Get woodhouse deploy-ready

• Merges to master trigger auto-deploy with no manual steps
  • to the ops GKE cluster
  • We could use https://gitlab.com/gitlab-com/gl-infra/k8s-workloads/tanka-deployments, by triggering a pipeline via the API from master builds of woodhouse.
• Metrics are scraped by prometheus
• Logs are routed to elasticsearch
  • I'm not sure what index GKE deployments are routed to by default. If they end up somewhere vaguely sensible without breaking everything, that might do for this issue. We could split out another issue to route woodhouse logs to woodhouse's own indexes
  • Spun out into https://gitlab.com/gitlab-com/gl-infra/infrastructure/-/issues/11601
• a staging environment exists
  • we can use the gstg GKE cluster
  • there should be a simple, documented workflow for deploying an unmerged MR to staging
  • it should be configured with dummy integrations: no pagerduty keys, staging slack channels, and a woodhouse-test gitlab project for webhooks, incident issues etc
  • we can use a woodhouse-staging slack channel to announce that we're "borrowing" the staging environment for an unmerged MR
• Incident channel cleanup runs regularly on a CI cronjob
  • moved to cutover plan: https://gitlab.com/gitlab-com/gl-infra/infrastructure/-/issues/11602