Add support for wildcard, *, for SMB Server Shares in Windows module, Perfmon Metricset

[sachin-frayne]

Describe the enhancement: Add support for wildcard, *, for SMB Server Shares in Windows module, Perfmon Metricset

Describe a specific use case for the enhancement or feature: When you have multiple SMB Server Shares, and you want to collect data from all of them one would assume that you can use wildcard like with other systems but and error is thrown:

ERROR module/wrapper.go:259 Error fetching data for metricset windows.perfmon: failed reading counters: failed collecting counter values: No data to return.

# This config works when providing only one instance "_Total"
- module: windows
  period: 60s
  metricsets:
  - perfmon
  ...
  perfmon.queries:
  - object: "SMB Server Shares"
  instance: "_Total"
  ...

# But when we change it to get all the instances, we get the errors explained above:
- module: windows
  period: 60s
  metricsets:
  - perfmon
  ...
    perfmon.queries:
    - object: "SMB Server Shares"
    instance: "*"
    ...

instance: "*" works for something like "LogicalDisk" but not for "SMB Server Shares" it seems.
I think that the main reason for this is that windows has a strange notation for server shares: \\*\C$ for example.
The instances contain * and \, so I suppose this feature is currently not supported.

[elasticmachine]

Pinging @elastic/elastic-agent-data-plane (Team:Elastic-Agent-Data-Plane)

[willemdh]

Can someone tackle this issue please. Perfmon monitoirng is such a basic functionality and it currrently does NOT work for months already....

[willemdh]

The silence at Elastic is deafening lately. Some feedback after all these months would have been nice.

[willemdh]

15 months later and still no way to index perfmon counters containing wildcards... :(

[MakoWish]

I'll see your 15 months and raise you another three months.

[nimarezainia]

We are currently unable to commit to a time frame for addressing this issue due to other higher priority development activities. However if anyone has a PR for this issue that would allow you to move forward, we would certainly prioritize its review and have it merged.

@jlind23 @pierrehilbert

[willem-dhaese]

2 years later and we still cannot index perfmon metrics in a stable way. I guess the same issue exists in Agent?