Real Load News

Real Load Synthetic Monitoring Generally Available!

Real Load Synthetic Monitoring Generally Available.

Performance testing as a main and synthetic monitoring as a dessert? Now available at Real Load.

Recently we’ve launched a new version of our Real Load portal which adds the ability to periodically monitor your applications. The best thing to it is that you can re-use already developed load testing scripts to periodically monitor your applications. It makes sense to re-use the same underlying technology for both tasks, correct?

And, of course, nobody forces you to have the main followed by the dessert. You can also first have the dessert (… synthetic monitoring) followed by the main (… load testing). Or perhaps the dessert is the main for you, you can arrange the menu as it best suits your taste.

How it works

As for load testing, you’ll first have to prepare your testing script. Let’s assume you’ve already prepared it using our Wizards. Once the script is ready, the only remaining thing to do is to schedule it for periodic execution.

Configuring Monitoring Groups and Jobs

You can setup so called Monitoring Groups which will be made up of a number of Monitoring Jobs. Each monitoring job executes one of the prepared test scripts.

In this screenshot you’ll see one Monitoring Group called “Vinnuo APIs” (in the red box) which executes one test script, highlighted in the green box.

Scheduling execution interval and location

The key properties you can configure on a Monitoring Group are:

  • Execution interval: Down to 1 minute, depending on licensing level.
  • Execution timeout: How long to wait for the job to complete, before considering it failed.
  • Max. Data Storage: How long to retain job execution results. Up to approx. 1 year, depending on licensing level.
  • Measuring Agents: The location (agents) to executed the monitoring jobs from. We recommend executing the job from at least 2 agents.

Last, you can enable/disable execution for the group by using the Execution Enabled toggle switch.

Configuring Monitoring Job

Next, you’ll add at least one Monitoring Job to the Monitoring test. Simply select from one of your projects a Test that was previously prepared.

In this example, I’ve picked one of the test that generates the relevant REST API call(s).

Next you’ll need to configured these key parameters relating to the execution of the test script. If you’re familiar with the load testing features of our product, these parameters will be familiar:

  • Number of Users: The number of Virtual Users to simulate. Given this is a monitoring job, this would typically be a low number.
  • Max test Duration: This will limit the duration of the job. Again, given this is a monitoring job the Max duration should be kept short.
  • Max Loops per User: The maximum number of iterations of the test script executed by each Virtual User. One iteration should typically be sufficient for a monitoring job.

Alerting Groups and Devices

A synthetic monitoring solution wouldn’t be complete with an alerting functionality.

You can configure Alerting Groups to which you can assign a number of different device types to be alerted. Supported device types are:

  • Email: An email address to deliver the alert to.
  • SMS: SMS alerting. This is subject to additional costs (SMS delivery costs).
  • Webhook: You can configure a WebHook, for example integrating with an existing alerting system.

In this example, I’ve created an alerting group called “SafeArea IT”…

… to which I’ve assigned one email alerting device. Needless to say, you can assign the same alerting group to multiple

Now that the Alerting Groups are configured, you can configure alerting at the Monitoring Group or Monitoring Job level, whichever best suits your use case. Simply click on the alert icon and assign the Alerting Groups accordingly:

Monitoring Dashboard

Once you’re done with the configuration, you’ll be able to monitor the health of your applications from the Real Time Dashboard. Please note that his dashboard is in evolution, and we’re adding new features on a regular basis.

For now, you’ll be able to:

  • See the overall status of all your synthetic monitoring jobs.
  • Look at the results of the last test execution for each job by clicking on the graph symbol pointed at by the red arrow.
  • Look at the logs of the last execution by clicking at the symbol pointed at by the green arrow.
  • Look at the overall scheduling log by clicking at the symbol pointed at by the blue arrow.

This screenshot shows the logs collected for the last job execution (green arrow). Note the additional logfiles (higlighted in the green box) that you can look at.


Regardless of whether your looking for a Synthetic Monitoring or Perfomance Testing solution, we can satisfy both needs.

