Cluster Management

TigerGraph Cloud allows you to adapt your resources as your data volume and query complexity grow.

1. Resizing & Scaling

You can modify your instance configuration at any time from the Actions menu.

Instance Size (CPU/RAM)

• Scaling up increases the available RAM for large graph computations.
• Downtime: Resizing instance types involves a short period of downtime while the VM is reprovisioned.

Disk Volume

• You can upgrade your disk size independently of your instance size (Professional tier only).
• Rule of Thumb: Ensure your disk size is at least 3x your RAM to account for OS, logs, and temporary engine files.

2. Solution Lifecycle

Managed solutions have three primary states of control:

Action	Result	Billing
Stop (Pause)	Shuts down the VM. Data is frozen in persistent storage.	Instance billing stops; Storage billing continues.
Start	Restarts the VM. System requires a "warm-up" period to load data to RAM.	Full billing resumes.
Terminate	Deprovisions the VM and all attached disk storage.	All billing stops.

3. Monitoring Activity

The Cloud Portal provides a dedicated monitoring view to track:

• Cloud Spend: Current and projected costs.
• Resource Usage: Basic CPU and Disk metrics (for deeper metrics, use the Informant API).

[!CAUTION] Termination is permanent. Once a cluster is terminated, all graph data and attached storage are deleted and cannot be recovered.