Track: Bare Knuckle Performance

10:35am - 11:25am

Scaling Up Performance Benchmarking

Performance scaling along with responsiveness is very important to both on-premise as well as cloud deployments. Often, deployments can be very simple to an extremely complex web of services where there is no single magic bullet for performance scaling. As part of the development of a very complex benchmark, we learned many lessons related to scaling up, scaling out and responsiveness. We would like to share application architecture decisions, observations points, problem analysis methodologies etc. which can be applied when architecting, deploying and analyzing real production applications. We will share how application level, as well as system level metrics, change for various application architecture decisions as well as high-level JVM analysis to correlate tail latencies.

11:50am - 12:40pm

Performance Open Space

1:40pm - 2:30pm

“Quantum” Performance Effects: Beyond The Core

CPU microarchitecture is hidden from developers by compilers, virtual machines, etc. Should developers know everything about CPU microarchitecture? It looks like learning quantum mechanics for cooking. Recently we heard how some microarchitecture features could lead to security breaches like Spectre/Meltdown. In this session, we won't talk about vulnerabilities. We will talk how (and how big) CPU microarchitecture details may have an influence on applications performance. Could it be visible by end-users? How to avoid misjudgment when we estimate our code performance? CPU is huge (not in size) that is why the talk is limited to those parts which located out of computational core (mostly caches and memory access).

2:55pm - 3:45pm

The Trouble With Memory

It is our experience that 60-70% of all the applications we analyze incur a performance penalty due to how it consumes and/or retains memory. Yet if you ask people what is their biggest performance headache it’s unlikely that they’d recognize memory as being #1. I have been involved in analyzing numerous applications for just these kinds of issues and in this session, I will cover the tell tale signs and symptoms that your application is one of those suffering from some sort of memory issue.  I'll discuss what steps you can take to cure this problem and will also cover how the JVM can both help reduce the memory ~strength of your application.  What should you do?"

4:10pm - 5:00pm

Tuning a Runtime for Both Productivity and Performance

Developers generally enjoy increased productivity when working in a managed code environment. What might not be obvious to them is the amount of investment in the managed code’s Virtual Machine to enable both productivity and high performance. Microsoft’s .NET team would like to share how they designed the runtime environment to balance convenience, fast startup, serviceability, low latency, high throughput. For examples services such as JIT compilation, TypeSystem, garbage collection all provide convenience but come at a cost. The challenges presented are common to many environments – you do not need to be a .NET developer to attend this talk.

5:25pm - 6:15pm

Brewing Java Applications in Sigma Managed Clusters

In handling the largest online shopping event in the world, the emergence of large-scale software deployments in the data center has led to many challenges. We will cover two of them: (1) evaluating and estimating software performance at scale, and (2) optimizing software for resource management. This presentation addresses the two challenges by bringing the knowledge of software performance monitoring in the data center to the world of applying performance analytics. We will describe how effective decisions can be made in a case study.