Sign up for a free account on our portal and click at “Sign up” (no Credit Card required). Then reach out to us at so that we can get you started with your first project.

Happy monitoring and load testing!

Support for SSL Cert Client authentication in Proxy Recorder

Added support for SSL Client certs to proxy recorder

Do you need to record HTTP requests for your test script that require SSL Cert Client authentication? Tick, we support this use case now…

How does it work…

The Proxy Recorder has been enhanced to support recording against websites or applications that require presenting a valid SSL Client certificate.

From an high-level point of view this is how things work:

  • 1 - SSL Client certificates are uploaded to the Real Load Portal. Each certificate is associated with an hostname (or IP address) of the target server, so that the Proxy Recorder knows when to present the SSL Client certificate.
  • 2 - The Real Portal will then share the SSL Client certs with Proxy Recorder. Currently Cloud Hosted proxy recorders are supported.
  • 3 - The tester then executes the steps to be recorded and included in the test script.
  • 4 - When the Proxy Recorder attempts to access hosts that required SSL Client authentication, the relevant SSL Client will be applied.

SSL Client Certificate Configuration

SSL Client certificates in the .pfx/.p12 format need to be uploaded to the Real Load Portal server.

The configuration of such SSL Client certificates in the Real Load Portal server is done by going to the Remote Proxy Recorders menu item and then clicking on the certificate symbol:


Then provide details about the certificate your uploading. Importantly the target server host must exactly match the hostname (or IP address) that will appear in HTTP requests.


Done. Once uploaded, using the Proxy Recorder attempt to access a resource that requires SSL Client Cert authentication. You should be able to access the resource.

Some SQL performance testing today?

SQL load testing

Most performance testing scenarios involve an application or an API presented over the HTTP or HTTPS protocol. The Real Load performance testing framework is capable of supporting essentially any type network application, as long as there is a way to generate valid client requests.

Real Load testing scripts are Java based applications that are executed by our platform. While our portal offers a wizard to easily create tests for the HTTP protocol, you can write write a performance test application for any network protocol by implementing such a Java based application.

This article illustrates how to prepare a simple load test application for a non-HTTP application. I’ve chosen to performance test our lab MS-SQL server. What I want to find out is how the SQL server performs if multiple threads attempt to update data stored in the same row. While the test sounds academic, this is a scenario I’ve seen leading to performance issues in real life applications…


Key requirements to implement such an application are:

  • You’ll need Java client libraries (… and related dependencies) implementing the protocol you want to test. In this case I’ll use MicroSoft’s JDBC driver and Hikari as the SQL connection pool manager.
  • You’ll need to determine what logic your load test application should execute. In this example, I’ll run an update SQL statement.
  • You’ll need to determine the metrics you want to measure during test execution. We’ll collect time to obtain a connection from the pool and the time to execute the SQL operation.
  • Make sure the Measuring Agent has network access to the service to be tested (… MS-SQL DB in this case).
  • Last, you’ll need some Java skills to put together the load testing application or access to somebody that will do that for you.

Step 1 - Implement the test script as a Java application

Using your preferred Java development environment, create a project and add the following dependencies to it:

  • DKFQSTools.jar - Required for all performance testing applications
  • mssql-jdbc.jar (The MS-SQL JDBC driver)
  • hikari-cp.jar (JDBC connection pooling)
  • slf4j-api.jar (Required by Hikari)

In NetBeans, the dependencies section would look as follows:

Once the dependencies are configured in your project, we’ll implemented the test logic (the AbstractJavaTest interface). For this application, we’ll create the below class.

