[velero] Bump velero version to v1.7.0

[jenting]

Special notes for your reviewer:

Since from v1.7.0, the velero container image a distroless, there is no shell command anymore.
We leverage the bitnami/kubectl in the initContainer phase, copy the sh and kubectl command to /tmp folder.
After that, the velero container is able to use the sh (pipe redirection) and kubectl command to upgrade the CRDs by shared mount point /tmp with command

/velero install --crds-only --dry-run -o yaml | /tmp/kubectl apply -f -

Note that: we deprecate the kubectl.initResources and use .resources instead since the Velero container image CPU/Memory request/limit should be the same as Velero Deployment.

Checklist

[Place an '[x]' (no spaces) in all applicable fields. Please remove unrelated fields.]

• DCO signed
• Chart Version bumped
• Variables are documented in the values.yaml or README.md
• Title of the PR starts with chart name (e.g. [velero])

[sarahhodne]

I don't think the sh or kubectl is required anymore, I was able to get this to work in our cluster by just running velero install --crds-only, it uses the same credentials that kubectl apply would be using.

[jenting]

Thanks for the suggestion.

But the velero install --crds-only only applies when the CRD resource does not exist.
Velero won't apply the patch/put change when the CRD resource exists already. Ref to code.

Therefore, if the user upgrades from the old Velero version to the new Velero version, after successfully update, the CRD resource is still the old Velero version.

[sarahhodne]

Ah, the help text of the --crds-only argument ("Useful for updating CRDs for an existing Velero install.") had me confused, and I didn't realise that it failed 🤦🏻‍♀️

Filed a ticket about the help text: vmware-tanzu/velero#4236

[simon-anz]

Bump? We're keen to deploy distroless Velero, and we use these charts.