everythingpossible - Fotolia


How mainframe maintenance intersects with x86 server upkeep

Mainframe maintenance overlaps with some x86 server upkeep tasks. Patching and storage management play a role in keeping mainframes up to date.

Mainframes and x86 systems are different, but if you're responsible for managing both, take heart; there's some overlap between mainframe and x86 server maintenance.

Patch management is one of the more common mainframe maintenance tasks. You can't use Windows Server Update Services or similar PC-based tools to handle mainframe patch management, but you can mimic the patch management policies that you use on PCs and x86 servers.

For example, if your internal patch management policy mandates that you apply patches to production systems within 14 days of release, you can apply this to your mainframe management workflow.

Some organizations retain old mainframe libraries in case they have to run an older version of the software following a restore operation. Because these old libraries are not always up to date, top mainframe provider IBM cautions organizations to make these libraries non-executable. This way, an attacker cannot access an old library to exploit a long-patched vulnerability.

Storage for mainframe maintenance

Another common maintenance task for both x86 servers and mainframes is storage management. On IBM's z/OS, the storage management subsystem handles any storage management tasks. When you need to allocate storage to a data set, you must specify the storage requirements and select a storage class, a management class and a data class.

A data class consists of a set of allocations and space attributes that are assigned to the data set. The storage class lists the performance and availability requirements for a given data set. Assuming the mainframe uses object storage, the storage management subsystem uses the storage class to pass information about object placement to the object access method (OAM).

The management class determines how you manage the data set and passes information to the OAM that pertains to data retention and backup requirements for mainframe maintenance.

Resource partitioning and virtualization

Many of the x86 servers in data centers act as virtualization hosts; you probably know how to provision resources to run virtualized workloads. Virtualization is not unique to x86 systems, however. Mainframes also support virtualization.

IBM mainframes have a feature known as the Processor Resource/Systems Manager. This feature enables the logical partitioning of the central electronics complex (CEC). In some documentation, the CEC is also referred to as the central processor complex. In other words, the Processor Resource/Systems Manager lets you divide the mainframe's CPU resources into a series of logical partitions, or virtual machines.

Logical partitions are often used to enable mainframe consolidation. An organization with several older mainframes can consolidate them into a single new mainframe, with each running in a separate partition. Like their x86 counterparts, each mainframe partition can receive dedicated CPU resources or share CPUs between multiple partitions.

The tool of choice for configuring partitions is the Hardware Management Console, which is useful for a variety of tasks beyond partitioning. You can use the console to perform root cause analysis of system-level failures and to assist in system firmware updates for mainframe maintenance. Likewise, the console can help you back up and restore configuration and customization data for hardware components.

Dig Deeper on Data center hardware and strategy

Cloud Computing
and ESG