Vega
Vega Runtime

Runtime Operations

Day-to-day operation of Vega Runtime: starting and stopping it, using the public CLI, monitoring health and status, reviewing logs, and planning maintenance.

Overview

This guide covers the day-to-day operation of Vega Runtime: starting and stopping it, using the public CLI, monitoring health and status, reviewing logs, handling common issues, and planning maintenance and upgrades. Operate Vega Runtime through approved procedures and do not infer hidden product behavior from local process state alone.

Concepts

TermMeaning
ReadinessThe Runtime reports it can accept work permitted by its current deployment.
HealthA public operational signal about Runtime or Capability condition.
StatusCurrent information about Runtime state, sessions, work, availability, and warnings.
DiagnosticOffline guidance for public configuration, identity, and Runtime data location.
Maintenance windowAn approved period for controlled updates or operational changes.

Runtime CLI

Invoke the public CLI as python -m vega.

TermMeaning
runStart Vega Runtime.
migratePrepare Runtime data for the approved normal or development mode.
statusProvide status guidance for a running Runtime.
diagnosticProvide offline configuration, identity, and storage-layout guidance.
Operator
python -m vega
run
migrate
status
diagnostic
Vega Runtime

Start a reviewed production configuration:

bash
python -m vega run --mode normal \
--config /etc/vega/base.yaml \
--identity /etc/vega/registration.yaml \
--log-level INFO

run supports--mode, repeatable--config,--identity,--log-level,--live-voice, and--audio-device.migratesupports the public mode selection for normal and development operation. Checkpython -m vega --helpfor the options in your installed release.

Warning. Runmigrateonly as part of an approved installation or upgrade procedure, after verifying the required backup and deployment targets.

Runtime status and health monitoring

Check Runtime status before accepting operational work and after startup, restart, or configuration change. Status can include the Runtime lifecycle state, active session, active tasks, Capability availability, version, and visible warnings. The exact fields depend on release and deployment permissions.

Health monitoring uses Runtime readiness and the health of Capabilities required by the work. A healthy Runtime may still decline a request when no eligible Capability is available or a Capability precondition is unmet.

Vega Runtime
Installed Capabilities
Runtime status
Runtime health
Capability health
Operator or monitoring system
Operate, investigate, or stop

Monitor readiness, repeated task failures, capability unavailability, storage capacity, and host-level conditions required by the approved deployment.

Metrics and logs

Metrics are operational measurements used to identify trends and alert on conditions that need attention. Use the publicmetrics_enabledconfiguration setting only according to your approved observability policy. The Runtime does not make an internal monitoring design a public contract.

Logs provide operational evidence for startup, configuration, status, and task issues. UseINFOfor routine production logging and temporarily increase logging detail only for approved diagnosis. Centralize logs with access control, retention, and redaction policies.

bash
python -m vega run --mode normal --log-level INFO
Never put credentials, personal data, or sensitive operational information in support tickets or shared log excerpts.

Troubleshooting

Use the problem → cause → resolution sequence below.

ProblemLikely causeApproved response
Runtime does not become readyConfiguration, identity, storage, or required credential issueCheck visible status, approved paths, logs, and service environment.
Capability is unavailablePackage not installed, not eligible, or unhealthyVerify the approved Capability set and its operational prerequisites.
Recovery needs attentionPrior task outcome is uncertainInspect the real-world condition and follow the recovery procedure before retrying.
Logs are missing or incompleteIncorrect log policy or service environmentVerify deployment logging configuration and access-controlled collection.
Observed problem
Collect status & logs
Identify scope
Apply one approved change
Verify readiness
Escalate if unresolved

Upgrading and maintenance

Plan upgrades as controlled operational changes:

  1. Review release notes and compatibility for Runtime, configuration, and installed Capabilities.
  2. Verify a recoverable backup and tested rollback path.
  3. Stop Runtime cleanly and resolve any active recovery condition.
  4. Install the approved release and run the required public migration step.
  5. Start the Runtime and perform a low-risk readiness and Capability check.
  6. Record the resulting version, time, and verification outcome.
Review release
Verify backup
Orderly stop
Install release
Run migration
Verify readiness
Resume operation
Rollback