RKE/Rancher snapshot restore: not a valid zip file

Hi!

I’ve setup a HA Rancher cluster and hooked it up to an external (private) S3 bucket. By configuring rke (configuration details below) I’ve managed to get automatic snapshots. However when I try to restore said snapshots it fails with the following error:

INFO[0006] Container [etcd-download-backup] is still running on host [node]
INFO[0007] Waiting for [etcd-download-backup] container to exit on host [node]
INFO[0007] Removing container [etcd-download-backup] on host [node], try #1
FATA[0007] Failed to download etcd snapshot from s3, exit code [1]: time=“2020-05-04T12:19:11Z” level=fatal msg=“Unable to decompress [/backup/test.zip] to [/backup/test]: zip: not a valid zip file”

The zip was created by RKE/Rancher, so it doesnt seem reasonable that it should fail due to it not being a zip file.

Etcd backup config:

backup_config: 
  enabled: true
  interval_hours: 1
  retention: 90
  s3backupconfig:
    access_key: "<access_key>"
    secret_key: "<secret_key>"
    bucket_name: "<user>"
    endpoint: "<endpoint>"
    custom_ca: |- ...

Commands used:

rke snapshot-save -config cluster.yml -name test

Worked fine!

rke snapshot-restore -config cluster.yml -name test

Failed with the above error

Any ideas or suggestions on how I can fix this problem?

Best regards
Axel

The S3 Bucket is a NetApp Storagegrid installation

I’m working on a fix for this, what version of StorageGRID are you using?

We are currently using version 11.3.0.1.

This was resolved in a later version of RKE.