Skip to content

Maintenance window guidelines

Opsflw Connect coordinates maintenance windows across shared platform services so application teams can plan changes with minimal downtime. Follow these guidelines when requesting or executing maintenance work.

Scheduling windows

  1. Submit a change request with the proposed scope, environments, and fallback plan.
  2. Confirm the window does not overlap with tenant blackouts or critical campaigns.
  3. Reserve a secondary window in case contingency actions require rollback or follow-up tasks.

The change advisory board reviews high-impact requests weekly. Routine patches with pre-approved procedures may use the express path described in the operations runbooks.

Communication checklist

  • Publish the maintenance details in the status page and #opsflw-notices Slack channel 72 hours ahead.
  • Send targeted emails to impacted tenant contacts with timeline, expected impact, and rollback triggers.
  • Host a live bridge during the window for real-time coordination and incident escalation.

Document outcomes and lessons learned in the change record before closing the request.

Execution best practices

  • Freeze configuration drift with opsflowctl lock --scope <environment> before work begins.
  • Monitor platform health via the Elastic Stack dashboards and automated synthetic checks.
  • Follow the rollback plan immediately if SLO breaches persist beyond the agreed error budget.

Update the maintenance calendar after completion so future planning remains accurate.