Best Practices When Deploying Linux on the HP ProLiant DL980 (updated December 2012)

20
Figure 9: Disabling 44-bit System Address Mode using RBSU
Use NOOP in the I/O scheduler
The default I/O scheduler is Completely Fair Queuing (CFQ), which attempts to balance throughput and latency. However,
today’s intelligent storage subsystems usually perform best by using the NOOP I/O scheduler, which disables all I/O
scheduling operations in the operating system. This I/O scheduler may be configured system-wide using the kernel boot
option elevator=noop or on a per-device basis using the following command:
echo noop > /sys/block/<device>/queue/scheduler
Increase crash kernel size
The HP ProLiant DL 980 running Red Hat Enterprise Linux 5.X or SUSE Linux Enterprise Server 10 supports up to 1 TB of
memory, so you should use the kernel boot option crashkernel to adjust the crash kernel size accordingly. Keep in
mind that the Linux kernel has a maximum limit of 896 MB for crash kernel size.
Known issues and workarounds
SUSE Linux Enterprise Server 10 SP3 stops responding under high utilization
To ensure that SUSE Linux Enterprise Server 10 SP3 does not stop responding under high utilization, upgrade to SUSE
Linux Enterprise Server 11 SP1 if possible. SUSE Linux Enterprise Server 11 SP1 has passed all vendor certifications.
As a workaround, reduce the number of logical processors on the HP ProLiant DL980 by disabling Hyper-Threading. This
prevents the server from halting.
Installation hangs with SUSE Linux Enterprise Server 10 SP3 + Netxen driver
Installation of SUSE Linux Enterprise Server 10 SP3 with a Netxen driver sometimes hangs due to a non-functional LAN-
on-Motherboard (LOM). The workaround here is to add brokenmodules=netxen_nic to the kernel boot parameters
and then use the driver DVD to manually load the correct version of the driver.