While MySQL systems power a lot of the world, how much do you really know about how they are tested? This session will provide an overview of several functional testing tools available for MySQL-based systems, with a focus on mysql-test-run, the random query generator (randgen), and dbqp / kewpie.
Topics will include the history, function, and usage of each system with the goal being to both inform attendees about how their favorite dbms is tested and to show them new ways to test their own systems.
Not a DBA but your work thinks you are? This is the session for you. We will give you a crash course into the most valuable variables, places to look, and gotcha's. This is not designed to replace traditional training but rather to ensure you at least have the most basic skills to ensure your foray into the world of MySQL DBA's gets off on a good foot.
In the beginning, all websites start from a single data store, but the abundance of open source solutions today make it apparent that read and write bottlenecks are quickly reached and some form of scaling is eventually necessary. The two extreme choices are to vertically scale and hope that hardware outpaces your capacity requirements, or to horizontally scale by sharding the data so that additional capacity is simply a matter of adding shards.
Some indexing structures are write optimized in that they are better than B-trees at ingesting data. Other indexing structures are read optimized in that they are better than B-trees at query time. Even within B-trees, there is a tradeoff between write performance and read performance. This talk will present the basics of write optimization, comparing different approaches and what the differences mean for end Users.
New Relic is the largest SAAS based Application Performance Monitoring provider. We have 9 MYSQL shards that collectively handle million data points per second. Our original disk subsystem was 15k HDD drives in a RAID-5 configuration. As our site load grew, the disk subsystem was starting to behave less predictably and was becoming a bottleneck. After several months of research and testing we decided to make the plunge to SSD.
This talk will chronicle our experience moving from HDDs to SSDs and will include:
As MySQL powers more and more business critical applications it is very important to be able to upgrade MySQL server with as low risk as possible. In this presentation we will go into details of best practices used by Percona to perform hundreds of MySQL upgrade projects as well as learn how to use Percona Toolkit to assist in simplifying those operations.
Now it is well known that MySQL can be used as a extremely fast NoSQL database (reference: http://yoshinorimatsunobu.blogspot.com/2010/10/using-mysql-as-nosql-stor...). NDBAPI and HandlerSocket are popular NoSQL interface, but for many application users "NoSQL to InnoDB with Memcached" will probably be the most easy-to-use one because you can access to MySQL via very popular memcached protocol (reference...
We will present results from benchmarking a MySQL Galera cluster under various workloads and also compare them to how other MySQL high-availability approaches perform. We will also go through the different ways you can setup Galera, some of its architectures are unique among MySQL clustering solutions.
* MySQL Galera
** Synchronous multi-master clustering, what does it mean?
** Load balancing and other options
** WAN replication
** How split brain is handled
** How split brain is handled in WAN replication
pt-table-checksum is a free tool which safely and efficiently verifies that data on MySQL replicas is identical to the master. The tool simplifies the complex task of accurately comparing live data without adversely affecting the servers. This session will demonstrate basic usage, discuss important dos and don'ts, and highlight certain technical issues. Those new to pt-table-checksum will learn how to get started confidently and correctly. Those familiar with pt-table-checksum 1.0 will learn the new features and enhancements of 2.0 while it is demonstrated.
Advanced users of MySQL replication have reason to be excited about MySQL 5.6 improvements. There are, in fact, many features that users have been waiting for since the inception of MySQL.
There are features that have been available for some time in MySQL 5.5, like semi-synchronous replication and heartbeat, which aren't as known as they should. We will cover these for starters.
Interested in looking at database performance issues from application or the client's side of things? The New Relic product is particularly well suited for this type of investigation. In fact, Percona and New Relic have a partnership whereby Percona provides a license to use New Relic to its paid support customers. During this talk I will show how to drill into database performance issues from application's perspective.
This talk is for those who want to know the technical what, when, where and why of choosing a search engine for SQL/NoSQL database solutions. It will assist individuals and businesses in developing a technical framework to answer the typical questions that one encounters when choosing a search option. Such as, do I need/want to add a search component to my architecture, how should one go about choosing a search engine and when does a commercial or open source search engine make sense for me/my business?
Non-blocking or event-driven programming is in high fashion today. Using non-blocking technology, it's possible to write single-threaded daemons that serve high concurrency workloads and interact with slow backend/external systems (e.g., databases and webservices).