Of particular relevance are these methods:

  • declareStatistics(): This is where you declare statistics metrics to be collected as the test is executed.
  • executeUserSession(): This method is invoked for every virtual user to be simulated. Note the SQL update statement that will be executed as part of this test script.

import com.zaxxer.hikari.HikariDataSource;
import com.zaxxer.hikari.pool.HikariPool;
import java.sql.Connection;
import java.sql.SQLException;
import java.sql.Statement;
import java.time.Instant;
import javax.sql.DataSource;

@AbstractJavaTest.ResourceFiles(fileNames = {})
public class MSSQLTest extends AbstractJavaTest implements AbstractJavaTestPeriodicThreadInterface {

    private static HikariPool pool;
    private static HikariDataSource dataSource = null;

     * Static Main: Create a new instance per simulated user and execute the
     * test.
     * @param args the command line arguments
    public static void main(String[] args) throws SQLException, NoSuchFieldException, IllegalArgumentException, IllegalAccessException {
        stdoutLog.message(LOG_INFO, "Max. Java Memory = " + (Runtime.getRuntime().maxMemory() / (1024 * 1024)) + " MB");

        dataSource = new HikariDataSource();

        java.lang.reflect.Field field;
        field = dataSource.getClass().getDeclaredField("pool");
        pool = (HikariPool) field.get(dataSource);

        // log test specific resource files, annotated by @AbstractJavaTest.ResourceFiles at class level
        try {
            // get all generic command line arguments

            // create a new instance per simulated user
            for (int x = 0; x < getArgNumberOfUsers(); x++) {
                new MSSQLTest(x + 1);

            // start the test
            stdoutLog.message(LOG_INFO, "[Start of Test]");
            try {
                // start the user threads

                // wait for the user threads end
            } catch (InterruptedException ie) {
                stdoutLog.message(LOG_WARN, "Test aborted by InterruptedException");

            stdoutLog.message(LOG_INFO, "[End of Test]");
        } catch (Exception ex) {
            stdoutLog.message(LOG_FATAL, "[Unexpected End of Test]", ex);
        } finally {

     * Close all output files.
    private static void closeOutputFiles() {

    // - - - vvv - - - instance  - - - vvv - - -
    private CombinedLogAdapter log = new CombinedLogAdapter();

     * Constructor: Create a new instance per simulated user.
     * @param userNumber the simulated user number
     * @throws IOException if the user statistics out file cannot be created
    public MSSQLTest(int userNumber) throws IOException {

    public void declareStatistics() {
        declareStatistic(0, STATISTIC_TYPE_SAMPLE_EVENT_TIME_CHART, "Get connection from pool", "", "Execution Time", "ms", 0, true, "");
        declareStatistic(1, STATISTIC_TYPE_SAMPLE_EVENT_TIME_CHART, "Exec SQL Update stmnt ", "", "Execution Time", "ms", 1, true, "");

    public void executeUserStart(int userNumber) throws Exception {
        // start a periodic thread that reports summary measurement results measured across all simulated users
        if (userNumber == 1) {
            AbstractJavaTestPeriodicThread periodicThread = new AbstractJavaTestPeriodicThread(this, 1000L, this);


    public int executeUserSession(int userNumber, int sessionLoopNumber) throws Exception {
        long measurementGroupStartTime$0 = System.currentTimeMillis();

        // 1- Get a connection from pool
        Connection connection = null;
        try {
            connection = dataSource.getConnection();
        } catch (Exception e) {
            log.message(LOG_ERROR, e.getMessage());
            return SESSION_STATUS_FAILED;
        addSampleLong(0, System.currentTimeMillis() - measurementGroupStartTime$0);

        // 2 - Prepare SQL statement
        Statement st = connection.createStatement();
        String SQL = "update TEST_TABLE set VALUE_NUM = '7058195060625506304' where DATA_URI = '2566' AND DATA_URI = '0' AND DATA_ID = '-1'";

        // 3 - Execute statement
        long measurementGroupStartTime$1 = System.currentTimeMillis();
        addSampleLong(1, System.currentTimeMillis() - measurementGroupStartTime$1);

        // end of passed session

    public void executeUserSessionEnd(int sessionStatus, int userNumber, int sessionLoopNumber) throws Exception {

     * Called periodically by an independent thread with the context of the
     * first simulated user. Reports summary measurement results which were
     * measured over all simulated users.
     * @param abstractJavaTest the context of the first simulated user
     * @throws Exception if an error occurs - logged to stdout
    public void onPeriodicInterval(AbstractJavaTest abstractJavaTest) throws Exception {

    public void onUserSuspend(int userNumber) throws Exception {

    public void onUserResume(int userNumber) throws Exception {

    public void executeUserEnd(int userNumber) throws Exception {


    public void onUserTestAbort(int userNumber) throws Exception {



Step 2 - Upload Java app and dependencies to Real Load portal

Once you’ve compiled your application and generated a jar file (… make sure the main class is mentioned in the META-INF/MANIFEST.MF file) you’re ready to configure the load test in the Real Load portal.

