Percona Live 2017 Open Source Database Conference

April 24 - 27, 2017

Santa Clara, California

Slides from Percona Live

Input-output performance problems are on every day agenda for DBAs since the databases exist. Volume of data grows rapidly and you need to get your data fast from the disk and moreover - fast to the disk. For most databases there is a more or less easy to find checklist of recommended Linux settings to maximize IO throughput. In most cases that checklist is good enough. But it is always better to understand how it works, especially if you...
When your database is growing, you definitely need to think about other techniques like database sharding. SPIDER is a MariaDB Server / MySQL storage engine for database sharding. Using SPIDER, you can access your data efficiently across multiple database backends. In this time we will introduce the following things. 1. why SPIDER? what SPIDER can do for you? 2. when SPIDER is right for you? what cases should you use SPIDER? 3. how...
Uber made waves in 2016 by publishing a paper describing their reasons for moving from PostgreSQL to MySQL. So, were they right? Mostly, no. We'll talk about the points they raise about PostgreSQL vs MySQL, and how they could have stayed on PostgreSQL while avoiding or mitigating the problems they faced... and honestly talk about how PostgreSQL can improve to meet their criticisms.
Clients are implementing changes at a fast rate, putting extra demand on us, the humans behind the task. How do we reduce human errors, while keeping up with the demand and still provide value? Would you get into a plane if the pilot didn’t check the aircraft? Ever had surgery? How do you know the surgeon didn’t leave a sponge behind? Checklists! While not glamorous or fancy, Human Reliability checklists will reduce human errors when...
Percona Server for MongoDB is fully compatible with MongoDB Community Edition, and has some cool extra features (some of which are available only in MongoDB's Enterprise Edition). The features include: - Two extra storage engines: Percona Memory Engine and MongoRocks (based on RocksDB) - Audit - External authentication - HotBackup for WiredTiger and MongoRocks - Rate limiter extension The presentation is made by a Percona...
Overview: There are a number of built-in MySQL commands and third party tools out there for generating data to help tune and diagnose your servers. All that information is great, but not always easy to parse through; especially if you are in the middle of an emergency. What do you do when you need to see only the non-sleeping processlist threads, pick out the top 5 longest currently running queries, or show the oldest open transactions...
There are challenges in taking an existing infrastructure from a data center to Google Cloud. It is an understatement to say that there are plenty of differences between the two. For our in-house setup, VitalSource uses MHA with MHA-Helper managing the VIP for us. In GCE, we needed to separate the access management from the topology manager. In this lecture, we will discuss the solutions that we reviewed and tested, such as: ProxySQL,...
gh-ost is a new tool by GitHub which changes the paradigm of MySQL online schema changes, designed to overcome today's limitations and difficulties in online migrations. gh-ost is: - Triggerless: no triggers placed; - Pausable: can suspend master writes altogether - Lightweight: makes a low impact on the master database - Controllable: one can interact with an executing gh-ost process, get info and reconfigure parameters -...
MongoDB database size will grow larger and larger. This talk will cover several ways to control MongoDB data size: 1 Use wiredTiger storage Engine 2 Expire Data from collections by Setting TTL 3 Manually purge data and rebuild Mongo databases 4 Automatically purge data
MySQL® is an extremely popular and widely used RDBMS. Apache® Ignite™ is the leading open source in-memory computing platform which can provide speed and scale to RDBMS-based applications. Apache Ignite is inserted between existing application and data layers and works with all common RDBMS, NoSQL and Hadoop® databases. Join Denis Magda, GridGain Product Manager, and Peter Zaitsev, CEO and Co-Founder of Percona, as they discuss how you...
This is a tutorial about MyRocks: RocksDB Storage Engine for MySQL (https://github.com/facebook/mysql-5.6). MyRocks is a project to utilize space and write optimized LSM database with MySQL. MyRocks supports lots of useful features like below. - Less than half space compared to compressed InnoDB - Much smaller write amplification compared to InnoDB - Reverse order index so that descending scan can be a lot faster - Crash safe slave...
The CAP theorem points to unavoidable tradeoffs between consistency and availability when the network can partition. This decision heavily impacts system performance and cost. Current database design forces application developers to decide early in the design cycle, and once and for all, where they sit in this spectrum. At one extreme, strong consistency, as in Spanner or CockroachDB, requires frequent global coordination; restricting...
Orchestrator is a MySQL topology manager and a failover solution, used in production on many large MySQL installments. It allows for detecting, querying and refactoring complex replication topologies, and provides reliable failure detection and intelligent recovery & promotion. This session walks through orchestrator setup, deployment and usage best practices. We will focus on major functionality points and share authoritative...
In this session we will share our experiences using gh-ost for long-running schema changes in a multi-tier replication topology under high-volume traffic. This will include sharing how gh-ost can work with topologies using the Tungsten Replicator. We’ll share the ideal combination of commands to use in this setup, and discuss reasons we chose gh-ost over our old friend pt-online-schema-change. Operational controls are important with long-...
MySQL query tuning is often more art than science and it can quickly eat up a lot of time. In this presentation, Janis Griffin, performance evangelist and 20+ year database expert, will share her secrets for determining the best approach for tuning queries by utilizing the performance schema (specifically instrumented wait events and thread states), query execution plans, SQL diagramming techniques and more. Regardless of the complexity of...
This talk presents the updated information around current MongoDB backup options. We will discuss how MongoDB dumps work, different options to use for those dumps, binary backup options, the latest Percona Labs backup tool, new hot backup features, and how Mongo Ops Manager works. Additionally, we will cover updates and changes due to the 3.4 world, what features are and are not in the Percona Labs tool, official support status and...
We will cover what new features are in both MongoDB Community 3.4 and Percona Server for MongoDB 3.4. We will talk about the new graph functions, parallel balancing, sharding improvements, MongoDB zones, and many more new or improved areas of MongoDB. This talk is to give the community a lay of the land on what is working, and what might be good to wait to use. If you are new to MongoDB or have not yet switched to 3.4, this will be...
MyRocks is a popular open source storage engine for MySQL, built on top of RocksDB storage library. For the past few years, Facebook spent plenty of time on it, to make it better and well-known, and claimed that MyRocks provides much better compression ratio and much smaller write amplification. rumor or truth? We(Alibaba Database Team) spent a whole year on RocksDB and MyRocks: do thorough tests, understand the implementation, make...
Last year, we joined percona live the very first time and bring the world AliSQL, a MySQL branch maintained by Alibaba Database Team. We share many improvements we made on AliSQL, including Column Compression, Double Redo Buffer, "Hot SKU", SQL Firewall, and so on. Some of these improvements have been absorbed by the Mainstream. Nevertheless, this is not the end of the story of AliSQL, this year we made even more breakthrough...
When you upgrade to a new environment, you may wish to test each of the unique queries that enter the database for a safe upgrade, but how can you collect a complete set of unique queries on a heavily loaded database? When you stay on the same environment, you may want to know the performance of each of the unique queries for tuning purpose, how do you collect the data. In this lightning talk, I would like to introduce the concept and a...