Recent Changes -

747SOnTreadmillsExperiencesScalingUptimeAndManagement

Sysadmin.747SOnTreadmillsExperiencesScalingUptimeAndManagement History

Hide minor edits - Show changes to markup

January 11, 2009, at 05:56 PM by verne - Just copying and pasting from sysadmin.miniconf.org, since it's linked from the schedule
Added lines 1-19:

747s on Treadmills: Experiences scaling uptime and management in the face of the Thundering Masses - Matt Moor

In 2007, Bulletproof Networks got the opportunity to host a major donations drive for a notable Australian charity. Hilarity ensued as we realised just how much load a PHP driven website could generate.

In 2008, we decided a little more forethought and a little less melting might be a good idea. Over a period of three months, we designed and redesigned the architecture, planned and tested our assumptions and debated with the developers.

Through a combination of VMware, really large boxes, and lots of open source software we've managed to survive hundreds of thousands of donations, tens of megabits of traffic, and multiple major hardware failures. All in six weeks.

In this talk, I'll talk about what we did right, what we did wrong, and the realities of scaling PHP web applications without significant developer time.

Bio

Matt Moor is a Senior Architect at Bulletproof Networks, a Managed Hosting Provider, and has been using and working with Open Source for around 10 years, 7 of which as a professional sysadmin.

Matt spends his time designing, building and troubleshooting networks based around IPTables, LVS, Apache, Squid, VMware and Cisco. Occasionally, people make him do stuff with Enterprise applications, too.

Matt has been involved in the Open Source community since 2001, and was on the committee of the Sydney Linux Users Group from 2005 - 2008. Matt was also heavily involved in building and operating the LCA2007 network.

Length: 50 minutes

Page last modified on January 11, 2009, at 05:56 PM