Quickstart

Launch a real Linux machine, run something risky, and rewind it to a known-good state — the whole loop, in one sitting.

Base URL: examples use $CLUNDRA_API as a placeholder for your Clundra API host. Set it to the host shown in your dashboard.

Create an account, then mint an API key. Everything below works with either a logged-in dashboard session or an API key.

Sign up at Clundra.
Go to Account → API Keys and click Create New Key, choosing the machines:read, machines:write, and machines:delete scopes.
Copy the key immediately — it starts with gopt_live_ and is shown only once.

Pass it in the X-API-Key header on every request:

X-API-Key: gopt_live_your_key_here

2. Create a machine

In the dashboard: click Deploy machine, pick a tier, and it's running in seconds. Prefer the API? Creation is synchronous — the call returns a machine that is already running, with an IP:

curl -X POST $CLUNDRA_API/machines \
  -H "X-API-Key: gopt_live_..." \
  -H "Content-Type: application/json" \
  -d '{ "tier": "small" }'

Grab the returned machine id — the next steps use it as $ID.

3. Run a command

It's a real Linux environment. Run a command from the machine's exec console in the dashboard, or over the API:

curl -X POST $CLUNDRA_API/machines/$ID/exec \
  -H "X-API-Key: gopt_live_..." \
  -H "Content-Type: application/json" \
  -d '{ "command": "uname -a" }'

The response contains the command output and a success flag: { "output": "...", "success": true }.

4. Create your first checkpoint

Before you do anything risky — a migration, an upgrade, an experiment — capture a checkpoint. In the dashboard, click Checkpoint and give it a label. Over the API:

curl -X POST $CLUNDRA_API/machines/$ID/checkpoints \
  -H "X-API-Key: gopt_live_..." \
  -H "Content-Type: application/json" \
  -d '{ "label": "before-migration" }'

The response returns a checkpointId — save it as $CPID. The label is optional.

5. Rewind

Now break something — then put it back. Click Rewind to your checkpoint in the dashboard, or:

curl -X POST $CLUNDRA_API/machines/$ID/rewind \
  -H "X-API-Key: gopt_live_..." \
  -H "Content-Type: application/json" \
  -d '{ "checkpointId": "'$CPID'" }'

Rewind restores the machine's state from that checkpoint and the machine stays live — same machine, same id, its state moved back. Your other checkpoints are untouched and still rewindable.

One limit to know: a machine with a mounted volume cannot be checkpointed, rewound, or forked. Keep reversible machines volumeless. See Checkpoints & Rewind for details.

Next steps

Machines — sizes, lifecycle, persistence, and deletion.
Checkpoints & Rewind — what a checkpoint captures and how restore behaves.
Fork — branch a brand-new machine from any checkpoint.
Networking — expose a service with a public IP and firewall rules.