Creating Table for Disk Usage using Visual Builder

I'm attempting to create a table visualisation to display disk usage for all logical volumes for all monitored servers. This will take the form of a row for each server with a column per logical volume.

5 volumes will be displayed, but not all servers will have all 5 volumes.

I've set up PerfMon counters for each of the 5 volumes in metricbeat.yml and am using 'perfmon.ignore_non_existent_counters' to exclude any volumes that may not be present on a server.

However, I am only seeing percentages being displayed for those servers that have all 5 volumes.

Can anybody confirm if the ignore_non_existent_counters is a working option under Metricbeat 6.2.4? Or do I need to upgrade to 6.3 to use this?

A basic table showing disk usage percentages in this way can't be that difficult surely, so please tell me I've completely missed the point somewhere ....

I suggest switching to the Data Table visualization type and trying something like this:

Thanks for the feedback Spencer.

I've given that a go and have all values returned for C: which is correct as this volume is present on all servers.

However, also including the E: and F: drives that are only present on some servers, results in the data only being shown for those servers. Any server with one or more of these volumes unavailable does not return any data (see image).

This is why I am wondering if the IGNORE_NON_EXISTENT counters option is not being recognised.

Are there any other possible reasons for disappearance of data under these circumstances?

Kind Regards

It sounds like when you add the columns for e and f to your visualization you get different data, which makes me think it's not a beats problem but something based on the way Kibana is querying...

There's definitely something amiss somewhere, as I'm seeing a similar problem when using other perfmon counters.

For example, the following entry in metricbeat.yml is correct as far as I and the config test are concerned, but I fail to see any values returned. I only have this running on one server at present and Kibana identifies with ony this single server.

  • instance_label: Process.PoolPaged.Bytes
    instance_name: Process
    measurement_label: PoolPaged.Bytes
    query: '\\machinename\process(listen)\pool paged bytes'

Are there any known issues in Kibana that would exhibit this sort of behaviour?

Kind Regards

There are a lot of reports and I'm not sure if someone has reported something similar to this, but you can search around the forums and https://github.com/elastic/kibana/issues and see if you find anything that sounds similar.

Thanks Spencer

Whilst I have the opportunity to do so, I'm going to uplift everything to V6.4.2 in any case.

I'll run through this again once I've done so and let you know what happens.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.