The `logical_disk.volume-exclude` is not working...

[Hummdis]

Current Behavior

I have several servers that have a dedicated PageFile drive, so I exclude them from monitoring. This has worked in the config.yaml file for a while, but since v0.30.5 it's not worked.

collectors:
  enabled: "[defaults],cpu_info,remote_fx,process,pagefile,mssql"
  collector:
    process:
      include: ['wuauserv', 'windows_exporter']
    # For MSSQL server metrics
    mssql:
      classes-enabled: accessmethods,bufman,databases,genstats,locks,memmgr,sqlstats,sqlerrors,transactions,waitstats
    textfile:
      directories: textfile_inputs
    logical_disk:
      volume-exclude: "(^S|Y):$"

As mentioned, this has worked. This is not a new configuration.

The servers in question range from 2012 R2 Datacenter (qty. 1) to 2019 Datacenter. All updated through at least March 2025.

While one server has two drives, most the others just have one and that line is something like:

collectors:
  collector:
    logical_disk:
      volume-exclude: "E:"

Expected Behavior

To ignore the specified volumes.

Steps To Reproduce

Simply use v0.30.6 and specify ignoring a volume and it'll still show.

Environment

• windows_exporter Version: 0.30.6
• Windows Server Version: 2012 R2 Datacenter to 2019 Datacenter

windows_exporter logs

There are no errors or other issues reported.

Anything else?

Here's an example metric from the server with both the S: and Y: drives told to ignore:

# HELP windows_logical_disk_avg_write_requests_queued Average number of write requests that were queued for the selected disk during the sample interval (LogicalDisk.AvgDiskWriteQueueLength)
# TYPE windows_logical_disk_avg_write_requests_queued gauge
windows_logical_disk_avg_write_requests_queued{volume="C:"} 363713.7848944
windows_logical_disk_avg_write_requests_queued{volume="E:"} 71266.8805749
windows_logical_disk_avg_write_requests_queued{volume="F:"} 50.463904299999996
windows_logical_disk_avg_write_requests_queued{volume="HarddiskVolume1"} 169.4592121
windows_logical_disk_avg_write_requests_queued{volume="L:"} 92.08660859999999
windows_logical_disk_avg_write_requests_queued{volume="S:"} 165123.3229188
windows_logical_disk_avg_write_requests_queued{volume="X:"} 2.8009106999999998
windows_logical_disk_avg_write_requests_queued{volume="Y:"} 165094.2029606
windows_logical_disk_avg_write_requests_queued{volume="Z:"} 2.6181711

[jkroepke]

Your config might not correct at this point. this would be the correct one:

collectors:
  enabled: "[defaults],cpu_info,remote_fx,process,pagefile,mssql"
collector:
  process:
    include: ['wuauserv', 'windows_exporter']
  # For MSSQL server metrics
  mssql:
    classes-enabled: accessmethods,bufman,databases,genstats,locks,memmgr,sqlstats,sqlerrors,transactions,waitstats
  textfile:
    directories: textfile_inputs
  logical_disk:
    volume-exclude: "(^S|Y):$"

The key collector where never below collectors

[Hummdis]

Well, that's 100% on me then. I'll get that fixed now.

However, that brings up a new question in that why is this working on the 241 other Windows servers we monitor in the current format?

[Hummdis]

Yup. I just updated the five servers we're seeing this on and it's fixed with the correction to the key's location.

[jkroepke]

Maybe E: is a CD-ROM drive and doesnt not contain a volume? In that case, it get ignored anyways.

[Hummdis]

No, these are all VMs across AWS, Azure, and GCP. The "E" drive is an attached disk that's dedicated for PageFile.

[jkroepke]

Is the E drive avalible, if you clear the config?

[Hummdis]

I'm not sure what you mean. The E drive reports regardless if I have it in the bypass config or not for those 5 systems. However, the indentation issue still allowed for it to be omitted in the other 241 systems. We have monitoring in place that would have triggered alerts because those drives are 100% full and in use by the Windows PageFile.

[jkroepke]

Maybe there is an unknown bug in windows_exporter, but the next version will ensure an config validation which should prevent this.