CONCEPT Cited by 1 source

Resource lifecycle management

The practice of tracking every provisioned resource through its complete lifecycle — creation, active use, approaching expiration, and cleanup — with full transparency and attribution at each stage.

Definition

In self-service infrastructure platforms, resources left without lifecycle management become orphans: abandoned environments consuming cost, violating platform limits, and obscuring true utilisation. Resource lifecycle management applies TTL defaults, notification chains, and automatic cleanup to prevent this drift.

Key Properties

1. TTL-based expiration — every resource gets a default lifetime with optional extension
2. Notification transparency — provisioning, approaching-expiry, and deletion events are visible to the owner
3. Resource-level independence — different resource types (workspaces, catalogs, databases) carry independent lifecycles; deleting one doesn't cascade to unrelated peers
4. Auditability — every lifecycle event is logged with who/what/when/why
5. Automatic cleanup — expired resources are reclaimed without human intervention

Why It Matters at Scale

At small scale, manual cleanup works. At 2,600+ active deployments across 3 clouds (Databricks FEVM's operating point), platform limits become a ceiling without automated lifecycle management. Unity Catalog and Lakebase both have hard limits on concurrent instances — centralised lifecycle control keeps those limits from binding.

Seen In

• sources/2026-07-23-databricks-self-serve-infrastructure-vending-machine — 90-day default TTL for builder environments, Slack notifications at each lifecycle event, independent catalog lifecycles