Product Comparisons
Apache JMeter
This chart shows a list of the differences between Web Performance Load Tester™ 4.0 and Apache JMeter 2.3 to the best of our knowledge.
| Feature | Load Tester | JMeter |
|---|---|---|
| Create a testcase | Record a testcase by interacting with your site in your browser. | Built-in proxy recorder handles some testcases but cannot record SSL sites - a 3rd party tool is required, which works only with Internet Explorer. |
| Correlation of dynamic fields | Automatic (and extensible) | Manual (except for cookies, which are automatic) |
| .NET VIEWSTATE support | Automatic | Manual |
| Multiple Connections | Modern browsers (IE8, FF 3.5, etc) open as many as six simultaneous socket connections per server when downloading components for a page. By default Load Tester executes a testcase using the same number of socket connections as the browser did during recording. This can be customized for each testcase if desired. | Supports only a single socket connection - resulting in innacurate reporting of total web page duration. |
| SSL Support | Records and plays back SSL, including client certificates. | SSL playback only. |
| Server Monitoring | Load Tester's Advanced Server Agents can monitor dozens of critical metrics on Windows (including .NET) and Linux. | None. |
| Multiple Simultaneous Test Cases | Multiple test cases can be combined into a load profile based on percentages, where each case can be a different percentage of the total load and have other unique settings. For example, 20% of the users could be searching via modems, 20% searching via DSL, 20% making a purchase, etc. | A single test case can run at one time. |
| View and Compare content | A built-in viewer shows the pages as recorded and when replayed. Replays and failed pages can be easily compared to the original recording. | no |
| Support | Load Testing experts | Google and user groups |
| Data Analysis | Extensive reports give a clear picture of all aspects of the performance of your site. Data can be viewed by time or summarized by user level. Results can be analyzed against performance goals, which can be set for the testcase and customized for each page and url. | Records URL duration and errors. Simple tables and charts. |
| How many users can your website handle? | Calculates how many users the website can handle based on performance goals set by the tester. | Not supported. |
| Which web pages are slow? | Sorted web page duration report identifies which pages are slow and under what conditions. | Not supported. |
| Think Time | By default, uses recorded think time for each web page. Can be customized to vary the think time randomly or use no think time. | No way to record think times. No concept of a web page, only urls. Uses same value for each URL or or random values. |
| Usernames & Passwords | Authentication wizard automatically detects and configures usernames and passwords for form field, NTML (Windows Domain), client certificate, and basic authentication. | Supports form field and basic authentication, but they must be configured by hand by editing each URL. |
| Bandwidth Simulation | Virtual users can connect to the web server using different simulated bandwidths. With a load profile 50% of the users could connect using a DSL bandwidth, and 50% with 56K modem bandwidths. | Not supported. |
| Client Certificates | Yes | No |
| ETag support | Yes | No |
| Automatically validate 304 respones | Yes | No |
| Measures Waiting Users | Yes | No |
| Run stepped-ramp tests | Yes | No |
| Supports the cloud | Yes | No |
| Software update installation | Automatic | Manual |
| Simulation Efficiency | Load Tester's highly tuned I/O engine can simulate more than 3500 users on a single computer. 500 users is typical for moderately complex testcases on modern hardware. | A simple 6 web page test with images and think time between pages consumed 90% of the CPU. |
| Load Balancing | The performance and memory usage of all test computers is monitored to insure accurate performance measurements. While controlling multiple computers the task of simulating users is load balanced among them using dynamic performance measurements of the test computers. | Not supported. |
| Multiple IP Addresses | Each virtual user can come from a different IP address. |
Not supported. |
| NTLM | Automatically supported in both recording and playback. | Not supported. |
| Session Tracking | The following session techniques are automatically detected and configured via a wizard:
Cookies URL rewriting URL parameter Hidden Form Field |
Session tracking must be detected and configured by hand by inserting controllers. |