After logging into the portal, create a new project (… “MSSQL” in the below screenshot) and a new Resource Set (“Test 1”). Upload your performance test application jar file (“RealLoadTest3.jar” in this example) and all other dependencies.

Once everything is uploaded, define a new test by clicking on the Resource Set (“Test 1”). Make sure you select the .jar file you’ve developed as the “Executing Script” and tick all the required dependencies in the Resource list.

Step 3 - Execute load test

You’re now ready to execute your performance test. When starting the test job select how many threads (Users) should execute you test application, ramp up time and test execution duration.

While the performance test is executing, you’ll notice that the metrics you’ve declared in the Java source code appear in the real time monitoring window:

If you keep an eye on MS-SQL Management Studio, in the activity monitor you’ll notice that resource locking is the wait class with the highest wait times. Not so surprisingly I might add, given the nature of the test.

Also note that the waiting task number is very close to the number of virtual users (concurrent threads) simulated, approx. 100.

Once the test completed, you can review collected metrics. The graph at the bottom of this screenshot shows execution times throughout the test of the SQL update statement, as load ramped up.


As you can see, it’s quite straightforward to prepare an application to performance test almost any network protocol.

Should you have a requirement to performance test an exotic protocol and your current tool doesn’t allow you to do so, do not hesitate to contact us. Perhaps we can help…

Thank you for reading and do not hesitate to reach out should you have any Qs.

New Feature - URL Explorer

Easy handling of random values in your load test scripts

An exciting new feature was added in Real Load v4.7.3. The URL Explorer feature allows you to quickly handle session specific random values that might appear in your load test script requests and responses.

In a nutshell, this is how things work:

  • Record a session using the Proxy Recorder.
  • Locate random values that might appear in your test script.
  • Search for these values in the recording using the URL Explorer.
  • Extract and assign these values to a variable.
  • Finally assign the value of the variable to all locations in the load test script where the same value appears.

All of this is documented in this short video (7 minutes) which walks you through the above process.

As always, feedback or questions are welcome using our contact form.

Oracle and async I/O... A world of a difference

What a difference enabling async I/O in Oracle makes…

While running a load test against an API product that I have to deal with in my other day to day job, I’ve noticed something in both the results and at the OS level (… on the DB server) that didn’t make much sense.

Odd results

The results of the performance test where somewhat OK but kinda unstable (… some strange variances in the response times). This graph tells the story better than 1000s of words.

Note the green line (transactions per second) going all over the place:

At first I suspected some sort of issue with the application server (Weblogic) and the DB Connection Pool. But all looked good there…

Then I’ve cast an eye on Oracle Enterprise Manager and noticed that most of the DB waits were related to I/O, although the storage of this particular test DB is located on a reasonably fast NVME SSD.

So I started looking at I/O stats on the Oracle Linux server hosting this DB. Being a lab DB, it’s more or less a standard Oracle install with not much performance tuning applied. Nor am I an Oracle expert that knows all secrets of the trade…

Anyways, there was one thing that somehow didn’t stack up: At the OS level, the % spent by the CPU in iowait was sporadically incredibly high (… 70%+ or so) with the CPU idle time plunging to less than 10%:

After reading various online articles about this, most of which suggested beefier HW or rewrite the app the app so that it would be more efficient with commits, it dawned on my that perhaps Oracle wasn’t using async I/O when writing to disk causing these high waitio stats.

