…ncorecache is enabled.

…o take CPUs by uncourecache grouping.

…ncorecache'

adapt to the older codebase

Signed-off-by: Francesco Romani <[email protected]>

In order to improve the observability of the resource management
in kubelet, cpu allocation and NUMA alignment, we add more metrics
to report if resource alignment is in effect.

The more precise reporting would probably be using pod status,
but this would require more invasive and riskier changes,
and possibly extra interactions to the APIServer.

We start adding metrics to report if containers got their
compute resources aligned.
If metrics are growing, the assingment is working as expected;
If metrics stay consistent, perhaps at zero, no resource
alignment is done.

Signed-off-by: Francesco Romani <[email protected]>

wire in the l3 group alignment metrics

Signed-off-by: Francesco Romani <[email protected]>

make sure to log the patched version

Signed-off-by: Francesco Romani <[email protected]>

add more verbose logs by default to make the LLC algorithm
debuggable.

Signed-off-by: Francesco Romani <[email protected]>

fix logging to reflect what takeFullUnCoreCache
actually did; with the buggy code it was still inferrable
from the logs, but it has to be explicit.

Signed-off-by: Francesco Romani <[email protected]>

Signed-off-by: Francesco Romani <[email protected]>

A log in `topology` subpkg was too spammy. Bump its verbosiness.

Signed-off-by: Francesco Romani <[email protected]>

no uncore cache can be treated as all CPUs belonging
to the same group, so alignment is implicit and always guaranteed.

Signed-off-by: Francesco Romani <[email protected]>

further improve throubleshooting to check if the LLC alignment
code is engaging.

Signed-off-by: Francesco Romani <[email protected]>