Posts Tagged DevOps

New Enterprise Systems Maturity Model

Does your shop use maturity models to measure where you stand and where you should be going compared with industry trends and directions? Savvy IT managers often would use such a model to pinpoint where their organization stood on a particular issue  as part of their pitch for an increased budget to hire more people or acquire newer, faster, greater IT capabilities.

Today maturity models are still around but they are more specialized now. There are, for instance, maturity models for security and IT management. Don’t be surprised to see maturity models coming out for cloud or mobile computing if they are not already here.

Earlier this year, Compuware introduced a new maturity model for the enterprise data center.  You can access it here. Compuware describes the new maturity model as one that helps organizations assess and improve the processes for managing application performance and costs as distributed and mainframe systems converge.

Why now? The data center and even the mainframe have been changing fast with the advent of cloud computing, mobile, and big data/analytics. Did your enterprise data center team ever think they would be processing transactions from mobile phones or running analytic applications against unstructured social media data? Did they ever imagine they would be handling compound workloads across mainframes and multiple distributed systems running both Windows and Linux?  Welcome to the new enterprise data center normal.

Maybe the first difference you’ll notice in the new maturity model are the new types of people populating the enterprise data center. Now you need to accommodate distributed and open systems along with the traditional mainframe environment. It requires that you bring together completely different teams and integrate them. Throw in mobile, big data, analytics, and social and you have a vastly different reality than you had even a year ago.  And with that comes the need to bridge the gap that has long existed between the enterprise (mainframe) and distributed data center teams. This is a cultural divide that will have to be navigated, and the new enterprise IT maturity model can help.

The new data center normal, however, hasn’t changed data center economics, except maybe to exacerbate the situation. The data center has always been under pressure to rein in costs and use resources, both CPUs and MIPS, efficiently.  Those pressures are still there but only more so because the business is relying on the data center more than ever before as IT becomes increasingly central to the organization’s mission.

Similarly, the demand for high levels of quality of service (QoS) not only continues unabated but is expanding. The demand for enterprise-class QoS now extends to compound workloads that cross mainframe and distributed environments, leaving the data center scrambling to meet new end user experience (EUE) expectations even as it pieces together distributed system QoS work-arounds. The new enterprise IT maturity model will help blend these two worlds and address the more expansive role IT is playing today.

To do this the model combines distributed open systems environments with the mainframe and recognizes different workloads, approaches, processes, and tooling. It defines five levels of maturity: 1) ad hoc, 2) technology-centric, 3) internal services-centric, 4) external services-centric, and 5) business-revenue centric.

Organizations at the ad hoc level, for example, primarily use the enterprise systems to run core systems and may still employ a green screen approach to application development. At the technology-centric level, there’s an emphasis on infrastructure monitoring to support increasing volumes, higher capacity, complex workload and transaction processing along with greater MIPS usage. As organizations progress from internal services-focused to external services-focused, mainframe and distributed systems converge and EUE and external SLAs assume a greater priority.

Finally, at the fifth or business centric level, the emphasis shifts to business transaction monitoring where business needs and EUE are addressed through interoperability of the distributed systems and mainframes with mobile and cloud systems. Here technologies provide real-time transaction visibility across the whole delivery chain, and IT is viewed as a revenue generator. That’s the new enterprise data center normal.

In short, the new enterprise maturity model requires enterprise and distributing computing come together and all staff work together closely; that proprietary systems and open systems interoperate seamlessly. And there is no time for delay. Already, DevOps, machine-to-machine computing (the Internet of Things), and other IT strategies, descendents of agile computing, are gaining traction while smart mobile technologies drive the next wave of enterprise computing.

, , , , , , , , , , , , , , , , , ,

Leave a comment

Cloud Management Platforms (CMP) Come of Age

Every organization deploying applications to the cloud will want CMP, advises Gartner in its Cloud Management Platforms, 2013 report.

In a different Gartner report, titled Cool Vendors in Cloud Management, the researchers note: As enterprises deploy private clouds, interest in managing them is rising. Innovative CMP vendors—what Gartner dubs Cool Vendors—can help manage infrastructure resources to deliver service quality, security, and availability for workloads running in cloud environments.