I finally bumped into a few articles talking about async I/O settings in Oracle and found a few useful SQL queries…

This one will assist in figuring out whether async I/O is enabled on your Oracle DB files:


… leading to a result like this. Note that for all files async IO is disabled…

So I decided to enable async I/O with these few SQL commands:


… and then checking again. As you can see, async I/O is enabled now:

NAME                                               ASYNCH_IO
-------------------------------------------------- ---------
/opt/oracle/oradata/AAOP74/datafile/o1_mf_system_j ASYNC_ON

/opt/oracle/oradata/AAOP74/itblspc01.dbf           ASYNC_ON
/opt/oracle/oradata/AAOP74/datafile/o1_mf_sysaux_j ASYNC_ON

/opt/oracle/oradata/AAOP74/datafile/o1_mf_undotbs1 ASYNC_ON

/opt/oracle/oradata/AAOP74/dtblspc01.dbf           ASYNC_ON

NAME                                               ASYNCH_IO
-------------------------------------------------- ---------
/opt/oracle/oradata/AAOP74/datafile/o1_mf_users_jz ASYNC_ON

/opt/oracle/oradata/AAOP74/btblspc.dbf             ASYNC_ON
/opt/oracle/oradata/AAOP74/cm.dbf                  ASYNC_ON
/opt/oracle/oradata/AAOP74/cm.idx                  ASYNC_ON
/opt/oracle/oradata/AAOP74/bodtblspc.dbf           ASYNC_ON
/opt/oracle/oradata/AAOP74/boitblspc.dbf           ASYNC_ON
/opt/oracle/oradata/AAOP74/DTBLSPC03.dbf           ASYNC_ON
/opt/oracle/oradata/AAOP74/ITBLSPC03.idx           ASYNC_ON
/opt/oracle/product/19c/dbhome_1/dbs/reportdt.dat  ASYNC_ON

14 rows selected.

Smooth sailing….

Time to re-run the load test with my preferred tool and the results look encouraging.

As you can see the green results line is much more stable. Not only that, but number of transactions per second (TPS) increased to approx. 136 from 101 in the previous run. Response times also went down somewhat, from 90 to 70ish msecs.

The CPU waitio stats also dramatically improved on the Oracle server:

To summarize, it makes sense to scratch beyond the surface of performance bottlenecks before investing in HW upgrades or so… Sometimes the solution is a low-hanging fruit waiting to be picked.

External references:

ORACLE-BASE - Direct and Asynchronous I/O

I/O Configuration and Design

Apache HTTPD on FreeBSD and Linux Load Test

Comparison of infrastructure resource usage between Linux and FreeBSD HTTPD instances

For various reasons, I’ve had to perform a series of tests to ensure our Measuring Agent can generate traffic from a large number of source IP addresses. Aside from validating that capability, the by-result of the test is a somewhat interesting comparison of a FreeBSD and Linux based Apache HTTPD server.

Generating Load From Multiple IPs

First, a quick overview of what I wanted to prove: I needed to make sure that we can run a Load Test simulating a large number of source IP addresses. To validate this requirement, I’ve configured one of our Measuring Agents with approx. 12k IP addresses. I’ve used a bash script to do that, as otherwise it would take forever. All IPs are assigned as aliases to the NIC from where the load will be generated, and all IPs are within the same /16 subnet.

Finally, I’ve configured my Real Load test script with two additional steps:

  1. Step 0 that selects a random IP address configured on the NIC and stores it in a variable.
  2. Step 2 that instructs the load test to use as src IP the address stored in the variable.

Infrastructure Details

The hypervisor is a Windows 2019 Server Standard edition machine, running Hyper-V and fitted with an somewhat old Xeon E5-2683v3 CPU. The measuring agent and the tested servers are connected to the same virtual switch.

The Linux and FreeBSD VMs are minimal instals of their distributions, onto which I’ve installed the latest Apache HTTPD build offered by the built in software distribution mechanisms. That’s why the HTTPD versions are not identical.

In order for the results to be somewhat comparable, I’ve deployed the same set of static HTML pages on both servers. I’ve also aligned several key HTTPD config parameters on both systems, as shown in this table.

