[docs/collector] Disaster recovery recommendations for Daemonset deployments

[mx-psi]

Desired feature or idea: We would like to document recommendations and best practices related to Daemonset deployments with the Collector, since these are prone to have catastrophic failures with traffic increases. We don't necessarily want to add documentation ourselves, but possibly link to 'best practices' documents, specially for things that are not Collector-specific.

Additional context: Part of the discussion in open-telemetry/opentelemetry-collector/issues/11499, see https://docs.google.com/document/d/1BO7hMg0K7zFQ0219z61yRj-2JMMTW_HfoX3Vt6kXW8E/edit?disco=AAABblcXbKI