Administration

The Installation section gets Tracedown running. This section is about what you deal with once it is real: once it holds data you cannot regenerate, once other people depend on it being up, and once the placeholder secrets you copied from the shipped .env.example have become a liability rather than a convenience. Nothing here is required to evaluate Tracedown. All of it is required to operate it.

If you are moving a trial installation toward something people rely on, read Secrets & Encryption first and Backup & Restore second. Those two pages cover the failures that are unrecoverable; everything else on this list can be fixed after the fact.

The pages

Page	Read this when…
Secrets & Encryption	Before you expose the install to anyone. Covers PLATFORM_AES_KEY, JWT_SECRET, the shipped development values, and what can and cannot be rotated.
Certificate Authority	You are enrolling agents, an agent certificate is expiring, or you need to understand the CA that underpins scheduler↔agent mTLS.
Backup & Restore	Before you have data worth losing. The encryption key and the database must both survive, and they must be backed up separately.
Retention & Aggregation	Your database is growing, you want longer history, or you are deciding how many aggregate-worker replicas to run.
Body Stores	Agents are somewhere your object storage is not, or response bodies have to stay in a bucket you own.
Scaling	Probes are queuing, one host is no longer enough, or you need to know which services can run more than one replica.
Monitoring Tracedown	You want Tracedown's own health in Prometheus or Grafana — the monitoring system needs monitoring too.
Upgrading	You are moving to a new version and want to know the migration and ordering rules.
Troubleshooting	Something is wrong and you want the symptom-to-cause map rather than a tour of the architecture.