“The benefits become obvious as enterprises start to deploy serious business workloads in the cloud, and these benefits all address bottom line concerns,” says Steven Henning, CMO, ServiceMesh.  The benefits revolve around grabbing new market opportunities fast and accelerating the introduction of new business capabilities that bring in customers and revenue.  ServiceMesh is one of the four cool CMP vendors named in Gartner’s report along with Cloupia, Eucalyptus, and Nimbula.

For DevOps to really hum, particularly when working at the speed of the cloud, it must work with IT to provision policies and governance across the software development lifecycle to create a common application platform and resources that support any language, streamline the entire application lifecycle, and generate market-ready applications faster than competitors. That’s why you want CMP technology.

CMP helps IT too. Often IT overhead and infrastructure management inhibit the allocation of developer resources and slow the organization’s ability to drive innovation at the speed of market demand. But by provisioning applications across the application lifecycle with pre-configured IT production-like containers for the application code, both IT and developers bypass the bottlenecks that occur at each phase of the software development lifecycle and speed delivery of applications to the cloud.

As 451 Research notes: Working where application development meets IT operations, DevOps teams moving to the cloud need tools to design application architecture, deploy it into production, and prevent unauthorized configuration drift.

451 Research was particularly impressed with ServiceMesh’s Agility Platform. Designed as an enterprise-class CMP, Agility embodies ServiceMesh’s conviction that platforms and applications drive business value, not infrastructure. Accordingly, the company believes that cloud governance has to be both application-centric, for today’s needs, and extensible, for the needs of tomorrow. Not surprisingly, the company is a major advocate of collaboration between application development teams and IT operations teams.

The 451 researchers observed that Agility Platform’s existing policies, event correlation capabilities, and application context awareness can all be harnessed to advance automated configuration management. It also plays nicely with the in-place CMDB, Puppet Master Server, and any other third-party tools.  In short, it is designed to address exactly the needs of enterprise DevOps.

Gartner, too, liked Agility Platform:  It identifies ServiceMesh’s strengths as a strong policy engine and lifecycle management. Furthermore, it is differentiated by connectors to third party providers and integration with other management systems. Finally Agility Platform provides policy-driven management, governance, security, and lifecycle management capabilities for private or hybrid cloud initiatives;  enabling IT organizations to manage all phases of the cloud computing lifecycle including the plan, build, and run stages by leveraging its policy-based framework. In addition, ServiceMesh’s cloud-native architecture can expand automatically through horizontal scaling to handle increasing system demands, according to Gartner.

But as Gartner noted, there are three others among its cool vendors in cloud management:

Cloupia offers its Unified Infrastructure Controller (UIC) as its cloud management platform (CMP). The UIC is software that enables the management of both private and public cloud services. It supports a self-service interface, workflow engine, and an orchestrator for virtual machine (VM) and physical infrastructure provisioning, Gartner reports.

Eucalyptus is a little different by providing Infrastructure-as-a-Service (IaaS) capabilities through an open-source software-based infrastructure for implementing on-premises clouds on existing infrastructure. The company’s strategy, according to Gartner, is to remain an IaaS platform, continue to advance its functionality, and build out a partner ecosystem to offer a more complete cloud solution.

Nimbula offers Director 2.0, which provides capabilities to build, manage, scale, and secure cloud environments across on-premises pools of virtualized IT resources as well as resources in public clouds, such as Amazon’s EC2. Like most of the others it uses the representational state transfer (REST) API to interface with cloud resources—both on premise and in public clouds—and it provides automation to manage the allocation of those resources, similar to ServiceMesh’s policy-based automation engine.

No vendor provides a complete CMP solution yet although ServiceMesh comes close. Still, you may need additional tools or may have to integrate multiple CMP approaches. But the CMP market is evolving fast; if you wait six months, each product will be better. Just don’t wait too long or you may miss fleeting opportunities. Tools like these enable DevOps to move that fast.

, , , , , , , , ,

1 Comment