Application life cycle management

[Rob Beekmans]

Application life cycle management 

Applications are the gateway to data employees are working with. Proper functioning of applications is key for any organization and issues with applications will be felt instantly. It is therefore important that applications are patched, updated and maintained to ensure they are secure, performant and available. 

Application life cycle management is key in this, you do not want applications patched in a production environment without proper testing. Recently two incidents took place where this occurred, and the result were outages that span hours to days. 

Application and image management requires rigorous testing, multiple phases of testing to determine if an updated version or patch is working as expected. Skipping these phases and deploying straight into production is like jumping off a cliff blindfolded, it may work but there comes a day when there is no water down the cliff. 

How does this application life cycle work?

Applications go through a life cycle; the first phase is the Develop phase where technical tests are conducted, the application is installed by an admin. Standard technical test to see if it opens, and no errors popup and it does not break other applications. The development phase is when the application is introduced in the organization. The Development phase is seen as a playground environment, one to get to know the application and its quirks.

The second phase of life cycle management is the Testing phase where the application, update or patch is evaluated against some dummy data (or an old backup of the production data). Simple queries, a printout, opening reports are done here. If all goes well it is handed over to the acceptance phase.

The third phase of life cycle management is the Acceptance phase, I would say the most crucial phase of life cycle management. Here key users of the application use a complete functional test plan to evaluate all functionalities used during the day. Only when this phase is signed off, is it released to production.

If all previous phases went well, the application, new, updated or patched rolls into Production. When patches or a latest version are released, it starts all over. The TAP process is ongoing for every application, all year long.

Application life-cycle management will look like a lot of effort “just” to update or patch an application. It is a lot of work, but it guarantees proper working of applications and ensures no loss of productivity when deploying an application, updated version or patched version. With the right test procedure and detailed functional test plans, it is a process to go through. Preparation is key in this, getting your test plans in order, it is one time write-up that will enhance your application management to Enterprise level. 

How often is a production environment image updated?

That is a relevant question when thinking about application life cycle management. Can you update the desktop image when a vendor releases an important update or patch, or do you have to wait until the end of the month for the image update?

 

Each dot on the chart is an update or patch, minor or important, functional, critical or security focused. No customer will update their image each time a patch is released for an application. That would be impossible as images contain dozens of applications at least if not close to hundreds. You’d be updating around the clock. 

The dilemma of every organization, how do I keep my application landscape up to date while not updating my image around the clock? 

App Volumes takes application management out of the image. 

With Omnissa App volumes, application management is no longer part of image management. The image can be managed on a regular pattern, bi-weekly, monthly, quarterly, whatever works best. Applications are managed and delivered by App volumes, On-demand even with Apps on-demand. Certain applications will be installed in the image, but the vast majority will be delivered on-demand.

Your application life cycle management is enhanced with App volumes. Within App volumes an application has four packaging stages. The app volumes packaging stages are named New, Tested, Published and Retired. You could overlay this with the application life cycle phases, App volumes “Tested” phase maps with the “Test” phase of application life cycle management. It is handed off to acceptance and once all tests are successfully completed, it is “Published”. At the end of each application life cycle phase an App Volumes stage starts, once the application has been through acceptance phase it will be published (App Volumes stage) which is a production phase until it is replaced by a recent version. When the recent version enters production, the outdated version will be Retired (App Volumes stage). The diagram below shows this process in more clarity. 

Older versions are retired but can still be assigned if needed, and a rollback is easy to do. App Volumes has a 99%+ compatibility record that is verified with customers who have over a thousand apps deployed with App volumes. 

Application life cycle management, unique in the market 

When we look at the VDI (Virtual Desktop Infrastructure) and DaaS (Desktop as a Service) solutions in the market, there is no equivalent to what Omnissa is offering. There are solutions that offer (near-time) application delivery or application layering but none of them have life cycle management functionality with their products. What that means is, admins need to keep track of testing and packaging phases themselves. Which is fine for one application but becomes a nightmare with hundreds or thousands of applications. 

Looking towards the broader EUC (End User Computing) market, outside the VDI/DaaS vendors, we see vendors with an application delivery solution but again they lack life cycle management features. 

Omnissa App Volumes is unique in this market for its life cycle management, on-demand delivery of applications and being part of a digital workspace that includes UEM (User Endpoint Management) to manage and deploy applications outside the VDI/DaaS realm. 

Benefit to TCO (Total Cost of Ownership) 

Besides the fact that you can patch faster and thus are less exposed to security threats (that should be enough reason alone), you will see a positive effect on your TCO. 

Think about it, with most applications outside of the image, you need fewer images, fewer desktop pools. Fewer images and pools relieve an admin of repetitive work. Admins can be assigned to other tasks than just managing desktop images or pools. 

Production is hindered by image updates for application patches. Every update of the image requires an update to the pools, no matter how you spin that, it impacts production. With non-persistent desktops it is easier perhaps but with persistent desktops it does impact on user experience. With the applications out of the image, the impact is minimal. It is a win-win situation, better security, better user experience, and better TCO. 

What is next?

If we extend application life-cycle management into modern management, the number of steps or phases extends. We define eight activities of modern management that correspond to application life-cycle management and extend them.

Eight Activities of modern management is something that requires a different blog on itself. It is something that levels up a customer application management and deployment. We will discuss this in a next blog post.

Summary  

App volumes and apps on-demand, available with most Horizon subscriptions, is a unique solution for organizations want to take application delivery and management to the next level. There is no other VDI (Virtual Desktop Infrastructure) or DaaS (Desktop as a Service) solution out there offering this level of management and delivery. 

If you’re a Horizon customer, you may already have access to the feature today and if not, check out Horizon by contacting us at https://www.omnissa.com/contact-us/.