A playground for experimenting with different server models.
We're using the following physical machines to perform these tests:
| Name | OS | Role | CPU | RAM | NIC | Notes |
|---|---|---|---|---|---|---|
| perfsvr | Windows Server 2012 R2 | Web Server | Xeon E5-1650 | 32 GB | Intel® Ethernet Converged Network Adapter X540-T1 10GbE | |
| perfsvr2 | Ubuntu 14.04 LTS | Web Server & Load Generator | Xeon E5-1620 | 32 GB | Intel® Ethernet Converged Network Adapter X540-T1 10GbE | |
| perf02 | Windows Server 2012 R2 | Load Generator | Xeon W3550 | 24 GB | Broadcom NetXtreme Gigabit Ethernet (BCM5764) | |
| perf03 | Ubuntu 14.04 LTS | Load Generator | Xeon W3550 | 12 GB | Broadcom NetXtreme Gigabit Ethernet (BCM5764) |
The machines are connected to a 5-port Netgear GS105 Gigabit switch.
We're using wrk to generate load from one of our Linux boxes (usually perfsvr2). We also have WCAT set up on perf02 but it as it doesn't support HTTP pipelining we've stopped using it for now.
For each stack, variations of the load parameters and multiple runs are tested and the highest result is recorded.
Similar to the plain text benchmark in the TechEmpower tests. Intended to highlight the HTTP efficiency of the server & stack. Implementations are free to cache aggressively and remove/disable components that aren't required in order to maximize performance.
| Stack | Server | Req/sec | Load Params | Impl | Observations |
|---|---|---|---|---|---|
| ASP.NET 4.6 | perfsvr | 65,383 | 8 threads, 512 connections | Generic reusable handler, unused IIS modules removed | CPU is 100%, almost exclusively in user mode |
| IIS Static File (kernel cached) | perfsvr | 276,727 | 8 threads, 512 connections | hello.html containing "HelloWorld" | CPU is 36%, almost exclusively in kernel mode |
| IIS Static File (non-kernel cached) | perfsvr | 231,609 | 8 threads, 512 connections | hello.html containing "HelloWorld" | CPU is 100%, almost exclusively in user mode |
| ASP.NET 5 on WebListener (kernel cached) | perfsvr | 264,117 | 8 threads, 512 connections | Just app.Run() | CPU is 36%, almost exclusively in kernel mode |
| ASP.NET 5 on WebListener (non-kernel cached) | perfsvr | 107,315 | 8 threads, 512 connections | Just app.Run() | CPU is 100%, mostly in user mode |
| ASP.NET 5 on IIS (Helios) (non-kernel cached) | perfsvr | 109,560 | 8 threads, 512 connections | Just app.Run() | CPU is 100%, mostly in user mode |
| NodeJS | perfsvr | 96,558 | 8 threads, 1024 connections | The actual TechEmpower NodeJS app | CPU is 100%, almost exclusively in user mode |
| Scala | perfsvr | 204,009 | 8 threads, 1024 connections | The actual TechEmpower Scala plain text app | CPU is 68%, mostly in kernel mode |
| libuv C# | perfsvr | 300,507 | 12 threads, 1024 connections | Simple TCP server, not real HTTP yet, load spread across 12 ports (port/thread/CPU) | CPU is 54%, mostly in kernel mode |
Like the Plain Text scenario above but with HTTP pipelining enabled. Only stacks/servers that support pipelining are included.
| Stack | Server | Req/sec | Load Params | Impl | Observations |
|---|---|---|---|---|---|
| NodeJS | perfsvr | 148,934 | 8 threads, 1024 connections | The actual TechEmpower NodeJS app | CPU is 100%, almost exclusively in user mode |
| Scala | perfsvr | 764,775 | 8 threads, 1024 connections, pipelining 15 deep | The actual TechEmpower Scala plain text app | CPU is 46%, mostly in kernel mode, 1Gbps NIC saturated (need to re-run on 10GbE) |
| libuv C# | perfsvr | 2,379,267 | 36 threads, 288 connections, pipelining 15 deep | Simple TCP server, not real HTTP yet, load spread across 12 ports (port/thread/CPU) | CPU is 100%, mostly in user mode, gets slower as test runs longer, issue somewhere |
Coming soon...
Coming soon...
This project is part of ASP.NET 5. You can find samples, documentation and getting started instructions for ASP.NET 5 at the Home repo.