Upgrades¶

Platform¶

Re-run the installer with a newer image, reusing the same state directory and passphrase:

STORAGE_PATH=<same-path-as-the-original-install>
PULUMI_CONFIG_PASSPHRASE=<same-passphrase-as-the-original-install>
HF_TOKEN=<your-huggingface-token>

docker run --rm -it \
  --network host \
  -e PULUMI_CONFIG_PASSPHRASE \
  -e HF_TOKEN \
  -v "$HOME/.kube/config:/.kube/config:ro" \
  --mount "type=bind,src=${STORAGE_PATH},dst=/var/shaide-installer" \
  ghcr.io/axem-solutions/shaide/installer:oss

The installer diffs desired against current state and applies only what changed.

The state directory and passphrase must match the original install. Without them the installer cannot read previous state and will attempt a fresh deployment.

Before upgrading¶

Read the release notes for breaking changes.
Back up installer state and platform data - see Backup and restore.
Upgrade non-production first.

Kubernetes¶

Upgrade the cluster through your provider, or with the RKE2 upgrade playbook on-prem. Keep every node on the same version - shaide does not support mixed-version clusters.

Rollback¶

Re-run the previous installer image. Rollback is not always clean: schema or storage-format changes may not reverse. Restore from backup when in doubt.