Update Alert Routing Server configuration/documentation

[nicholaspcr]

Summary

References https://github.com/TheThingsIndustries/lorawan-stack-support/issues/1476#issuecomment-4858617725
References https://github.com/TheThingsIndustries/lorawan-stack/pull/4898 (doc includes the new behavior defined in this PR)

Documents the Alert Routing Server (ARS) for self-managed Enterprise customers: a configuration reference, the AWS CloudFormation deployment steps, and the gateway outage alerting rules.

Screenshots

screenshot of the new pages

Changes

• Add an Alert Routing Server configuration reference (ars.* options), with a warning that the component is only available on AWS CloudFormation deployments, not Docker or Kubernetes.
• Document ARS deployment in the AWS ECS guide: the templates involved (2-7-queue-sqs, 4-1-secrets, 4-2a-configuration, 5-4-ecs-services) and how Tenant Resource Monitoring's Alertmanager forwards alerts to ARS.
• Mark the Alerting feature as available on Enterprise (AWS CloudFormation only) in addition to Cloud.
• Document gateway outage alert timing: alerts stay open at most 24 hours, after which the customer is notified that the gateway reconnected or that monitoring has stopped.

Notes for Reviewers

...

Checklist

• Scope: The referenced issue is addressed, there are no unrelated changes.
• Run Locally: Verified that the docs build using make server, posted screenshots, verified external links. Test with HUGO_PARAMS_SEARCH_ENABLED=true if style changes will affect the search bar.
• New Features Marked: Documentation for new features is marked using the new-in-version shortcode, according to the guidelines in CONTRIBUTING.
• Style Guidelines: Documentation obeys style guidelines in CONTRIBUTING.
• Commits: Commit messages follow guidelines in CONTRIBUTING, there are no fixup commits left.