Zen Load Balancer Performance Tests Part 2

All right. Time to upgrade the entire thing as described in the "Zen Load Balancer 3.0.3 Perfomance and Security Customization"series.

In summary:

Zen Load Balancer 3.0.3 Perfomance and Security Customization Part 5 was skipped, no NIC bonding was necessary;
Zen Load Balancer 3.0.3 Perfomance and Security Customization Part 4 was implemented, with the 1GB+ recommended settings and no high-memory usage settings;
Zen Load Balancer 3.0.3 Perfomance and Security Customization Part 3 was skipped, no point of checking for invalid packets or rate limit. This is just a test server;
Zen Load Balancer 3.0.3 Perfomance and Security Customization Part 2 was implemented;
Zen Load Balancer 3.0.3 Perfomance and Security Customization Part 1 was implemented (no filesystem tuning though), although the system's memory is 2GB, which means the only benefit here is the NX protection which won't be necessary since this is a test server (really, this should actually produce some overhead).

Let's go over my Zen Load Balancer's stats again:

Memory: 2048MB
CPU #1: 3.0GHz

I decided to use G-WAN as the backend, mainly because:

a) I hadn't used it before and it was a good opportunity to do so;
b) They brag a ridiculously high performance throughput -you know you have to at least check if it's partly true, and if it is, all the better as I'll stress Zen even more;
c) They have an enormous amount of haters; from other webserver fanbois who refuse to accept that a different webserver than the one they use might be better, to company plants that ticket non-existant vulnerabilities to try to prove that G-WAN isn't as good as its publishers claim. This alone can make me install software just out of spite.

So G-WAN it is, then!

The webpage that I will serve will be generated by this little C++ servlet here:

root@gwan:/opt/gwan_linux64-bit/0.0.0.0_8080/#0.0.0.0/csp# cat bench.cpp
// ============================================================================
// C++ servlet example to benchmark Zen and G-WAN Web Application Server   (http://gwan.ch/)
// ----------------------------------------------------------------------------
// bench.cpp: Concatenate a bunch of "Hello world!" until it gets to 400KB (encodings included).
//
// This code is based on the hello.cpp example by Thomas Meitz (gwan at jproxx dot com)
// ============================================================================
// imported functions:
//   get_reply(): get a pointer on the 'reply' dynamic buffer from the server
//   xbuf_cat(): like strcat(), but it works in the specified dynamic buffer
// ----------------------------------------------------------------------------
#include "gwan.h" // G-WAN definitions
#include <string>

using namespace std;

class Hello
{
public:
   void whatsup(xbuf_t* reply, string& str)
   {
      xbuf_cat(reply, (char*)str.c_str());
   }
};
// ----------------------------------------------------------------------------
int main(int argc, char *argv[])
{
   string h("Hello World! ");
   do {
   h+=h;
   } while (h.length() < 240000);
   Hello hello;
   hello.whatsup(get_reply(argv), h);

   return 200; // return an HTTP code (200:'OK')
}
// ============================================================================
// End of Source Code
// ============================================================================

These were the farm settings:

Tests with Zen Load Balancer, customized as per the recommendations in my blog

The test will be the same as before; we'll send 10, 50 and 100 requests concurrently and 10000 in total, all trying to access the output of our script.

The tests will be run 5 times each. After that, we'll find the median value according to that point in time so as to eliminate any statistical aberrations.

D:\apache\Apache24\bin>abs -c10 -n10000 -f TLS1 https://192.168.0.99:443/?bench.cpp
This is ApacheBench, Version 2.3 <$Revision: 1554214 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking 192.168.0.99 (be patient)
Completed 1000 requests
Completed 2000 requests
Completed 3000 requests
Completed 4000 requests
Completed 5000 requests
Completed 6000 requests
Completed 7000 requests
Completed 8000 requests
Completed 9000 requests
Completed 10000 requests
Finished 10000 requests


Server Software:        G-WAN
Server Hostname:        192.168.0.99
Server Port:            443
SSL/TLS Protocol:       TLSv1,RC4-SHA,1024,128

Document Path:          /?bench.cpp
Document Length:        425984 bytes

Concurrency Level:      10
Time taken for tests:   110.865 seconds
Complete requests:      10000
Failed requests:        0
Total transferred:      4262600020 bytes
HTML transferred:       4259840000 bytes
Requests per second:    90.20 [#/sec] (mean)
Time per request:       110.865 [ms] (mean)
Time per request:       11.086 [ms] (mean, across all concurrent requests)
Transfer rate:          37547.54 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        4   13   6.9     11     106
Processing:    45   98  24.3     97     756
Waiting:        1   11   8.0     10     103
Total:         57  111  25.0    109     775

Percentage of the requests served within a certain time (ms)
  50%    109
  66%    114
  75%    117
  80%    120
  90%    127
  95%    134
  98%    146
  99%    161
 100%    775 (longest request)

Whoa, got to say, this looks extremely promising. Let's do this four more times, find the median values and create a chart out of it.

This is our Zen's CPU usage for 10 concurrent connections:

This is not bad at all. No excessive CPU usage, and good performance as well for an HTTPS farm; apachebench reported around 90 requests per second.

For 50 concurrent connections:

Insane. Once again, no excessive CPU usage, and good performance for an HTTPS farm; apachebench reported almost 89 requests per second!

For 100 concurrent connections:

And this seals it. With our recommended performance upgrades, Zen is both faster and uses a lot less CPU. Just compare this with our previous results! For the record, apachebench reported around 80 requests per second.