There was an error while loading. Please reload this page.
docs: update production PostGIS operations
docs: make failed Job evidence actionable Document the dedicated-alert contract and preserve failed Job evidence until operators have verified recovery and completed manual cleanup. Constraint: Failure logs must remain available for investigation while duplicate emails stay quiet. Rejected: Automatic Job TTL | delayed responders could lose the only container logs. Confidence: high Scope-risk: narrow Directive: Delete only the investigated failed Job, never its CronJob or database resources. Tested: Markdown diff check, live route tests, kubectl Job and container log diagnostics. Not-tested: GitHub Wiki rendered anchor layout.
docs: publish the verified dev PostGIS runbook Document the deployed dev health and migration contract, GitOps email alert ownership, and safe k3s recovery boundaries without implying production has been promoted. Constraint: Production remains plain PostgreSQL pending backup/restore rehearsal. Rejected: One shared dev/prod PostGIS contract | contradicts the live production state. Confidence: high Scope-risk: narrow Directive: Update prod health guidance only after the separate PostGIS promotion. Tested: Live dev health, PreSync history, Gmail firing/resolved mail, and diff checks. Not-tested: Production PostGIS because it is explicitly out of scope.