PerfBytes

Back when I grew up in Minnesota we played a game called "Duck, Duck, Gray Duck where one my friends would run around the circle of all us kids and tap us on the head - around and around. It's basically the same way a load balancing algorithm works to enable horizontal scalability of large systems. Except that the load balancer would not randomly tap one of us and say Gray Duck and then go sprinting around to take our spot. That's not actually how node eviction works in clustering. In this episode of PerfBytes we are joined by our old friend Alon Girmonsky from Blazemeter to discuss the common challenges with modern load balancers; both software and hardware types. Dan Bartow gives a rant about the past 7 years of bad load balancer performance. And James introduces a whole new type of pre-outage performance outage on News of the Damned. This episode of PerfBytes is formally sponsored by www.flood.io.

It's different for mobile. The variable resources like network and cpu, the limited space and screen size all force you to get your game on for mobile application performance. In this episode Ophir Prusak from Blazemeter.com joins us to discuss the bits that fly off your application as the rubber of your fat and lazy code hits the hard pavement of the device. James tries to laugh like Vincent Price and Mark reminds us about the European restrictions protecting nuisance birds. This episode of PerfBytes is formally sponsored by Flood.io (www.flood.io).

It's time for PerfBytes LIVE! Please join us again at STPCON in October 2013 for a live broadcast recording of the PerfBytes podcast featuring your wildly banal hosts Mark Tomlinson and James Pulley. We are celebrating the one year anniversary of the PerfBytes podcast and this special episode will be broadcast LIVE from the STPCON conference and will include our coverage of the STPCON news, the performance year in review, the News of The Damned, industry headlines, extemporaneous speaking from actual performance testers and live Q&A with the members of the audience. Attendees in-person will be entered into a drawing to win a pair of PerfBytes shoes! As supporters of the Software Test Professionals performance testing community, the PerfBytes podcast is powered by creative community-driven contributions and the overwhelming desire to improve performance testing practices
across our industry. And pancakes.

Look in the mirror. Do you see someone who loves BBQ more than showering? Do you see someone wearing a wool cap and jacket just to keep warm while working in the datacenter? Can you say the word "No!" to a stakeholder? Do you know how many seconds there are in 1 day? If you answered yes to any of these questions, there's a good chance that you are a performance tester. In this episode we explore the performance personality in the style of comedian Jeff Foxworthy. We chat also chat about Art Bell's underwater pyramids, supercomputers in Japan, bitcoins encrypted by drug dealers, the lessons learned (or not) by Healthcare.gov and of course the News of the Damned wouldn't be complete without a repeat crash of an event ticket website in the UK.

Performance left to the end of the project? Only two weeks on the schedule for doing load testing, regardless of the complexity or requirements? No qualified performance engineering talent assigned to the team? Sounds like you need a class called: Performance for Project Managers 101. This episode will cover all the common anti-patterns for performance project planning; scheduling, resources, tools, environments and objectives.

What's the one thing in common with both hard drives and pizza? In this episode of PerfBytes we discuss yet another essential physical resource common to computer performance: the disk! Whether it's local hard drives, network attached storage, nfs, SAN array or flash SSD - the disks are a common bottleneck that can leave your application in a lurch; sluggish, slow and lifeless.

He's dead Jim. In the last 2 weeks of August 2013 there have been a nearly unbelievable number of system-wide outages at some of the world's largest internet service providers. Microsoft, Amazon, Google...almost everyone except Twitter found their systems lying face down in the mud recently. Even the NASDAQ and NYSE struggle to keep their shared server running in their "Neutral Zone" putting nearly six trillion trades at risk. Techno-guru John Curtis joins us for this entire episode dedicated to the News of The Damned. Satan admits to not being a hipster.

Have you seen my car keys anywhere? The older we get the more we might tend to forget about the exciting topic of performance tuning for memory. We're talking about garbage collection, memory leaks and caching strategies. Special guest Scott Moore from Northway Solutions joins us for this whole episode from his home studio in Nashville, TN. The famous newborn Prince George is already crashing websites just as his mom does, as we cover Swaddle-geddon on the News of the Damned. James says the word kerfuffle, twice. Seriously, I swear I left my car keys right here in the studio on the desk.

Continuous Deployment, plus Continuous Release might not quite equal Continuously High-Performing apps! As you might have heard, the developers and the ops guys got together and decided to start sharing tricks and secrets. They even figured out how to start doing each others' jobs! But where does performance fit in this highly-automated new paradigm for application development and release? Do you know what kind of beverage fits with DevOps? Do you have a tattoo?

Bellman asked Ford, "Do you know the way to San Jose?" More importantly, do you know how many hops it takes for your packet to get there? In this episode of PerfBytes we'll focus on the infrastructure that allows all the different components of your system to work together: the network. Understanding the essential sources of latency for what makes your network transfer slow, what causes congestion and how to measure the health of your network during a load test. Ping!