Disk Ops In Progress Metric missing when using 24.04.1-Ubuntu

**Describe the problem**

CRDB Version

Cockroach version is 25.2.8, but it looks like the linux version is more important. Tested multiple CRDB versions and all have the problem when paired with the specific Linux Ubuntu version.

Working:
aws #29-Ubuntu SMP
Not working:
aws #14~24.04.1-Ubuntu SMP

We found the issue on the nodes that aren't showing this metric is that `sys.host.disk.iopsinprogress` is always 0

Found that Linux is properly reporting/logging disk activity, but cockroach is not translating it to/incrementing the `sys.host.disk.iopsinprogress` metric value in `crdb_internal.node_metrics`.

In a nutshell, cockroach seems to be failing to get the metric from Linux.

**To Reproduce**

Setup a cluster on any version of Ubuntu that isn't 24.04.1 - run a workload, view the `sys.host.disk.iopsinprogress` is not always 0.

Setup a cluster on Ubuntu 24.04.1 - run same workload, `sys.host.disk.iopsinprogress` is always 0.

**Expected behavior**
`sys.host.disk.iopsinprogress` reflects correctly on Ubuntu 24.04.1

**Additional context**
Primary impact is ensuring that this version isn't used or that customers will need to upgrade from this version.

Jira issue: CRDB-58073