Parameter Measuring Agent FreeBSD HTTPD VM Linux HTTPD VM
OS Version RH 8.4 13.0 Oracle Lnx 8.4
RAM 4 GBs 4 GBs 4 GBs
vCPUs 10 4 4
HTTPD Version n/a 2.4.53 2.4.37
HTTPD MPM n/a event event
ServerLimit n/a 8192 8192
MaxRequestWorkers n/a 2048 2048
ThreadsPerChild n/a 25 25

See further down for other tuning parameters applied to the HTTPD VMs.

Load Test Execution and Result Metrics

I’ve then executed a 20 minutes 1000 VUs load test which, which is configured to maximize the number of HTTP requests generated. Apache is configured to server some static HTML pages, made up of text and some images.

This table summarizes metrics observed once the max. load was reached, approx. 10 minutes into the test. The PDF reports allow you to have a better glance into the test results.

Metric Linux HTTPD FreeBSD HTTPD
User CPU usage 21% 20%
System CPU usage 47 % 70%
Avg reqs/s 8.8k 10.3k
Avg network throughput 1.1 Gbps 1.3 Gbps
Hyper-V CPU usage 10% 11%
Test report PDF Linux Report PDF FreeBSD Report PDF
Test progress screenshot


  • CPU usage was measured with the “iostat 20” command.
  • Hyper-V CPU usage was taken from Windows Admin Center.

And the winner is…

… is difficult to pick, to be honest.

  • CPU usage, as measured by Hyper-V was a little bit higher for FreeBSD. CPU metrics measured within the VMs seem to indicate an overall higher CPU usage by FreeBSD (… in particular System CPU). Perhaps the Linux NIC driver is better optimized for Hyper-V.
  • FreeBSD HTTPD seems to deliver an higher throughput (network and avg requests/s).
  • FreeBSD HTTPD also seems to offer an higher HTTP Keep-Alive efficiency, which might partially explain the higher throughput.
  • Observations (like CPU usage, etc…) were averaged by “eyeballing” metrics displayed on screen. Expect some rounding error…

Assuming I had time to spend to better tune and align the two platforms, I might have been able to squeeze out a bit more performance from each server, but I doubt that would have materially changed the result in favor of one OS or the other. Obviously I’m happy to be proven wrong…

Feel free to email us with your feedback, I’ll be more than happy to test any further tuning suggestions.

OS Tuning

Below the OS level tuning that was applied to the Linux and FreeBSD servers. I didn’t have time to research in full each of the parameters mentioned below, they were mentioned in various other online sources and adopted. I’ve implemented the ones that seemed to make most sense…

Linux HTTPD (/etc/sysctl.conf)

The last 2 tunables were required to prevent the Linux server stopping accepting connections for various reasons…

fs.file-max = 524288
net.ipv4.tcp_max_syn_backlog = 2048
net.ipv4.tcp_synack_retries = 3
net.ipv4.tcp_max_orphans = 65536
net.ipv4.tcp_fin_timeout = 30
net.ipv4.ip_local_port_range = 16384 60999
net.core.somaxconn = 256
net.core.rmem_max = 1048576
net.core.wmem_max = 1048576

FreeBSD HTTPD (/etc/sysctl.conf)


Real Load Portal Generally Available!

The Real Load Portal is open for publich registration.

The Real Load Portal is open for publich registration.

Feel free to register for an account and trial our product by following the registration instructions here. You’ll be given a two weeks 100 VUs demo license and no credit card is required to sign up!

If you’d like a one-on-one session to guide you through the first steps of how to use our product, please do no hesitate to contact us at .

Happy testing!

Desktop Companion 0.24 Released

Quick update walkthrough video

The Desktop Companion is a Desktop GUI that allows you to manage several Real Load aspects directly from your desktop.

We’ve released the latest update and this 5 minutes video illustrates the key changes.

Happy watching!

My system just got faster!

My system just got much faster… I wonder why.

Today I’ve started executing a lengthy performance test against a SOAP API to seed the underlying DB. For various reasons, I need to replicate the daily DB volume increase of a production system in my own lab DB.

