hostmetrics: Remove slab reclaimable from cache

[rmfitzpatrick]

Is your feature request related to a problem? Please describe.

Currently the library used by the hostmetrics receiver to report cached memory includes slab reclaimable memory in its calculation despite this not being the literal value provided by /proc/meminfo (introduced in shirou/gopsutil#617). This can be unexpected as the receiver's memory scraper also provides usage and utilization values for the slab reclaimable value itself and the fact that this happens isn't advertised in this project.

Describe the solution you'd like

Calculate the linux cached memory by subtracting the advertised slab reclaimable memory so the reported values reflect what's provided by procfs.

Describe alternatives you've considered

Leaving this alone is viable, but I think at minimum the aggregate value should be advertised in the scraper readme.

Additional context

I've not done a comprehensive audit of different offerings but anecdotally sysstat/sar and the legacy sfx smart agent don't conflate the two values.

[github-actions]

Pinging code owners: @dmitryax. See Adding Labels via Comments if you do not have permissions to add labels yourself.

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[fatsheep9146]

@rmfitzpatrick do you have interest in contributing this?

[rmfitzpatrick]

@fatsheep9146 yes I'd be happy to open a PR w/ the proposed change.

[fatsheep9146]

@fatsheep9146 yes I'd be happy to open a PR w/ the proposed change.

Thanks! It's yours! @rmfitzpatrick

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[dmitryax]

There is an ongoing effort to move slab states to another metric:

• Add available memory state for hostmetrics receiver #7417
• [receiver/hostmetrics] Add available memory state (optionally) #19149

@rmfitzpatrick feel free to weigh in

I believe we can keep the cached state as is if we remove the slab states

[rogercoll]

I think this solution will also need to modify the used value provided by Gosputils as it is made by cached state: https://github.com/shirou/gopsutil/blob/0969c9436bc0d9bf190face9c4a3c4a430b98523/mem/mem_linux.go#L313

    // Total memory removing slab from catched
    memStats, _ := mem.VirtualMemory()
    newCached := memStats.Cached - memStats.SReclaimable
    newUsed := memStats.Total - memStats.Free - memStats.Buffers - newCached
    fmt.Println(memStats.Total == newUsed+newCached+memStats.Free+memStats.Buffers)

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• receiver/hostmetrics: @dmitryax @braydonk

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[github-actions]

This issue has been closed as inactive because it has been stale for 120 days with no activity.