I'm interested in collecting various CPU metrics from UIM (v8.51) for all of the "computer systems" that are registered within our UIM instance. I'd like to use the REST API approach to do this, and I've deployed/tested/validated the "webservice_rest" package. The goal is to leverage the REST approach in Splunk as a modular input in order to index and visualize the metric data on that platform.
I have used the "/ci/metricdefinitions/" call to identify the metrics I'm interested in (for example, "% CPU Time")...but I'm not sure what call I would use to generate the metric data for all of the computer systems. Any assistance would be greatly appreciated.
can you have a look here:
QoS Calls - CA Unified Infrastructure Management Probes - CA Technologies Documentation
Looking at the calls, don't see what you want, and instead the call is based on a specific system name, source, target.
Target for example could be disk C:, or D:, or CPU 1, CPU2, those are all different targets.
probably need to start withGet Targets for QoS-Name and Source
guessing you want current and not historical so it would beGet Raw QoS Data
Thank you both for the guidance. I'm also curious to know a couple of things relevant to the REST call:
a) If I want to collect metrics for all "sources" (i.e., computers) at once as opposed to specifying a single source, how would I accomplish that? Can I use a wildcard approach?
b) When I was looking through disk metrics, I found some targets defined as "C:/". How do I specify that target in the REST call? Do I need to escape the backslash somehow?
A) based on the documentation it does not seem possible to get the data for all sources in one call.
B) use 'Get QoS Data (Using the Table ID)' get table ID from s_qos_data
Thank you, David...this is a great starting point for me to work with.