Operate
Provision a hosted instance, scope who sees which rows, and prove a backup restores.
- Steps
- 4
- Time
- about three hours
Operate is the path for running Antfly for someone else. The work is access, backups, and observability on a hosted instance, plus enough of the Kubernetes operator to serve a client who hosts their own.
Each step is a guide from the docs, in the order the work happens. Read the page, then do the checkpoint before moving on. The checkpoints are the exercises the workshop runs live.
- 01
Getting Started with Antfly Cloud
10 min
Provisioning, connecting, and the first scoped key are the three things every client engagement starts with.
This page answersProvision a managed Antfly instance, connect to it, and create the first scoped API key.
Checkpoint - 02
Cloud Access Control
Organizations, groups, instance grants, and row filters are how one instance serves several teams without any of them seeing another's rows.
This page answersOrganization administration, Cloud API keys, instance grants, groups, and row filters.
Checkpoint - 03
Backups and Observability
Schedules, restores, metrics, events, and audit logs are what a client asks about in the second meeting. Do each one once before then.
This page answersBackup schedules, restores, downloads, metrics, events, and audit logs for Antfly Cloud instances.
Checkpoint - 04
Quickstart
Some clients run their own clusters. The operator gives a self-hosted deployment the same shape as a hosted one, so what you learned above carries over.
This page answersDeploy your first Antfly cluster on Kubernetes
Checkpoint
The Operate capstone
[PLACEHOLDER: the Operate capstone brief. It states the task, the constraints, and the pass criteria, and says that hidden queries are part of grading and that LLM and agent use is allowed. Replace before the PR merges.]