System information

n
When allocating LUNs or RAID groups for ESX/ESXi systems, multiple operating systems use and share
that resource. As a result, the performance required from each LUN in the storage subsystem can be much
higher if you are working with ESX/ESXi systems than if you are using physical machines. For example,
if you expect to run four I/O intensive applications, allocate four times the performance capacity for the
ESX/ESXi LUNs.
n
When using multiple ESX/ESXi systems in conjunction with vCenter Server, the performance needed from
the storage subsystem increases correspondingly.
n
The number of outstanding I/Os needed by applications running on an ESX/ESXi system should match
the number of I/Os the HBA and storage array can handle.
Resolving Performance Issues
The vSphere Client offers extensive facilities for collecting performance information. The information is
graphically displayed in the vSphere Client. The vSphere Client updates its display periodically.
You can also use the resxtop vSphere CLI command that allows you to examine how ESX/ESXi hosts use
resources. For information about resxtop, see the Resource Management Guide or vSphere Command-Line Interface
Installation and Reference Guide.
Resolving Path Thrashing
If your server is unable to access a LUN, or access is very slow, you might have a problem with path thrashing
(also called LUN thrashing). Path thrashing might occur when two hosts access the LUN through different SPs
and, as a result, the LUN is never actually available.
Only specific SAN configurations in conjunction with the following conditions can cause the path thrashing:
n
You are working with an active-passive array. Path thrashing only occurs on active-passive arrays. For
active-active arrays or arrays that provide transparent failover, path thrashing does not occur.
n
Two hosts access the same LUN using different storage processors (SPs). For example, the LUN is
configured to use the Fixed PSP. On Host A, the preferred path to the LUN is set to use a path through SP
A. On Host B, the preferred path to the LUN is configured to use a path through SP B.
Path thrashing can also occur if the LUN is configured to use either the Fixed PSP or the MRU PSP and Host
A can access the LUN only with paths through SP A, while Host B can access the LUN only with paths through
SP B.
This problem can also occur on a direct connect array (such as AX100) with HBA failover on one or more nodes.
Path thrashing is a problem that you typically do not experience with other operating systems:
n
No other common operating system uses shared LUNs for more than two servers. That setup is typically
reserved for clustering.
n
If only one server is issuing I/Os to the LUN at a time, path thrashing does not become a problem.
In contrast, multiple ESX/ESXi systems might issue I/O to the same LUN concurrently.
Fibre Channel SAN Configuration Guide
68 VMware, Inc.