beautypg.com

HP Matrix Operating Environment Software User Manual

Page 162

background image

system. To associate the characteristics of the real system from which data is being imported
to the simulation system, use the -o -S options with the

capprofile

command.

When converting standalone servers to VMs, be sure to include an appropriate virtualization
overhead in the utilization calculation for the new VMs. (See

CPU Virtualization Overhead %

for more information on this calculation.)

Socket count is missing or seems incomplete in population report
Socket information is not always available due to limitations in data collection from some platforms
or when leveraging data collection techniques that do not include socket count in their collection.
There is no remedy in these situations.
Workload performance seems inaccurate
Using benchmarks for data normalization affects workloads defined with static profiles in older
Capacity Advisor releases (prior to 6.0) because they use clock speed to determine performance.
Suggested action: Such workloads that are still used for planning should be redefined or edited in
Capacity Advisor 6.x or later to improve the accuracy of the performance measure. See

“Creating

a workload” (page 100)

and

“Editing a workload” (page 101)

.

CPU utilization seems high on managed nodes running Microsoft Windows 2003 Server
Agentless data collection might not be accurate on hyper-threaded systems. Suggested action:
Update Microsoft Windows 2003 Server so that it correctly reports the number of physical
hyperthreading-enabled processors or the number of physical multi-core processors. See

Microsoft

Support

for this issue.

Socket and core count seem off on Microsoft Windows 2003 and HP-UX 11 v2 (IA and PA) systems
Windows 2003 and HP-UX 11 v2 (IA and PA platforms) operating systems may not correctly report
the number of processor sockets in a multicore/hyperthreading environment. Suggested action:
The following patches are required to correct this issue.

For Windows 2003, apply the KB932370 patch. For more information, see

http://

support.microsoft.com/kb/932370/

.

For HP-UX 11 v2, apply the PHKL_37803 patch. You can download this patch from

ftp://

us-ffs.external.hp.com/hp-ux_patches/s700_800/11.X/PHKL_37803

. After downloading,

execute the command sh PHKL_37803, and install the depot file that is generated.

No patch is necessary for Windows 2008 or HP-UX 11 v3. This information does not apply to
HP-UX 11 v1.
Data seems to have disappeared
The profileID that Capacity Advisor uses to store whole-OS performance data (profiles) is based
on a system's network name, as currently discovered in Systems Insight Manager (as seen in the
System Page for the managed node).* If a system is later re-identified, but with a different network
name, any data collected under the old name will no longer be associated with the managed
system under its new network name. The data may still exist in the database, but because it is no
longer identified with the managed system from which it was collected, it cannot be viewed or
used for capacity planning.
The following are some examples of how this can occur:

An VMware ESX VM guest is initially discovered by Systems Insight Manager horizontally
only, with no IP address or host name (only the ESX host is explicitly discovered in Systems
Insight Manager, and VMware Tools are not installed/running on the guest). Later, the guest
is discovered with a valid host name/IP address (that is, VMware Tools are subsequently
installed on the guest). (Horizontal discovery can also find Microsoft Hyper-V hosts with a
similar effect if the guests are later rediscovered with valid host name or IP address.)

If a system is multi-homed and its multiple IP addresses or network names are discovered in
Systems Insight Manager, only one "system node" is recorded in HP SIM to represent that
server (the Primary IP Address). This node will contain the list of IP addresses and network
names for that system. HP SIM randomly selects which IP address and network name pair
becomes the default Primary IP Address used for the node's short name. If you edit the “System
Properties” and change the "Primary IP Address", any data associated with the previous
workload name reference will be “lost”.

162

Troubleshooting in Capacity Advisor