SOLVED
Michael,
You won't find public MTBF or MTTR figures for either Proliant or Integrity servers. The reason is that it all depends on the actual configuration you have, so publishing a "typical" number is open to interpretation (usually by the competition).
If you were to pull out some of the advantages of Integrity over Proliant, they would be:
Dynamic Processor Resiliency - servers can detect some cases of failing CPU and take the CPU offline before the failure (HP-UX only I think)
Double-chip spare - RAID-like levels of protection for memory, without the overhead of extra memory.
Intel Cache-safe technology (disabling failed cache-lines in processors)
Enhancedd PCIe IO error recovery (failing PCIe cards can be isolated and prevent whole system for crashing)
If I were to sum all this up I'd say that a hardware error on an Integrity system is:
i) Less likely to result in a complete system failure.
ii) Crucially is more "diagnose-able" - i.e. it's generally possible to use the "Machine Check end" architecture in Itanium CPUs to gather enough data at the point of panic to diagnose what failed. I've seen x86 systems where intermittent hardware errors were only fixed by a "trial and error" approach to component replacement, due to a lack of meaningful data on the state of the system at the point of failure.
For more info/data, I'd suggest you contact a local HP or reseller rep.
HTH
Duncan
I am an HPE Employee
