Platform Operations — Process Map

Companion content. Maps the operational surface area of the platform into process clusters.

Sources of runtime activity (universal census = 453)

Layer	Count	Adapter
DB-managed process defs	373	LIVE_DB_FN (dot/iucmd/job_kind/wf)
host crontab	(clustered)	host_crontab snapshot
systemd timers	(clustered)	systemd_timer snapshot
fs dot/bin	(clustered)	fs_dot_bin snapshot
fs scripts	(clustered)	fs_script snapshot
docker runtime	11	docker_container snapshot
KB SOP docs	partial	kb_sop_doc snapshot

Process clusters (17, from 143 orphan objects)

• Managed-process candidates (6): Backup/DR, Permission-Guard, Reconcile, Health-Monitoring, Content-Publish, Approval-Lifecycle.
• DOT-implementation (3): dot-bin-reconcile, dot-HC, dot-normalize — implementation detail of existing DOT processes.
• Components of existing processes: KB-SOP-link, process-discovery, cron-env-helper, DB schedulers (pollers).
• Owner-gated runtime: docker-runtime (11 services).
• Not-process: OS-infra (29), noise-quarantine (3), templates, infra-config.

Operational reading

Health monitoring, backup/DR, permission-guarding, reconciliation, content publishing, and approval lifecycle are the six processes the platform actually runs but had never named. Naming them (as candidates) is the substance of this phase; making them official is the owner's next move.

Scanner cadence

The universal scanner runs daily (systemd wf-universal-scanner.timer, 04:10) keeping the census fresh; new orphans land in wf_orphan_remediation_queue automatically.