Was this page helpful?
Caution
You're viewing documentation for an unstable version of ScyllaDB Open Source. Switch to the latest stable version.
This article will guide you through self-managed ScyllaDB deployment on GCP. For a fully-managed deployment of ScyllaDB as-a-service, see ScyllaDB Cloud documentation.
Active GCP account
Google SDK, which includes the gcloud command-line tool
ScyllaDB Image requires at least 2 vCPU servers.
Choose an instance type. See Cloud Instance Recommendations for GCP for the list of recommended instances.
Other instance types will work, but with lesser performance. If you choose an instance type other than the recommended ones, make sure to run the scylla_setup script.
See the following table to obtain image information for the latest patch release. For earlier releases, see GCP Images
Image Name |
Image ID |
|---|---|
scylladb-6-2-1 |
6004835462308999633 |
Launch a ScyllaDB instance on GCP with gcloud using the information from the previous step. Use the following syntax:
gcloud compute instances create <name of new instance> --image <ScyllaDB image name> --image-project < ScyllaDB project name> --local-ssd interface=nvme --zone <GCP zone - optional> --machine-type=<machine type>
For example:
gcloud compute instances create scylla-node1 --image scylladb-5-2-1 --image-project scylla-images --local-ssd interface=nvme --machine-type=n1-highmem-8
To add more storage to the VM, add multiple --local-ssd interface=nvme options to the command. For example, the following
command will launch a VM with 4 SSD, and 1.5TB of data (4 * 375 GB):
gcloud compute instances create scylla-node1 --image scylladb-5-2-1 --image-project scylla-images --local-ssd interface=nvme --local-ssd interface=nvme --local-ssd interface=nvme --local-ssd interface=nvme --machine-type=n1-highmem-8
For more information about GCP image create see the Google Cloud SDK documentation.
(Optional) Configure firewall rules.
Ensure that all ScyllaDB ports are open.
Connect to the servers:
gcloud compute ssh <name of the created instance>For example:
gcloud compute ssh scylla-node1
To check that the ScyllaDB server is running, run:
nodetool status
Manage your clusters with ScyllaDB Manager
Monitor your cluster and data with ScyllaDB Monitoring
Get familiar with ScyllaDB’s command line reference guide.
Was this page helpful?