Machine Check Architecture (Mca) Error Recovery; Advanced Settings Utility - IBM x3850 X6 Planning And Implementation Manual

7.5.1 Machine Check Architecture (MCA) error recovery

Hardware errors can now be presented to the operating system (OS) to allow for recovery.
This allows what used to be fatal errors to now be recoverable. The following operating
systems were due to be supported at the time of writing, and are in plan for future support of
this feature:
Windows Server 2012 (full)
Windows Server 2008 (partial)
Red Hat Enterprise Linux (RHEL) 6.4
SUSE Linux Enterprise Server (SLES) 12
VMware vSphere 5.1 (planned)
The following sequence describes the procedure when a hardware error is detected:
1. An uncorrectable hardware error occurs (Figure 7-30).
x3850 X6
Figure 7-30 Uncorrectable hardware error occurs
2. The operating system kills the virtual machine (VM) that was using that memory page and
marks that memory as bad (Figure 7-31).
x3850 X6
Figure 7-31 New VM is created
3. The operating system automatically spawns a new VM.
Sophisticated applications such as SAP and SQL Servers can also take advantage of this
technology, for example:
1. An uncorrectable hardware error occurs.
2. The operating system marks the page as bad and notifies the application.
3. The application rebuilds the records from storage.

7.6 Advanced Settings Utility

The Advanced Settings Utility (ASU) allows you to modify your server firmware settings from
a command line. It supports multiple operating systems, such as Linux, Solaris, and Windows
(including Windows Preinstallation Environment (PE). The firmware settings that can be
modified on the X6 platform include UEFI and IMM2 settings.
VM VM VM
Uncorrectable
error
VM
VM VM
Failed
memory
VM VM
Chapter 7. Management
221