The next BriefingsDirect Voice of the Customer performance engineering case study discussion examines how JetBlue Airways in New York uses virtual environments to reduce software development costs, centralize performance testing, and create a climate for continuous integration and real-time monitoring of mobile applications.
We’ll now hear how JetBlue cultivated a DevOps model by including advanced performance feedback in the continuous integration process to enable greater customer and workforce productivity.
To describe how efficient performance engineering has reduced testing, hardware and maintenance costs by as much as 60 percent, we’re joined by Mohammed Mahmud, the Senior Software Performance Engineer at JetBlue Airways in New York. The discussion is moderated by BriefingsDirect’s Dana Gardner, Principal Analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: Why is mobile user experience so very important to your ability to serve customers well these days?
Mahmud: It’s really very important for us to give the customer an option to do check-in, book flights, manage bookings, check flight status, and some other things. On flights, they have an option to watch TV, listen to music, and buy stuff using mobile devices as well. But on board, they have to use Fly-Fi [wireless networks]. This is one of the most important business drivers for JetBlue Airways.
Gardner: What sort of climate or environment have you had to put together in order to make sure that those mobile apps really work well, and that your brand and reputation don’t suffer?
Mahmud: I believe a real-time monitoring solution is the key to success. We use HPE Business Service Management (BSM), integrated with third-party applications for monitoring purposes. We created some synthetic transactions and put them out there on a real device to see … how it impacts performance. If there are any issues with that, we can fix it before it happens in the production environment.
Also, we have a real-time monitoring solution in place. This solution uses real devices to get the real user experience and to identify potential performance bottlenecks in a live production environment. If anything goes wrong there, we can get alerts from the production environment, and we can mitigate that issue right away.
Gardner: How have you been able to connect the development process to the operational environment?
Mahmud: My area is strictly performance engineering, but we’re in the process of putting the performance effort into our DevOps model. We’re going to be part of the continuous integration (CI) process, so we can take part in the development process and give performance feedback early in the development phase.
In this model, an application module upgrade is kicking off the functional test cases and giving feedback to the developers. Our plan is to take part of that CI process and include the performance test cases to provide performance feedback in the very early stage of the development process.
Gardner: How often are you updating these apps? Are you doing it monthly, quarterly, more frequently?
Mahmud: Most of them on a two- or three-week basis.
Gardner: How are you managing the virtual environment to create as close to the operating environment as you can? How do the virtualized services and networks benefit you?
Mahmud: We’re maintaining a complete virtualized environment for our performance testing and performance engineering. Before our developers create any kind of a service, or put it out there, they do mock-ups using third-party applications. The virtual environment they’re creating is similar to the production environment, so that when it’s being deployed out there in the actual environment, it works efficiently and perfectly without any issue.
Gardner: How has performance engineering changed over the past few years? We’ve seen a lot of changes in general, in development, mobile of course, DevOps, and the need for more rapid development. But, how have you seen it shift in the past few years in performance engineering?
Mahmud: When I came to JetBlue Airways, LoadRunner was only one product they had. The performance team was responsible for evaluating the application performance by running a performance test and give the test results with identifying pass/fail based on the requirements provided. It was strictly performance testing.
The statistics they used to provide were pretty straightforward, maybe some transaction response times and some server statistics, but no other analysis or detailed information. But now, it’s more than that. Now, we don’t just test the application and determine the pass/fail. We analyze the logs, traffic flow, user behavior, and how they behave, etc. in order to create and design an effective test. Now, this is more performance engineering than performance testing.
Early in the cycle
We’re getting engaged early in the development cycle to provide performance feedback. We’re doing the performance testing, providing the response time in cases where multiple users are using that application or that module, finding out how this is going to impact the performance, and finding bottlenecks before it goes to the integration point.
So, it’s more of coming to the developers’ table, sitting together, and figuring out any performance issue.
Gardner: Understanding the trajectory forward, it seems that we’re going to be doing more with microservices, APIs, more points of contact, generating more data, trying to bring the analysis of that data back into the application. Where do you see it going now that you’ve told us where it has come from? What will be some of the next benefits that performance engineering can bring to the overall development process?
Mahmud: Well, as I mentioned earlier, we’re planning to be part of the continuous integration; our goal is to become engaged earlier in the development process. That’s sitting together with the developers on a one-to-one basis to see what they need to make sure that we have performance-efficient applications in our environment for our customers. Again, this is all about getting involved in the earlier stages. That’s number one.
Number two, we’re trying to mitigate any kind of volume-related issue. Sometimes, we have yearly sales. We don’t know when that’s going to happen, but when it happens, it’s an enormous pressure on the system. It’s a big thing, and we need to make sure we’re prepared for that kind of traffic on our site.
Our applications are mostly JetBlue.com and JetBlue mobile applications. It’s really crucial for us and for our business. We’re trying to become engaged in the early stages and be part of the development process as well.
Gardner: Of course it’s important to be able to demonstrate value. Do you have any metrics of success or you can point to ways in which getting in early, getting in deep, has been of significant value? How do you measure your effectiveness?
Mahmud: We did an assessment in our production environment to see, if JetBlue.com goes down for an hour, how much it’s going to cost us? I’m not authorized to discuss any numbers, but I can tell you that it was in the millions of dollars.
So, before it goes to production with any kind of performance-related issue, we make sure that we’re solving it before it happens. Right there, we’re saving millions of dollars. That’s the value we are adding.
Gardner: Of course more and more people identify the mobile app with the company. This is how they interact; it becomes the brand. So, it’s super important for that as well.
Mahmud: Of course, and I can add another thing. When I joined JetBlue three years ago, industry standard-wise our position was bottom on the benchmark list. Now, we’re within the top five in the benchmark list. So, we’re adding value to our organization.
Gardner: It pays to get it done right the first time and get it early, almost in any activity these days.
What comes next? Where would you like to extend continuous integration processes, to more types of applications, developing more services? Where do you take the success and extend it?
Mahmud: Right now, we’re more engaged with JetBlue.com and mobile applications. Other teams are interested in doing performance testing for their systems as well. So, we’re getting engaged with the SAP, DB, HR, and payroll team as well. We’re getting engaged more day by day. It’s getting bigger every day.
You may also be interested in:
- Strategic DevOps—How advanced testing brings broad benefits to Independent Health
- How always-available data forms the digital lifeblood for a university medical center
- How flash storage provides competitive edge for Canadian music service provider SOCAN
- Loyalty management innovator Aimia’s transformation journey to modernized IT
- Infrastructure as destiny — How Purdue builds an IT support fabric for big data-enabled IoT
- Cybersecurity crosses the chasm: How IT now looks to the cloud for best security
- Big data and cloud combo spark momentous genomic medicine advances at HudsonAlpha
- How software-defined storage translates into just-in-time data center scaling and hybrid IT benefits
- Securing data provides Canadian online bank rapid path to new credit card business
- How Allegiant Air solved its PCI problem and got a whole lot better security culture, too
- How European GDPR compliance enables enterprises to both gain data privacy and improve their bottom lines
- DevOps and security, a match made in heaven