Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Table of Contents

Hardware

Client host (epam1)

Approach

Benchmarks were run against two different versions of FIX Antenna Java (2.24.2 and 2.19.4), and the results were compared in order to detect possible regressions.

Hardware

FIX Antenna Machine

  • Intel(R) Xeon(R) CPU E5-

...

  • 2643 v3 @ 3.

...

  • 40GHz (2 CPU Hyper-Trading Enabled,

...

  • 24 Cores)
  • RAM 128 GB, 2133 MHz
  • NIC Solarflare Communications SFC9120 (Firmware-version: 4.2.2.1003 rx1 tx1)
  • Linux (CentOS 7.0.1406 kernel 3.10.0-123.el7.x86_64)
  • SolarFlare driver version: 4.1.0.6734a

...

Client Machine

  • Intel(R) Xeon(R) CPU E5-2643 2687W v3 @ 3.40GHz 10GHz (2 CPU Hyper-Trading Enabled, 24 20 Cores)
  • RAM 128 GB, 2133 MHz
  • NIC Solarflare Communications SFC9120 (Firmware-version: 4.2.2.1003 rx1 tx1)
  • Linux (CentOS 7.0.1406 kernel 3.10.0-123.el7.x86_64)
  • SolarFlare driver version: 4.1.0.6734a

...

Benchmarks

Single Session Echo Scenario

  • FIXAJ has one One acceptor session is configured on server hostthe FIX Antenna Java side.
  • Client application has one One initiator session is configured on the client hostapplication side.

 The overall process is following:

  1. Client The client application connects to the FIXAJ instance FIX Antenna Java and sends 200000 FIX 4.2 messages with at a rate of 5000 50000 messages per second.
  2. FIXAJ FIX Antenna Java receives the messages and matches it to the same session using business layer logic.FIXAJ responds to the client application with the same message in via the same TCP/IP connection (the same session).
  3. The client application collects the response time histogram.
  4. The process is repeated 5 times for each FIX Antenna Java version.

The response time measured by the client application is the difference between timestamps:

  • t1 - timestamp is taken right before sending a message to

...

  • the socket
  • t2 - timestamp is taken right after

...

  • receiving the same message

...

  • from the socket (from the FIX Antenna Java)

So the round-trip time formula is: RTT = t2 - t1 and measured in microseconds.

Results

Image Removed

...

Single session, usec

...

the measurement unit is microseconds.

The test scenario diagram:

Drawio
borderfalse
diagramNameFIX Antenna 2.24.2 Benchmark
simpleViewerfalse
width600
linksauto
tbstylehidden
diagramDisplayName
lboxfalse
diagramWidth540
revision1

Results

The performance of the FIX Antenna Java 2.24.2 release is almost the same as the performance of the FIX Antenna Java 2.19.4 release. There is no performance degradation or improvement.

Performance value slightly differs from run to run due to the nature of the system used to measure the performance.

Single Session Echo Scenario

Image Added

The table below contains a single run one-to-one comparison and may have some differences compared to the graphic above that shows a 5 runs comparison.


FIXAJ 2.24.2, usecFIXAJ 2.19.4, usec
Min9.2968.968
Max206.335210.687
Median9.6399.263
Average9.876569.56634
50%9.6399.263
75%9.7519.399
90%9.8639.599
95%9.9359.727
99%10.86310.519
99.9%108.159150.655
99.99%191.999194.303