feat(operator): add Prometheus ServiceMonitor support to valkey-operator chart

[somaz94]

Adds an optional Prometheus ServiceMonitor for the valkey-operator metrics
endpoint, gated on metrics.serviceMonitor.enabled (default false).
Resolves #191 and clears the # TODO: prometheus ServiceMonitor placeholder
in values.yaml. Mirrors the recently merged PodDisruptionBudget addition (#193).

The ServiceMonitor targets the existing operator metrics Service (*-metrics,
port http) and exposes the standard scrape knobs (interval, scrapeTimeout,
scheme, tlsConfig, honorLabels, relabelings, metricRelabelings, extra
labels/annotations, namespace). It renders only when both metrics.enabled and
metrics.serviceMonitor.enabled are true, so the default render is unchanged.

Validation (run locally):

• helm unittest . — 21 passed (10 new ServiceMonitor cases + existing PDB suite)
• helm lint . — 0 failed
• default helm template render byte-identical to main except the chart-version label (0.2.1 -> 0.2.2)
• helm template --set metrics.serviceMonitor.enabled=true renders a valid ServiceMonitor

resolves: #191

Note: Chart.yaml bumps 0.2.1 -> 0.2.2, colliding with my open #195
(topologySpreadConstraints) which also bumps to 0.2.2. They are independent;
whichever merges first, I'll rebase the other to the next patch.