ESB Performance Testing - Round 5
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

 
View only
 
 
ABCDEFGHIJKLMNOPQRST
1
ESB Performance Testing - Round # 5
2
3
Please refer to http://esbperformance.org for details of the ESB Performance Benchmark
4
5
This round was executed on a dedicated Amazon EC2 c1.xlarge (20 ECU / 8 Core / 7G RAM) node on the 3rd of October 2011
6
Network isolation was not required as all traffic was between the same node as the load generator, each ESB and the backend service was executed on the same node. All results were executed across a single instance without a shut down or restart in-between.
7
8
Each ESB was allocated a 2G Heap in memory. A 2G RAM Disk was made available at /tmp/ram
9
Sample message sizes ranged from 512 bytes to 100K bytes, with concurrency levels of 20, 40, 80, 160, 320, 640, 1280, 2560 users
10
11
Configuration source code for each ESB can be found at https://bitbucket.org/adroitlogic/esbperformance
12
13
Credits
14
Asankha C. Perera (asankha@adroitlogic.com), Founder and CTO, AdroitLogic
15
Ruwan Linton (ruwan@adroitlogic.com), Director of Engineering, AdroitLogic
16
Email: info@adroitlogic.com
17
18
Disclaimer
19
This benchmark was conducted by AdroitLogic, which develops the free and open source UltraESB included in this comparison
20
21
The ESB Performance suite was originally developed in June 2007 by the then WSO2 ESB team members, and has subsequently been used by Mulesoft, BEA and AdroitLogic in addition to WSO2.
22
23
These configurations have been built by AdroitLogic, sometimes with the help of the some of the ESB vendors, and sometimes without any help. They are also based on the previous resources published by WSO2 and Mulesoft, and the solutions we've come up with may have inherent limitations due to our limited understanding of some of the other ESBs and the limited time we had spent. Hence these configurations may not be optimal, and/or may not be optimally tuned for the scenarios tested
24
25
However, AdroitLogic has spent many days and nights to get the other ESBs to work under identical conditions as much as possible. We have strived to give each ESB 2G of heap memory, 300 worker threads, and set the socket timeout to 120 seconds. But due to limitations in available time and publicly available documentation etc, we could not set the threads and socket timeouts the same for each ESB.
26
27
We would be happy to receive feedback to improve the configurations used, and the tuning parameters for the different ESBs from the vendors and/or end users. We will include those updates into future rounds.
28
29
30
31
WSO2 and WSO2 ESB are trademarks of WSO2 Inc.
32
MuleSoft and Mule ESB are trademarks of MuleSoft
33
Fuse ESB is a trademark of IONA Technologies PLC
34
Petals is a trademark of EBM WebSourcing
35
JBoss is a trademark of Red Hat, Inc
36
Talend is a trademark of Talend
37
Apache ServiceMix is a trademark of the Apache Software Foundation
38
UltraESB and AdroitLogic are trademarks of AdroitLogic Private Ltd.
39
40
All other product and company names and marks mentioned are the property of their respective owners and are mentioned for identification purposes only.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Loading...
 
 
 
Introduction
Summary Data
Summary Graphs
Requests Per Second
Visible Response Time (ms)