IT consolidation is not just about server consolidation. The scope of these cost-saving projects can be very broad, and can involve anything from a small number of devices to multiple data centers. The benefits can be outstanding as well; some consolidation projects have resulted in a 15 to 1 reduction in the number of mail servers
or the reduction of 155 to just 12 major datacenters worldwide
A significant wave of IT consolidation has already occurred due to the popularity and usefulness of virtualization technologies that started in mainframes and mid-sized systems. Virtualization has now spread effectively to most hardware platforms including desktops, laptops and networking devices like firewalls and load balancers. A new wave of consolidating images to larger centralized systems designed to reduce costs and simplify management is also in progress. This next step in consolidation at the image level is enabled by the new economies of scale possible with mainframe and large-scale distributed computers.
Consolidation to Larger Centralized Systems
Server and storage virtualization are important consolidation building blocks. When consolidating production workloads, it’s best to utilize virtualization in a robust enterprise environment. For example, the Linux OS benefits from the mainframe’s capabilities and strengths, including strong data security, operational efficiency, high availability and striking performance. For IT consolidation projects, the Linux OS on IBM Z can result in data center simplification, more reliable operation and lower costs.
Tools are essential to the success of any consolidation project. Not only are they crucial in managing the up and running system—they’re also very important for sizing the newly consolidated environment. For server consolidation, the IBM Systems Workload Estimator, which is a web-based sizing tool for IBM Systems, is essential for completing early project sizing and planning. The scope of support includes IBM Z, Power Systems, IBM Flex Systems and IBM PureFlex Systems.
You can usethis tool to size a new system, to size an upgrade to an existing system or to size a consolidation of several systems. WLE will characterize the projected workload using either your measurement data or one of the many workload plug-ins. These plug-ins provide the necessary sizing guidance. Virtualization should be reflected in the sizing to yield a more robust solution, by using various types of partitioning and virtual I/O. WLE will provide current recommendations, along with growth recommendations for processor, memory and disk—whether internal or SAN—that satisfy your overall performance requirements. WLE can support estimates dealing with multiple systems, multiple partitions, multiple OSes and multiple sizing intervals. It’s difficult to imagine a more robust tool for consolidations to IBM systems.
Migration to Cloud for Advanced Automation and Workload Migration
Another part of this wave—something that’s best implemented in conjunction with larger centralized systems—is cloud services. Leveraging the capabilities of virtualization in the content of cloud services, like a mainframe private Linux cloud, does something well beyond the basics of virtualization. Automation—such as image provisioning and standardization using predefined and standard software imagesffers a strong starting point for lowering costs and speeding up day-to-day processes. Migration to the cloud is also effective as part of workload migration efforts. These consolidations can take place involving private, public and hybrid implementation projects for a variety of production and non-production workloads. Once the workload migration takes place, the door is open to exploiting fundamental cloud mechanisms that likely haven’t been used before, such as automatically starting (or shutting down) instances based on actual workload.
Provisioning and Management Is a Key Cloud Capability
Cloud provisioning and management products compliment the WLE by performing functions after the systems are configured and ready to be operated. Cloud computing is a modernization innovation for both z/OS and Linux. IBM Cloud Provisioning and Management for z/OS has features that focus on five important areas:
- Doing things on demand quickly and with fewer errors. On-demand self service provisioning of z/OS software makes available repetitive and automated processes to z/OS application developers to provision z/OS software. Application programmers use a z/OSMF graphical web-based user interface to perform these functions.
- Better process for handling computing resources. Improved control for users and system programmers eliminates the need for system programmers to be in the main path of provisioning. Users drive provisioning and obtain z/OS computing resources using actions that have been pre-approved by system programmers.
- Use of service catalogues to support elasticity and tracking. Catalog creation and resource tracking is used to create catalogs with customized z/OS software subsystem services that make use of the multi-tenancy and rapid elasticity of IBM z/OS.
- Predefined policies and resource pools for quicker allocation. Policy control uses predefined policies and resource pools to allow traditional z/OS system programmers to maintain control of system resources, and no longer requires their presence during the allocation process.
- Workflows to get things done quickly. Workflow authoring is used to build new or modify previously created workflows. These workflows are a crucial tool for user productivity in these cloud environments.
As we know, virtualization is a great tool, but combining it with cloud features like those found with IBM cloud provisioning and management amplifies its power. Most of these cloud features are significant new ideas that can have an important impact on the speed and reliability of the work being performed.
Changes in Consolidation Tactics Continue Through Modernization
The technical implementation of consolidation projects has changed considerably over the last decade. Ten years ago, companies were deploying virtualization as a consolidation tactic. They started with servers and extended that idea to any device (think networking) that could be virtualized to increase utilization and usefulness while lowering implementation and ongoing costs.
Today, centralizing images to IBM Z servers and including cloud-services software are major factor in new consolidation and modernization activities. It’s not just virtualization, but centralization and cloud, that impact these outcomes, and those projects are showing that the effort is definitely worth it.