I’ve prepared a Real Load test script and started hammering a server in my lab environment. I’ve noticed the performance wasn’t particularly good but that didn’t matter, as I wasn’t actually executing a performance test.

I’ve let the test run and went for lunch (… a sandwich). When I came back, I’ve noticed that my system became somehow much faster, raising from approx. 50 TPS to approx. 200 TPS. Each transaction represents a SOAP request…

See this graph from the real time monitoring window:

Knowing how this particular product works and knowing that typically the performance is limited by the performance of the underlying DB, I’ve started looking at various DB counters and one thing I’ve noticed is that the Response Time reported by MS SQL Studio on a particular DB file went down considerably (… from 100ms+ to 10-20ms).

That was curios…. why would this happen? I’ve then cast an eye on the metrics of my storage system (a TrueNAS self build…) and noticed that the ZFS L2 ARC read cache hits improved noticeably around that time. Notice the orange line, next to 0% hit ratio around 12PM and then raising to 90%+ after approx. 50 minutes.

Anyways… this just goes to show that having access to metrics of all infrastructure components during a load test is critical. But sometimes getting to these metrics can really be hard. Just need to persist to get to be bottom of things…..

Desktop Companion Enhancements

Features added to make recording of HTTP sessions more user friendly

This is a short update video to illustrate enhancements in the last update of the Desktop Companion.

Enhancements were done to the Proxy Recorder tab:

  • Allow adding page breaks as you navigate from page to page while recording.
  • Added a real-time counter of the requests being recorded.
  • Added a button to force the Desktop Companion window on top of others, so it doesn’t get hidden by browser windows.

All of the above is illustrated in this short video…

Desktop Companion Released

Conveniently manage AWS Measuring Agents from your desktop and more…

The Desktop Companion is a Desktop GUI that allows you to manage several Real Load aspects directly from your desktop.

It was freshly released in the last few days, plz do not hesitate to try it out. We’ve put together a short video that shows how to:

  • Prepare a simple load tests using the Recording Proxy on your Desktop.
  • Upload the load test to the Real Load Portal
  • Start an AWS EC2 Measuring Agent (Load Generator) from the Desktop Companion
  • Execute the load test script
  • Terminate the AWS EC2 Measuring Agent

All of this in 8 minutes…

It’s the first video I’ve ever had to publish, so I apologize in advance for the rather basic editing…

Real Load Plugins Introduction

Real Load plugins - Create, share or simply re-use.

A great new feature of the Real Load portal is that is allows you to share or simply consume plugins that have been prepared by others.

Plugins are written in Java. There are 3 types of plugin that are supported by the Real Load application:

  1. Session Element Plug-In - Typically used to generate custom data required by your load test script. For example:
  • Extract data from a DB.
  • Generate random data that follows a specific syntax.
  • Query an external webservice to obtain data to be injected in the load test.
  1. URL Plug-in - Allows you to modify request or response data:
  • Modify the HTTP request (…change the URL, etc…)
  • Modify response data.
  1. Java Source Code Modifier Plug-in - Allows to automtically modify a test script Java source code.

One of the key fetures of the product is that plugins can be optionally be published on the Real Load portal, for other users to consume. You can have a glimpse of available plugins here.

Interested in plugins but don’t know where to start? We’ll soon publish a getting started documentation on our website. In the meantime please reach out to us at

Real Load Demo Portal online

The Real Load Demo Portal is now up and running.

The demo portal is now available for selected customers who wish to evaluate the product functionality. You need an invitation code from us to sign up at the portal.

Quick Start Guide

  1. Navigate to and click at “Sign up”
  2. Enter your invitation code and follow the instructions
  3. Once you are signed in navigate to “Measuring Agents”
  4. Add the following Measuring Agent: port 8080
  5. Ping the Measuring Agent at application level
  6. Click at the “Wizards” icon to “HTTP Test Wizard”
  7. Define your first HTTP/S test, debug the test, save the session, generate the code and run your test

Note: The Measuring Agent has the following restrictions:

  • Maximum number of users per test job: 500
  • Maximum test job duration: 5 minutes

“alt attribute”

“alt attribute”

“alt attribute”