IBM Support in Austria After End of Life

The Layer Between Two Owners

Servers have a lifecycle owner and storage has a lifecycle owner. The SAN fabric between them frequently has neither. It was bought alongside the array, installed once, and is not reviewed by the server team because it is not a server, nor by the storage team because it is not an array. It ages out without appearing in anybody's plan.

Its Failure Is Designed to Be Invisible

A correctly built estate runs two independent fabrics, so losing a switch removes one path and nothing stops. That is the design working. It is also why the failure can persist unnoticed for months: no application is affected, no alert reaches anybody who would act, and from every dashboard reporting service health the estate looks perfect.

A Dual Fabric That Has Become a Single Fabric

This is the real exposure, and it is a state rather than an event. The estate is now one failure from losing storage access entirely, and nothing distinguishes it from a healthy dual-fabric estate except a check nobody is running. The next switch failure is the one that stops the business, and it will look sudden.

Inexpensive Parts, Disproportionate Consequence

A switch power supply and the optics on active ports are the common causes of a lost path, and both are cheap relative to what they protect. ICD supplies fabric switch hardware, optics, Storwize components and Power parts for generations past IBM end of support, tested and covered for one to three years.