Runtime management

Daily checks

Check instance access, platform services and application runtime in that order. A reachable homepage does not prove that backend functions are healthy.

Administrators can inspect operations, capacity and logs in system settings. Exact sections depend on the deployed version and account permissions.

Three kinds of records

• Operation logs explain who acted on a resource.
• Task execution records describe a scheduled run.
• Runtime logs provide service and application diagnostics.

Record the time, directory path, execution ID and error summary before correlating events. Keep private business data and credentials out of public reports.

Understand capacity

Resource statistics help identify database, object-storage and workspace usage. Database size is an allocation estimate, not an exact count of bytes written by each operation. Zero growth can mean existing allocated space was reused.

Log archiving is not a full database backup. Use Backup & restore for recovery planning.

Command-line diagnostics

For source deployments, run from the repository root:

go run ./cmd/kagectl doctor
go run ./cmd/kagectl status
go run ./cmd/kagectl verify
go run ./cmd/kagectl logs main

Resolve the earliest error, then verify again. See Deployment & operations for deployment and upgrade guidance.