associated with the identified memory region, the OS or hypervisor may take those pages out of
the available pool. If the memory pages are already committed to a virtual machine (VM), the OS
or hypervisor may choose to shut down the associated VM (Guest OS) or cause a Machine Check
Exception to the Guest OS. The associated VM (Guest OS) might in turn close down the affected
thread or process or it might shut down on its own. You can restart the halted VM or process once
you replace the bad memory at the server's next maintenance cycle.

NOTE:

Supported on Intel® E7-4800/8800 and E7 v2 or v3 based systems only
On select Operating Systems, Hypervisors and end-applications that support Machine
Check Architecture Recovery functionality.

Windows Server 2012 R2 and Hyper-V 2012 R2
Windows Server 2012 (requires Microsoft hotfix 2901012) and Hyper-V 2012
Windows Server 2008 R2 and Hyper-V 2008 R2
Windows Server 2008 and Hyper-V 2008
Redhat Enterprise Linux 6.4
SLES 11 SP3

HP Advanced Error Containment

The DL580 Gen9 features advanced PCIe error containment capability that prevents corrupted
data propagation. If data corruption is found on the PCIe bus, the system re-sets the PCIe root
port and re-starts the transaction. This feature helps protect customer data and prevents the error
from creating a chain of failures across the datacenter.

NOTE:

Supported on Intel® E7-4800/8800, E7 v2 or v3  based systems only

HP Advanced Fault Resiliency

Double Data Device Correction (DDDC+1) - Detects and corrects up to two DRAM chip
failures for every x4 DIMM plus single bit error in the server. DDDC requires memory
lockstep mode and may impact performance to varying degrees depending on workload
memory access characteristics.

NOTE:

 Supported on Intel® E7-4800/8800 based systems only.

Single Data Device Correction (SDDC+1) - Detects and corrects one DRAM chip failure for
every x4 DIMM in the server. Further corrects single-bit errors after one DRAM chip failure.
SDDC is supported in performance mode and does not impact system performance.

NOTE:

 Supported on Intel® E7-4800/8800 based systems only.

Online Memory Sparing - Tracks excessive number of correctable errors and copies the
contents of an unhealthy rank to an available spare rank in advance of multi-bit or persistent
single-bit failures that may result in future uncorrectable faults. Since a DIMM or a rank is
needed to perform sparing, this technique reduces the total amount of available memory by
the amount of memory used for sparing.
Memory Mirroring - System maintains two copies of all data. If an uncorrectable memory
error occurs, the system automatically retrieves the good data from the mirrored (redundant)
copy. Memory Mirroring consumes 50% of the system memory capacity.
Redundant Interconnects and Failover - The DL580 Gen9 system and Intel® Xeon® E7-
4800/8800 v3 processors combine to provide a high degree of fault tolerance and failover
capabilities for key processor-processor or memory-processor links or clocks.
Hot-plug redundant power supplies and hot-plug fans - The DL580 Gen9 server features
N+N redundant power supplies and N+1 rotor (each fan module contains two rotors)
redundancy  to keep your system running in the unlikely case a power supply or rotor
malfunctions.

Each fan contains two independent rotors. If one rotor fails, system continues
operation in a degraded condition with remaining rotors. If two rotors fail, system
initiates an OS shutdown. Two rotor failures is a highly unlikely event.
If a fan module (i.e. two rotors) is removed as a hot-plug event, the replacement fan
module must be installed within approximately 60 seconds before the system initiates

QuickSpecs

HP ProLiant DL580 Gen9 Server

Standard Features

DA - 15187   Worldwide QuickSpecs — Version 2 — 8.17.2015

Page  6