In this blog post, we'll take you through the improved CloudBurst workflow of Silk Performer 17. Take a look at the graphic below - it shows the workflow we will pursue in this article. We'll start with the registering process to create a new CloudBurst account, then we're going to estimate the costs of a cloud load test using the credit calculator on the Cloud Portal. Next, we're going to prepare and execute the cloud load test using the Cloud Agent Manager and the Workbench. And finally, when the cloud load test is completed, we'll review the costs in the Workbench and on the Cloud Portal. If you haven't yet tried out Silk Performer CloudBurst, this blog article will help you to get started.
Registering on the Cloud Portal
Our first step is to register on the Cloud Portal and to create a CloudBurst account. The CloudBurst account is your starting point if you want to execute cloud load tests. Open Silk Performer and click Tools > Cloud Agent Manager. The CloudBurst login dialog displays. Since we have no CloudBurst account yet, we click Register for Cloud Portal (1). A web browser is started and we are automatically directed to the Cloud Portal, which you can also access through cloud.borland.com by the way. We click Register on the right (2), and then we enter our personal data, enter the captcha string, and accept the terms and conditions.
When you follow these steps, you will receive an email with a link, which you need to click to confirm the registration. You will be redirected to a site on the Cloud Portal where you can select your preferred product. CloudBurst is the one to go for in this case. We have also followed this path and are now redirected back to the homepage of the Cloud Portal. Here we can login through the form on the top right of the site (3).
Estimating the costs of a cloud load test
Once we are logged in, we click the Services tab (1) and then CloudBurst service (2). Here on the CloudBurst service page, we can find a note that we are provided with a starting capital of 200 Micro Focus Credits. Before we can purchase additional credits, we have to specify a credit card. To do so we click Account on the top (3), then the Payment Details tab and the Set Payment Details button. A dialog displays where we can enter credit card details. For our quick sample load test that we are going to perform, the 200 Micro Focus Credits are sufficient though, so we navigate back to the CloudBurst service page.
When we click the Transactions tab (1), we can see a table that lists all CloudBurst operations and the number of remaining credits. Once we have executed our cloud load test, we'll come back to the Transactions page and review the costs of the test.
Next, we click the Calculator tab (2). It shows a credit calculater, which allows us to estimate the costs of a cloud load test with the settings we are planning to use. For our sample load test, we're entering the following values: web (3), 50 virtual users (4), 1 day (5), and 2 test hours (6). Based on these values, the calculator estimates that we will need 80 credits (7) for such a load test. We will now switch to the Silk Performer Workbench and prepare a load test with these values.
Preparing and executing a cloud load test
If you follow along with this blog article, you now have to create a new project, record a short sample script, and specify all your settings up to the Assign Agents step on the workflow bar. For our example, we've configured an Increasing workflow with 50 virtual users in the Adjust Workload dialog. These workload settings are important, as they will continue to show up in some of the upcoming dialogs. We go on with clicking Assign Agents. On the dialog that pops up, we click Use cloud agents. If you are not already logged in to your CloudBurst account, the login dialog displays. We enter the credentials from our newly created CloudBurst account and click Login.
The Prepare cloud agents dialog displays. We've already specified to use 50 virtual users for our load test in the Adjust workload dialog, this value now displays in the Workload field (1). In the field below, we specify to use 25 virtual users per agent (2). Then we enable the regions, where our cloud agents are supposed to be located. In our case, we select California (3) and Germany (4), and we click Next (5).
The Cloud Agent Manager launches and the Start Agents dialog displays. It gives us a number of important facts: Firstly, it displays the two cloud agents from Germany and California, we're about to start (1). Next, It shows us that 20 Micro Focus Credits will be reserved (2), once we start the cloud agents.
This is the so-called passive infrastructure rental fee. It is charged for each running agent and is calculated per hour. The description text includes the current fee, which is 5 credits per agent and per started hour (3). We've specified an automatic shutdown in 2 hours (4). This results in a passive infrastructure rental fee of 20 credits - 5 credits x 2 agents x 2 hours. To get a comprehensive overview over the fees that are associated with Micro Focus Credits, click the Help link on the dialog (5) or click this link: Micro Focus Credits. Now it's time to start the agents (6).
Once the cloud agents are ready, which can take a few moments, we can continue by clicking through the workflow until we reach the Workload Configuration dialog. We click the Agent Assignment tab, where we can choose whether we want to use cloud agents only or whether we want to use a mix of cloud agents and on-premise agents. We return to the Workload tab and click Run to start the load test.
Before the load test is actually started, another dialog comes up. It shows the estimated amount of credits that the load test will presumably require. The exact consumption cannot be calculated before the test is completed because it might be the case that not all virtual users will be started. We will simulate such a case by manually stopping the load test. In our example, the dialog shows an estimated consumption of 60 Micro Focus credits (1). These 60 credits are made up of 50 credits for our 50 virtual users (2) and 10 credits (3) for the infrastructure costs. This estimation assumes that all 50 virtual users are started as planned. We click Accept and Run to actually start the load test. Since we've configured an Increasing workload, the virtual users are starting up gradually. After a few moments we manually stop the test by clicking the little Stop All icon in the Monitor window.
Reviewing the costs
The Load Test Summary page displays and we click View debit information on the left in the Next Steps area. The dialog that pops up shows how many credits were actually debited. In our case it's 17 credits (1) and not 60 credits according to the estimation. This is because we stopped the test before it could start up and run all virtual users. Also note the infrastructure fees. The active infrastructure rental fee is 5 credits per agent and per hour, which equals 10 credits (2). The passive infrastructure rental fee is 0 credits (3). This row always shows 0, since the passive fee is always already covered.
Remember the Start agents dialog in the Cloud Agent Manager we've discussed above. That dialog displayed an estimated passive infrastructure rental fee of 20 credits. But note that these 20 credits were an estimation too. We initially planned to use the 2 cloud agents for 2 hours, which would result in 20 credits (5 credits x 2 agents x 2 hours). However, we're now going to manually stop the agents. We will then be charged a passive fee for 2 agents for only 1 hour, which equals 10 credits (5 credits x 2 agents x 1 hour).
So, we click OK to close the dialog and click Tools > Cloud Agent Manager. We select the two regions in the Cloud Agent Manager, click Stop Agents and Stop Selected. Let's review the amount of credits we were debited in the Cloud Portal. We switch to the Cloud Portal and click the Transactions tab (1). You probably have to wait some minutes until all data is delivered to the Cloud Portal. Once this is the case, the table shows the 17 credits (2) we were debited for the virtual users as well as 5 credits (3) for each started cloud agent. We now have a total of 173 Micro Focus credits left for further load tests (4).
The CloudBurst workflow has been much improved with Silk Performer 17. Trying out CloudBurst takes just a few simple steps to register and create a CloudBurst account. With the free starting capital of 200 Micro Focus credits, you can experiment and familiarize with working with cloud agents.
If you want to learn more about CloudBurst, you can read the CloudBurst Tips and Tricks blog article and watch the corresponding video. The Help also provides a comprehensive section about Silk Performer CloudBurst.
Watch the Improved CloudBurst Workflow video
To get an even better understanding of the CloudBurst workflow, watch the following video:
To learn more about all new features and enhancements Silk Performer 17.0 provides, take a look at the following blog posts:
- Released: Silk Performer 17.0
- HTTP Live Streaming
- Performance Explorer Enhancements
- Generating Clean Scripts
- Evaluating Agent Capacities
The Silk Performer Help is another comprehensive source of information:
If this article was useful for you, leave us a comment or like it. We appreciate any of your feedback.