[Review Docs] Ensure the documentation of control access to APM data

[lucabelluccini]

We have a documentation page here, but the page was created probably before:

• the existence of the traces-apm.sampled
• the existence of the new pre-aggregated data streams

It would be good to revise the page and verify what exactly needs to be done in order to apply the same technique in presence of the new data streams of APM.

FYI @raultorrecilla as discussed.

[endorama]

Upon a first review I think we are covered.

The data stream we use in 9.x:

• traces-apm<namespace>
• traces-apm.rum-<namespace>
• metrics-apm.internal-<namespace>
• metrics-apm.transaction.<metricset.interval>-<namespace>
• metrics-apm.service_destination.<metricset.interval>-<namespace>

• metrics-apm.service_transaction.<metricset.interval>-<namespace>
• metrics-apm.service_summary.<metricset.interval>-<namespace>
• metrics-apm.app.<service.name>-<namespace>
• logs-apm.error-<namespace>
• logs-apm.app.<service.name>-<namespace>

The doc suggests to create aliases on these patterns:

data stream	production alias name	staging alias name
traces-apm*	production-traces-apm	staging-traces-apm
logs-apm*	production-logs-apm	staging-logs-apm
metrics-apm*	production-metrics-apm	staging-metrics-apm

The patterns already cover all data streams, including TBS one (traces-apm.sampled) and metrics aggregations one (metrics-apm.internal, metrics-apm.service*, metrics-apm.transaction).

I will test all instructions to make sure they are correct but from all data streams shoudl already be covered.

[endorama]

Testing was postponed due to other work items.