Lenovo Support in Canada After End of Life

Redundancy Spent Is Not Redundancy Lost Loudly

A hot-swap fan fails in a ThinkSystem or System x chassis. The remaining fans compensate, the machine continues serving users without interruption, and a log entry is generated. Because nothing visible happened, the resulting ticket carries no urgency, and on a busy estate it can sit for months. The design worked exactly as intended, and the protection it provided has now been used up.

The Second Failure Behaves Completely Differently

A chassis down one fan has no margin left. When the next fan fails there is nothing to absorb it, so the outcome is throttling or a protective shutdown rather than another quiet log entry. The first failure was a warning delivered without consequence; the second arrives as an outage, and the interval between them is unknowable.

It Persists Because It Does Not Hurt

A failed drive in a protected array is the same situation — redundancy consumed, service unaffected — and it gets fixed promptly because it alarms loudly and everybody understands the array is exposed. A failed fan produces neither the noise nor the shared understanding, so it competes for attention against work that has visible consequences and reliably loses.

Treat It as an Open Incident

The useful change is classification rather than technology: a loss of redundancy is an incident that remains open until redundancy is restored, not an event that has been logged. Holding fan modules on site makes that a five-minute restoration rather than a procurement cycle. ICD supplies them across ThinkSystem and System x generations past vendor support, tested and covered for one to three years.