IBM today unveiled newly integrated versions of tools from its Rational and Tivoli stables, with the aim of permitting development and operations teams to work more closely throughout the application life cycle. Affecting nine products in all, the integrations marry tools from the test and development side to those used by IT staff for deployment and support, and are intended to improve efficiency for organizations using both. There...
were no announced price changes, and current licensees can upgrade at no additional cost.
"As we peel the onion, we discover that the developer is silos away from operations, and that doesn't work," said Dave Locke, director of products for IBM Rational. The problem, he said, is that work sometimes gets "lost in translation" as it travels from one part of the organization to another. Locke described an example in which Rational ClearQuest 7.1.1 is used by a development team to track changes and automate processes, and Tivoli Service Request Manager is the tool of choice for logging help requests.
"As service request tickets are written, they typically accumulate," Locke explained. "After some level of triage, they and handed over to development, where they have to be entered manually [into ClearQuest]."
The integration synchronizes the logging trouble tickets in Service Request Manager with the ClearQuest's bug-fix tracker, reducing human error and adding a new level of traceability to original service requests, Locke said. "They're populated automatically into ClearQuest, and handed to the development side. And they know know which developer wrote what code so [the ticket] can pop right into their queue. Nothing gets lost in translation," Locke said.
The next pairing -- Rational Asset Manager 7.2 with Tivoli Change and Configuration Management Database -- deals with application deployment, pre and post. Asset Manager keeps track of all the elements that go into development. "For example, developing for the Phone, you might have an iTunes server, which connects to credit card billing [app], which connects to music databases, which connects to a download server and then back to the desktop," Locke said. "That's a complex world with lots of assets brought to bear, each with its own release cycle. As a developer, I don't want to rewrite any of that code, and I do want to make sure I'm using the right pieces so that when we go live, we're using the latest of everything."
From the operations side, it's helpful for support teams to know which versions are current when trouble reports come in. "This integration is about forward and backward round-tripping. Knowing what I'm using to deploy to production, keeping track in an automated fashion of what you're using to build the software, where defects are coming in from and feeding those back to the development loop," Locke said. Of course, it's even better if you're also using an automated build system. "This automatically knows what versions of things are the most up to date and passes [the info] to everyone who needs it."
The integration of Rational Test Lab Manager 2.0 with Tivoli's Provisioning Manager and Application Dependency and Discovery Manager is a shot at achieving parity between development and production environments. When you set up a test lab, Locke said, you want the hardware, software, operating systems, browser, patches and configurations to match what's going on in production. "There are too many things to worry about to make sure it's set up the way the organization is going to use the software," he said.
"On the Tivoli side, it's about deployment to the real world." By marrying the two, according to Locke, data gathered by Tivoli in real time about what's going on in production can be fed back to the lab and used to configure images for provisioning to test systems. "It can be looped between deployed hardware, software and networking in real life operations and mimicked in the test lab without writing things down and without human error," he said.
Then there's Rational Performance Tester and Tivoli Composite Application Manager. To explain the integration, Locke describes a production environment in which users are complaining about application slowness (as if that's ever happened). "Your operations guy starts digging into alerts and finds nothing wrong with the network, so we want to tell development." But rather than simply throwing the problem over the transom, IT can peek into performance testing data and compare it with what they're hearing from users.
"Remember you're simulating the environment in the lab; production is always lightly different," said Locke. "This integration allows the Tivoli side to pass actual detail of discovery results back to functional and performance testing parts of test lab. We can now actually pass at the source code level."
The newly integrated tools are not intended as a monolithic application lifecycle management (ALM) suite, but rather are aimed at pockets of need. "Fifty percent of applications, once deployed, are rolled back so corrections can be made, and then rolled out again. That's expensive," Locke said. "We're proponents of better teams and better communications. And by integrating these applications, we're trying to get teams to work more effectively together."