Azure Monitor input plugin does not watch for new resources

[csdaraujo]

Relevant telegraf.conf

[[inputs.azure_monitor.subscription_target]]
  resource_type = "Microsoft.Compute/disks"
  metrics = [
    "Composite Disk Read Bytes/sec",
    "Composite Disk Read Operations/sec",
    "Composite Disk Write Bytes/sec",
    "Composite Disk Write Operations/sec",
  ]
  aggregations = [ "Average" ]

Logs from Telegraf

No valuable logs since the issue relates to new resources that are not identified/included by Telegraf after creation.

System info

Commit d7f3a0e, Ubuntu 22.04, Kubernetes 1.29

Docker

No response

Steps to reproduce

1. Install and run telegraf monitoring Azure resources
2. Create a new resource for a type that is monitored by Telegraf (example used disk, so a PVC in k8s)
3. Check if metrics for the new resources are sent

Expected behavior

New resources should be monitored by Telegraf without the need for a restart (which adds the new resource).

Actual behavior

New resource is not monitored until restart.

Additional info

There should be some kind of watch mechanism to include new resources.

[powersj]

Hi,

It looks like the only time we collect the resources is during Init during start up. The result of that is the resources to montior and a debug message about the total number of targets.

Are you suggesting some sort of timer to refresh that list? Is this something you want to put up a PR for?

I'd suggest making this opt-in with a config option like:

## Resource refresh interval
## Interval to refresh the targets used for montioring. Disabled by default
## and when passed a zero value.
# refresh_interval = "0s"

[csdaraujo]

Yes. That would work! Will work on a PR for us.

Thank you!