarks against different configuration values. Running WiredTiger with default settings is fine for most workloads but some deployments may require changes to the default settings and especially the WiredTiger core engine configuration variables...
Percona Monitoring and Management (PMM) is a wonderful solution, it aims to provide out of the box monitoring and management for your MySQL (and/or MongoDB) environment. By design it's delivered in just two distinct modules, a Docker container for the server components, and a package/tarball for the client components. This session will take a look at the applications used within PMM, such as Prometheus, Grafana, and Percona's...
In this case study, we will present Yahoo's backup and recovery strategy. We will present How to optimize backups of large amounts data using full, incremental, compress, parallel options. How to get the consistent backup using slave-info, safe-slave-backup, no-lock How to automate slave rebuild using Xtrabackup How to perform point-in-time-recovery How to monitor and validate the backups How to use multi-source replication to...
Yelp has grown rapidly over the last several years, and we were forced to quickly scale our infrastructure to keep up. While every company experiences growth in a different way, there are some realities that all growing and successful organizations will face: geographical expansion will become a requirement, launching new database clusters will become vital, and all hosts will become replaceable. You must have automation and a powerful...
Topic: MongoDB in AWS (MongoDB as a DBaaS ) The problem: How to automatically build self-healing mongo cluster specified by user on AWS. The solution: Using AWS CFN template and Service Catalog to deployment Mongo clusters to achieve: . Flexibility (specify nodes per shards and how many shards in cluster). . Auto-build MongoDB clusters. . Auto-recovery from resource failures....
How do you correctly direct traffic to your databases? How do you add new replicas or remove broken or retired replicas from the traffic flow? How do you scale elastically? Every company has their own way of how they solve the service discovery problem. Some use solely DNS, others use VIPs, Consul, …..etc. At Yelp we’ve had great success using SmartStack for service discovery in most of our infrastructure, and we wanted to use it for...
Come on a journey with Patrick as he walks you through his transformation from a PostgreSQL DBA to a MySQL DBA. This talk will be a story of Patrick's journey where we'll walk through some of the major differences between PostgreSQL and MySQL and how we use both of them at Yelp. We'll talk about query plans and the query optimizer, index types and their structure, and the issues around schema changes; highlighting some...
Even if InnoDB is universally recognized by DBAs as "the" engine to use for production data, there are use cases where InnoDB just does not fit. In this talk we will introduce the most popular engines available for MySQL to date, focusing on the limitations that plague InnoDB when it's about inserting rows at high rate and/or using a random primary key, like an UUID. We will describe how we worked around this limitations...
There are substantial improvements in the Optimizer in MySQL 8.0. Most noticeably, we have added support for advanced SQL features like common table expressions, windowing functions and grouping() function. We also made DBAs’ life easier with invisible index, and additional hints that can be used together with the query rewrite plugin. On the performance side, cost model changes will make a huge impact. We have made JSON support even more...
Never used Percona XtraDB Cluster before? Are you a seasoned veteran, wanting to see the current hot stuff?Come join this full-day tutorial where we will introduce you to the concepts of a fully functional Percona XtraDB Cluster. The first half of the tutorial will cover the basics of installation, bootstrap, IST, SST, Replication, Certification, DONOR selection, common-failure situations, and the garbd. The second half will cover...
Never used Percona XtraDB Cluster before? Are you a seasoned veteran, wanting to see the current hot stuff?Come join this full-day tutorial where we will introduce you to the concepts of a fully functional Percona XtraDB Cluster. The first half of the tutorial will cover the basics of installation, bootstrap, IST, SST, Replication, Certification, DONOR selection, common-failure situations, and the garbd. The second half will cover...
Postgres Logical Decoding feature has been added in 9.4 and it is continuously improving thankfully to its vibrant community. In this talk we are going to walk through its concepts, usage and some of the new things up to come in further releases. Logical Decoding is one of the features that are under the BDR implementation, allowing bi directional streams of data between Postgres instances. Also, it allows to stream data outside...
Yandex team has built one of the best opensource databases for analytics. It's fast, capable for storing petabytes of data and supports SQL. This talk is ClickHouse overview: features and benchmarks, plans and statuses, use cases and real users feedback. I will start with current ClickHouse status in DBMS market: how many users, what's the community size, and what's going on in contributing. In the main part I will...