Application maintenance keeps legacy systems serviceable

IT operations needs tools to ensure that it can manage old -- maybe ancient -- applications while deploying new apps with different structures.

The business world is disposable. What was hot and fresh in the morning becomes cliché by lunchtime and gone by dinner. But few things ever die off.

IT operations managers keeping pace with the latest and greatest also have legacy applications running for years or decades after install. The enterprise needs responsive, effective application maintenance tools so these systems perform up to standard without pulling operations staff away from new projects. Many firms need application maintenance and monitoring tools that work with modern apps as well.

Companies should start looking for solid operations tools at the well-established system management suppliers, including BMC Software, CA Technologies, Compuware, Hewlett Packard Enterprise (HPE), IBM and Oracle.

Mainframes keep humming

Mainframe systems continue to support many business apps. The baggage of legacy application software and add-ons does not migrate easily to other architectures. Mainframes also offer high-end performance, and companies are slow to change with this investment already made. Consequently, these systems hum along in corners of the data center supporting important legacy applications such as airline reservations, banking and retail systems. More than half (55%) of enterprise applications call upon the mainframe to complete transactions and 89% of CIOs stated that mainframe workloads are increasing and getting more varied, according to a Vanson Bourne survey.

Although new automation tools, such as Puppet, can automate traditional system management functions, employment listings site shows about the same number of open jobs seeking Puppet skills as open jobs for mainframe techs.

Mainframe operations managers feel as comfortable looking at text-based management screens as most ops pros do looking at fancy graphical user interfaces. But management tools need to marry the new with the old, and work with the multiple mainframe operating systems from Linux and Unix to IBM's z/OS.

IBM has a broad suite of management tools for its mainframe systems. The IBM z/OS Management Facility is browser-based and monitors mainframe performance. The tool features dashboards, wizards and interactive help to increase administrators' productivity. For traditionalists, the tool includes a green screen interface emulator.

To help manage modern applications, the IBM Tivoli OMEGAMON XE on z/OS Monitoring Feature for JVM provides resource-level monitoring of Java virtual machines. IBM Wave is built for mainframes running Linux and IBM z/VM. The browser-based GUI provides a view spanning partitions, servers, sites and geographies, and enables monitoring, provisioning, and management of projects and virtual servers. IBM Wave even includes a drag-and-drop capability that moves Linux virtual servers around on z/VM systems.

Compuware developed the Topaz platform with a GUI for managing IBM System z mainframe systems. Topaz is based on the open source Eclipse framework, a popular development and runtime environment created by IBM. Topaz maps graphically oriented interfaces onto the traditional green screen, so IT pros can manage the systems more easily. The Topaz visualizer shows relationships among different data objects, and it can display all the columns within a database table and highlight the connections one table has with other tables. This approach replaces the need to internally memorize database table structures and interact with the data through a series of commands.

System management has long been mundane work, requiring a lot of hands-on testing. Recently, the focus has shifted to more adding intelligence and automation to performance maintenance. For instance, CA Unified Infrastructure Management includes enhanced analytics, alerting and cloud monitoring capabilities. The analytics features a time to threshold function that predictively calculates and identifies potential performance threats, so ops staff can correct them before users see an effect. In addition, a time over threshold measurement enables IT staff to identify persistent performance issues to address.

Make way for multiple server technologies

Like mainframes, Unix servers often operate in multi-technology environments, hosting legacy apps alongside modern applications on hardware and in the cloud. Application maintenance tools must support a range of hardware operating systems, not only UNIX and Linux but also Microsoft Windows Server. Given the dispersed nature of processing, system techs are looking at performance on an end-to-end basis rather than at each element. As a result, these management tools need visibility into a wide range of hosts.

Oracle is leading supplier of UNIX hardware and also has a robust line of management tools. The Oracle Single System Management line includes tools for monitoring and configuring Oracle's SPARC Enterprise T-Series servers, Sun Fire x86 servers, blades and their chassis. All of these systems have Oracle Integrated Lights Out Manager service processors, which provide in-band and out-of-band interfaces for collecting monitoring data. With these tools, operations staff performs remote hardware monitoring, error and fault management; management of user accounts, and altering.

HPE IMC Application Performance Manager Software monitors applications running on a variety of systems, including UNIX, Linux, Windows, database, application, Web, mail, Web services and LDAP servers. The product streamlines application maintenance by consolidating traditional application monitoring with network monitoring via a common interface. This combination lets it provide a visualization of traffic flows, outlines performance characteristics, analyzes network health and notifies IT administrators of any hot spots or bottlenecks. The fault management functions include a user selectable alarming threshold that triggers automatic problem notification.

IT operations managers often scurry to put tools in place to manage the latest technical solutions. Even as they focus on those tasks, they must recognize that legacy applications on less-popular operating systems support important applications and put solutions in place to monitor those systems as well.

Next Steps

A guide to modernizing your apps

How to build an APM strategy for the app economy

Application performance monitoring tools get a workout

Dig Deeper on Application Maintenance on Production Systems