Quickstart¶
Block volume — attach one to an instance (during deployment or after), then on the instance:
New, empty volume? Format it first with mkfs.ext4 /dev/<TARGET>. See Attach a block volume for the full steps, including adding it to /etc/fstab so it survives a reboot.
Shared filesystem — create one from the Shared filesystems screen (or while deploying an instance), choose which instances can access it, then mount it on each:
See Create a shared filesystem and Mount a shared filesystem for details — including mounting across every node in a cluster at once. Attaching storage to an Instant Cluster instead of individual instances? See Use SFS with a cluster, since clusters use a separate virtiofs-based filesystem attached only at creation time.
Container Registry: authenticate Docker with your registry credentials, then tag and push your image:
echo '<SECRET>' | docker login vccr.io/<PROJECT_ID> \
-u 'vcr-<PROJECT_ID>+credentials-name' --password-stdin
docker tag myapp:v1.0.0 vccr.io/<PROJECT_ID>/myapp:v1.0.0
docker push vccr.io/<PROJECT_ID>/myapp:v1.0.0
Once pushed, the image appears on the Container Registry page in the console, where you can create a deployment from it directly. See Container Registry for the full steps, including referencing the image in a deployment. Each container deployment also gets a default 500 GiB scratch disk at /data for ephemeral working data.
Object Storage: create access keys under Project management → Credentials → Object Storage Access Keys in the dashboard (these are separate from your Cloud API credentials), then configure the Verda CLI or any S3-compatible tool:
Or, with the AWS CLI, run aws configure with your Verda access key and secret, then point at the Verda endpoint:
Object Storage may need to be enabled for your project by Support; it may not be available by default in a new project.