Getting to grips with the legacy application challenge

null

The pace of legacy retirement is accelerating. In 2016, analyst group Gartner predicted that over the following four years IT organisations would decommission three times more applications than in the preceding six. Retiring legacy applications is viewed as a way to free up resources to drive digital transformation and other strategic initiatives. It also removes a thorn in the side of the IT department and a hindrance to innovation.   

The problems of application ‘life support’

The legacy applications that are ripe for retirement are not simply old. They are systems that are no longer being updated with live data, either because they have been replaced by new applications or are now surplus to requirements. Most companies have at least one of these, but some will have hundreds. 

Companies find it difficult to discard these systems because they hold valuable historical information. This information might be required for operational reasons, such as responding to customer queries, or for compliance – meeting a statutory obligation to retain customer data for a number of years, for instance. The data held within legacy systems is also a valuable store of potential business insights that can be mined using analytics tools and therefore many companies are understandably reluctant to lose it. 

As a consequence, many systems that have reached the end of their natural life are kept ticking over on ‘life support’. However, the costs, and risks, of keeping these applications running are significant. High support overheads and maintenance charges, access difficulties, compliance concerns and technical skills shortages are all problems associated with legacy applications. Older systems are more vulnerable to malware and other security threats, and are incompatible with modern security mechanisms.

The solution: move the data 

If the data is still required, but the application is not, the logical solution is to move the data somewhere else so that the application can be retired. 

But where to?

Migrating large volumes of legacy data to a business application that replaced the original legacy system is rarely a viable solution. While you may want to take current, live data across, pumping many years’ worth of historical data into a new system will likely impair system performance and incur significant additional disk storage and backup costs.

Writing data in a raw format to a database, file or offline storage is also problematic. Separated from the original application, the raw data will be out of context and therefore lose its meaning. Business users rely on applications to process and interpret data in order to make it meaningful. Applications perform logic, such as calculating totals, and make translations, such as replacing the number ‘1’ in a database with the words ‘monthly billing’ when displayed on screen, for example. 

Equally, users expect data to be laid out on screens in a format that is as easy to read, search through, and interpret as when they used the original application. Dumping the data out of the original system without retaining these basic capabilities is likely to create more problems than it solves. 

A different approach is required that caters for the long-term data access needs of business users. If done correctly, legacy application decommissioning achieves this. A structured process is followed in order to move all data required by the business to a specialist online content repository. Critically, the data is easily accessible, and application business logic is embedded so that it remains meaningful, with no prior knowledge required.  The original application can then be retired, together with any supporting hardware and software.

A business-led approach to application retirement

Application decommissioning is often viewed as an IT initiative, but the involvement of business stakeholders is key to its success. Business users know which data in the original legacy applications is important to retain, and how it will be used in the future. Creating a cross-functional team that includes business users as well as IT helps to de-risk the project by ensuring that all requirements for use of the data are considered, and encourages buy-in from colleagues who may – initially at least – have reservations about losing their old applications.

Running a decommissioning programme

Today’s new business app is tomorrow’s legacy system. Applications will naturally come to the end of their life over time as they are replaced by newer technology or processes. You should therefore aim to create a programme to identify any applications suitable for retirement and decommission them efficiently on an ongoing basis.

To this end, Andy Kyte at Gartner suggests that enterprises appoint a dedicated decommissioning manager, or ‘application undertaker’, to oversee the decommissioning programme, build skills and knowledge and maximise economies of scale. In fact Kyte believes the lack of such an undertaker is the reason why many legacy systems are kept running way beyond their usefulness as a live application. 

Similarly, when creating a decommissioning programme it is important to have a senior management sponsor. Raising decommissioning to a strategic level with authority behind it will make it easier to embed the process as a natural part of the application lifecycle – and will command the full and active participation of all the right people from across the business on every individual project.

Following a repeatable process

Following a structured, repeatable process minimises risk by ensuring that all requirements for ongoing use of legacy data are captured at the outset.  

Running a workshop with business users at the outset will help the project team to understand the key business requirements that must be met, including which data must be kept accessible (and which can be discarded) and how users navigate through the system. 

Once the initial requirements are defined, a detailed scoping phase can be conducted. This will identify where the data currently resides and the existing data formats that are available, such as database tables, documents, voice recordings and so on. It should also identify any logic that must be applied to the data to retain the correct context for business users (such as substituting numbers or codes with meaningful text), along with any compliance obligations, for example to retain information in an original document format. 

Before completing the all-important data extraction and switching off the live application forever, as for any IT project there should be an acceptance testing phase. This will ensure that end users are happy and are able to access all of the information they need.    

Nine key capabilities to look for in a decommissioning solution

The technology chosen for decommissioning is a critical success factor. It can be tempting to choose a content repository based on what works for the first application – or set of applications – being retired. However, if the first system is not flexible enough to accommodate all future application retirement projects, a single organisation could end up using a variety of different solutions. The headache of managing multiple, disconnected content repositories would cancel out many of the benefits of decommissioning applications in the first place, such as ease of accessibility and compliance, the ability to pool legacy data in one place, and reduced running costs.  

To ensure that a decommissioning solution will be suitable for the broadest possible range of business requirements it should meet the following nine criteria:

  • Accessibility: provides business users with fast, easy access to all of the information they currently obtain from their legacy applications, using modern access methods such as web and mobile interfaces 
  • Data reliability: retains legacy data in its original business context, without the need for additional knowledge to interpret its meaning in the future
  • Data integrity: ensures that the data provides an accurate business record by adopting tamper proofing and non-repudiation measures
  • Usability: is easy for workers around the business to use without technical skills or prior knowledge of the original system, and provides analysis capabilities to deliver business intelligence
  • Compliance and information lifecycle management: enforces corporate policies to conform with statutory requirements for data management, including retention, destruction, redaction, and legal hold.
  • Performance: scales up to handle large volumes of data from multiple decommissioned applications and to provide access to as many people as required, with fast response times
  • Risk management: incorporates effective processes and safeguards to eliminate the risk of any data loss or business disruption during or after decommissioning.
  • Low total cost of ownership: minimises expenditure over the long term by using hardware and software with low maintenance costs and by limiting the need for ongoing IT support and development 
  • Security: provides effective mechanisms to maintain a safe and secure environment, including managing user authentication, user access rights, data security and data segregation.

Finding an effective way to manage the problem of an ever increasing portfolio of legacy applications through to end of life is an important challenge that affects most enterprises. The best long-term solution is to establish an ongoing decommissioning programme that will both raise corporate awareness of the benefits of retiring legacy applications, and make the decommissioning process itself easier by embedding it as an essential part of good IT practice.

Lynda Kershaw, Marketing Manager at Macro 4, a division of UNICOM® Global 

Image Credit: TechnoVectors / Shutterstock