HP Matrix Operating Environment 7.3 Update 1 Installation and Configuration Guide for Provisioning with Virtual Resources

Discovery order
For best results, HP recommends creating the following discovery tasks, and running them
sequentially in the suggested order.
1. Onboard Administrator
2. vCenter and SCVMM servers
3. ESX hosts and Hyper-V hosts
4. Management Processors
5. Non-HyperVisor hosts
6. DHCP range
If discovery tasks are run out of order, errors are likely. Typical errors due to running discovery
tasks in an improper order include:
• Association errors
For example, a server not associated with a management processor or virtual machine not
associated with a virtual machine host.
• Credential errors
Because discovery of systems can trigger additional discovery of associated systems or
management processors, if the correct credentials are not supplied for the associated systems,
their discoveries are likely to show errors. If you discover these associated systems with a
subsequent discovery task containing the proper credentials, they should then complete
successfully.
Large system discovery
If a discovery task discovers a very large number of systems, the re-identification done as part of
Configure/Repair can consume resources long enough to cause a subsequent discovery task to
time out which could result in discovery task in “skipped” status. If this occurs, re-run the skipped
discovery task and any subsequent failed discovery tasks in the original order, starting from the
task that timed out. You do not need to rerun tasks which completed prior to a “skipped” task.
If you encounter subsequent failures of the same type:
1. Repeat this process, but introduce delays between the executions of the discovery tasks.
2. Consider breaking up the discovery tasks into multiple smaller ones so that the Configure/Repair
action does not run long enough to cause a timeout of a discovery task.
Troubleshooting discovery issues 63