IBM DevOps Test Performance vs. Selenium

Go for IBM RPT if: 1. You're testing a Java-based Web application with HTTP protocol 2. You wanted to distribute the load across machines easily 3. Your team is in learning phase/not really introduced to a wide range of performance testing tools Do not go for IBM RPT if: 1. You wanted to test REST or any other advanced protocols 2. Your system under test demands a very high user load 3. Your application is written in .NET or any other platform except Java.

Incentivized

When you have to test the UI and how it behaves when certain actions are performed, you need something that can automate the browsers. This is where Selenium comes to the rescue. If you have to test APIs and not the frontend (UI), I would recommend going with other libraries that support HTTP Requests. Selenium is good only when you have no choice but to run the steps on a browser.

Incentivized

• Data Parameterisation/ Data Correlation is made simpler compared to its competitors
• Distributed Load Testing is easier to set up
• Performance metric gathering while a test is in progress
• Th look and Feel really helps a beginner to understand and work with.

Incentivized

• For any web based UI automation, Selenium is the best tool out there to automate your tests.
• It supports multiple coding languages like Java, Python, Ruby, C# etc.. to choose from.
• There is a huge community of users and can get many answers on StackOverFlow.
• It has lot of other plugins to make your tests even more efficient.

Incentivized

• Memory utilization could have been improved.(Eats up system's RAM)! It may crash if a test is conducted with the heavy load if adequate RAM is not available in the VM/host machine.
• Licensing could have been made simpler. IBM's licensing method is difficult to follow.
• Support for protocols other than HTTP. Not really up to the current trend.

Incentivized

• Selenium is pretty user-friendly but sometimes tests tend to flake out. I'd say roughly one out of twenty tests yields a false positive.
• Selenium software cannot read images. This is a minor negative because a free plug-in is available from alternate sources.
• Slowness may be a minor factor with Selenium, though this is an issue with basically any testing software since waiting on a site to execute JavaScript requires the browser to wait for a particular action.

Incentivized

We love this product mainly because of its high customization abilities and the ease of use. Moreover, its free and can be learned easily through online communities and videos. The tests are more consistent and reliable as compared to Manual tests. It has enabled us to test a large number of features all in one go, which would have impossible through manual tests. The reports generated at the end of the tests are really helpful for the QA and the development teams to get a fair view of the application.

Incentivized

For those who are unfamiliar with coding, there is a bit of a learning curve. There is plenty of helpful documentation and resources but it can take a little time to get the software up and running. Once you get the hang of how Selenium works, and what it can do, you realize how many things you can use it for, and how many processes you can automate.

Incentivized

The Selenium app has a pretty fat community of users. For the problems we are experiencing, we are primarily receiving support from these communities. In addition, there is widespread service support. Instant support is given to the problems we experience when we need Online support. We and our team are happy to provide this support, especially before important deployment processes

We did everything we needed to use it. Now we can execute our tests on different operational systems and browsers running few tests simultaneously. We also implemented Appium framework to execute our tests on mobile devices, such as iPhones, iPads, Android phones and tablets. We use SauceLabs for our test execution and Jenkins for continuous integration.

Incentivized

Cost/Licensing: While JMeter is an opensource testing tool from Apache, compared to IBM RPT and HP LoadRunner, RPT is much cheaper than Loadrunner. Functionality: JMeter provides basic functionalities which are adequate for performance testing, however advanced features are not available (such as load testing with GUI, reporting is very basic etc.). But when it comes to Loadrunner, it offers very broad features and supports a variety of protocols. So in this category, Loadrunner is a winner, but RPT is better than JMeter. Ease of operating: JMeter is easy compared to LoadRunner, but it has old GUI and look and feel is not that great to understand. Also, most of the things are to be done in a command line, non-GUI mode. While LoadRunner is very advanced with many options, which also confusing sometimes. But RPT, on the other hand, maintains a balance between simplicity and offering of different features. So winner: RPT.

Incentivized

At the time of adoption, there were not many other alternatives that were even close to being competitive when it comes to browser testing. As far as I know now to this day, there is still little competition to Selenium for what it does. Any other browser-based testing still utilises Selenium to interact with the browser.

Incentivized

• Accuracy in metrics, thus improving system's performance
• Costs less compared to competitor like HP LoadRunner
• Helped the team of beginners learn things quickly

Incentivized

• There hasn’t been a downside to using it yet other than we’ve got to update the programs we create for each change.
• This has saved us hundreds of hours of manpower by allowing our automation engineer to rapid fire tests.
• We are able to screenshot and save entire sites before and after launch with a program the automation engineer created
• We can compare large volumes of data against data in excel docs with a program created using Selenium

Incentivized