Delivering better mobile performance

Authors:  Mark Tomlinson and Esther Levine

Of all the reasons for using Shunra technology together with HP LoadRunner or Performance Center, none is more powerful than when you need to accurately simulate a mobile end-user experience.  Yes, it’s true that through all the years we’ve posted many blogs and videos about “real-world performance testing” using the integration between the HP and Shunra products.  Recreating actual network conditions allows for accurate assessment of impact on total response time and provides for precise analysis of response time across the client, network and server components. However, when it comes to the real world of a mobile end-user, the value and importance of emulating the network conditions is vastly more critical to the accuracy of your test design and results. 

 

Here are the basic concepts that make the network conditions for mobile performance different from other network simulations:

 

Latency:  Mobile is synonymous with high latency.  If you read nothing further, you should understand that this is the single most important network issue for mobile end-users.  Latency (a.k.a. delay) negatively impacts not only application response times, but also limits the multi-tasking ability on the device by making each dependent component request highly latent.  To make matters even worse, as the mobile devices moves location from one tower or access point to the next, the latency changes – it’s dynamic.  Your mobile application should be prepared for that, especially on synchronous UI interactive controls.

 

Loss:  When it comes to loss, there’s more than just packet loss to consider with mobile.  There’s connection loss – at the physical layer connecting the phone to the tower (LTE).  If you’ve ever been happily downloading data and then “poof” the download aborts – that’s connection loss.  In addition, consider connection switching, which is not the same as connection loss.  The networking component of the mobile operating system has more elaborate logic for handling periodic loss of connections and thus switching of connections as you move physically between towers.  When you are roaming, switching connection between carriers or providers occurs and that means that your applications must also perform well when these connection issues materialize.

 

Bandwidth:  The maximum throughput to a mobile device is dynamic, just like latency.  However, keep in mind that as we put more and more powerful processors into mobile devices, we will increasingly try to conduct more and more simultaneous activity.  That results in more simultaneous connections to the back-end services, which are required to keep the apps running smoothly.  The realistically available bandwidth is what you must consider – don’t just expect your end-users to always have a 4G or WiFi connection.  Even if they do have high-bandwidth, your app might not be the only one trying to use it at the same time.

 

Those are the three essential parts to the virtualized mobile network that you must consider as part of your testing, to ensure the test to be accurately designed and automated.  And there are several ways to combine Shunra products together with HP Software.

 

 

For HP LoadRunner and Performance Center:  we have a product called Shunra for HP Software which installs Shunra technology on the Load Generators and the Controller machines.  This integration has been built-in to HP LoadRunner and Performance Center for the last three years and is a mutually-developed integration that is built and tested jointly by Shunra and HP Software R&D.  The value of this solution is that you can easily add real-world mobile conditions to your existing test suites without re-coding or completely reconfiguring your test lab.

 Shunra Screen Shot.png

 

 

Using the HP Analysis available in HP LoadRunner together with the in-depth reports in Shunra’s Analytics, you’ll be saving time and effort in diagnosing and resolving performance bottlenecks or transactions that violated established response time goals.

 

For HP Service Virtualization:  if you are simulating an external service or server in the test environment, you should consider a few different options to be fully deployed with Shunra in the lab.  First, if you have just a single SV box set up, you could install the Shunra vCat product on the SV host and emulate the real-world mobile conditions for the different services that are being simulated on that box.  Secondly, if you have multiple different services used in a test for variable latency or switching, consider the Shunra Performance Suite (hardware appliance) as the main routing solution for your test environment.

 

For HP Functional Testing and Unified Functional Testing:  the Shunra vCat product can be installed on each machine where QTP is installed, and the virtual network emulation can be automated programmatically via the simple API calls or via the GUI interface.  That will enable you to integrate real-world mobile testing conditions on the network interface of the QTP workstation or virtual machines.”

 ----

We're also interested in hearing from you... and your experience with mobile testing.  If you have a few minutes, please share your perspective in our BRIEF Mobile Research survey.

  

If you’re interested in HP LoadRunner or Performance Center you can follow us - on the HP Performance Validation Blog  or even follow us @hploadrunner on Twitter

And if you’re looking for functional testing – you need to check out The Future of Testing Blog

Of course- go to www.hp.com/go/mobiletesting for all things related to HP mobile testing

Comments
Vera Obonyo(anon) | ‎01-18-2012 12:55 AM

if you are simulating an external service or server in the test environment, you should consider a few different options to be fully deployed with Shunra in the lab.  First, if you have just a single SV box set up, you could install the Shunra vCat product on the SV host and emulate the real-world mobile conditions for the different services that are being simulated on that box.  Secondly, if you have multiple different services used in a test for variable latency or switching, consider the Shunra Performance Suite as the main routing solution for your test environment.

 

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
Showing results for 
Search instead for 
Do you mean 
About the Author


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation