Hi,

The 0.1.2 version uses "apply" instead of "create" during the restoration. Could you release the last version to avoid problems ?

Thank you

I have a release of prometheus-operator called prom. The release key from prom.v1 configMap has value with character length 467908. We are using kubectl apply … command while creating the release configMaps.

kubectl apply adds annotation kubectl.kubernetes.io/last-applied-configuration in the configMap which has whole object in JSON format.

• prom.v1 configMap

  apiVersion: v1
  data:
    release: H4sIAAAAAAAC...
  kind: ConfigMap
  metadata:
    annotations:
      kubectl.kubernetes.io/last-applied-configuration: |
    {"apiVersion":"v1","data":{"release":"H4sIAAAAAAAC..."},"kind":"ConfigMap","metadata":{"annotations":{},"creationTimestamp":"2019-11-25T10:35:18Z","labels":{"MODIFIED_AT":"1574678151","NAME":"prom","OWNER":"TILLER","STATUS":"DEPLOYED","VERSION":"1"},"name":"prom.v1","namespace":"kube-system","resourceVersion":"3983651","selfLink":"/api/v1/namespaces/kube-system/configmaps/prom.v1","uid":"46925b3d-0f6f-11ea-abfa-02a3452c4edd"}}
    creationTimestamp: "2019-11-25T10:35:18Z"
    labels:
      MODIFIED_AT: "1574678151"
      NAME: prom
      OWNER: TILLER
      STATUS: DEPLOYED
      VERSION: "1"
    name: prom.v1
    namespace: kube-system
    resourceVersion: "3983651"
    selfLink: /api/v1/namespaces/kube-system/configmaps/prom.v1
    uid: 46925b3d-0f6f-11ea-abfa-02a3452c4edd

• Error from plugin:

  $ helm backup --restore utilities
  2019/11/26 15:54:34 extracting backup from file utilities.tgz
  2019/11/26 15:54:34 reading backup data
  2019/11/26 15:54:34 releases found to restore: prom, …
  2019/11/26 15:54:34 applying backup data to tiller (this command will fail if releases exist)
  2019/11/26 15:54:51 Error: command execution failed: [kubectl --namespace kube-system apply -f restore/manifests.yaml]
  2019/11/26 15:54:51 Warning: kubectl apply should be used on resource created by either kubectl create --save-config or kubectl apply
  configmap/cluster-autoscaler.v1 configured
  …
  The ConfigMap "prom.v1" is invalid: metadata.annotations: Too long: must have at most 262144 characters
  Error: plugin "backup" exited with error

• Versions

  $ kubectl version --short
  Client Version: v1.16.2
  Server Version: v1.14.6-eks-5047ed
  
  $ helm version  --short
  Client: v2.15.2+g8dce272
  Server: v2.15.2+g8dce272
  
  # helm-backup version
  # 0.1.2

Should we use kubectl create or kubectl replace instead of kubectl apply while creating the configMaps? I can work on this if we plan to make the change :)

Scenario:

1. helm backup --file dev.tgz dev in old cluster
2. Move secrets from one cluster to another
3. helm backup --restore --file dev.tgz dev in old cluster

Output:

19/08/01 14:56:03 applying backup data to tiller (this command will fail if releases exist)
2019/08/01 15:28:47 Error: command execution failed: [kubectl --namespace kube-system apply -f restore/manifests.yaml]
2019/08/01 15:28:47 configmap/XXX.v1 created
configmap/XXX.v10 created
... many many lines with configmap created
configmap/XXX.v99 created
Error from server (Conflict): Operation cannot be fulfilled on configmaps "XXX.v1": the object has been modified; please apply your changes to the latest version and try again
... many conflict errors.

Although all releases were created and helm ls confirmed it, no pods started.

Cannot install helm-backup plugin.
OS: macOS 10.14.3
Helm: v2.11.0

helm plugin install https://github.com/maorfr/helm-backup      
                                                                                                                                                       
Downloading and installing helm-backup  ...
https://github.com/maorfr/helm-backup/releases/download//helm-backup-macos-.tgz
tar: Unrecognized archive format
tar: Error exit delayed from previous errors.
mv: rename releases//backup to bin/backup: No such file or directory
mv: rename releases//backup.exe to bin/backup: No such file or directory
Installed plugin: backup

When installing on a Linux system, awk throws the following error:

awk: cmd. line:1: warning: regexp escape sequence `\"' is not a known regexp operator

Not able to backup specific release from namespace

helm backup same for restore as well.
Would there be any plan to implement it or anyone already had a alternate solution for it.

I can backup Helm charts when Tiller is installed into kube-system. But if I have Tiller installed into another namespace and I specifically state that in the argument, it is not able to find any releases.

kubectl get pods --namespace appabc
NAME READY STATUS RESTARTS AGE
test1-service-68b4b238fd-7frr4 1/1 Running 0 5d
test1-service-68b4b238fd-fcnv8 1/1 Running 0 5d
demo-service-6d75111d6c-ft8ht 1/1 Running 36 5d
demo-service-6d75111d6c-rjd55 1/1 Running 36 5d
tiller-deploy-8bb9111b5-clfn8 1/1 Running 0 5d

$ helm list --tiller-namespace appabc
NAME REVISION UPDATED STATUS CHART NAMESPACE
test1-service 1 Thu Nov 29 10:53:16 2018 DEPLOYED test-service-0.1.1-snapshot offeropt
demo-service 1 Thu Nov 29 10:53:10 2018 DEPLOYED demo-service-0.2.0 offeropt

$ helm backup --tiller-namespace appabc appabc
2018/12/04 14:10:11 getting tiller storage
2018/12/04 14:10:11 found tiller storage: configmaps
2018/12/04 14:10:11 getting releases in namespace "offeropt"
2018/12/04 14:10:12 found relases:
2018/12/04 14:10:12 getting backup data
2018/12/04 14:10:14 successfully got backup data
2018/12/04 14:10:14 writing backup to file offeropt.tgz
2018/12/04 14:10:14 backup of namespace "offeropt" to file offeropt.tgz complete (found releases: )

The version mentioned here is "0.1.2" which should be "0.1.3" based on the tag update

Please give us a support for Helm 3. It is a great tool. Maybe a doc in order to do a fork to implement support.

Hello I use Multiple path in my KUBECONFIG env var as explaind in the doc

However the plugin try to "stat" the config file and fail as it expect a single file:

2019/01/17 13:00:55 getting tiller storage
2019/01/17 13:00:55 stat /home/ggenot/.kube/config-1:/home/ggenot/.kube/config-2:/home/ggenot/.kube/config-3:: no such file or directory
Error: plugin "backup" exited with error

Note that I have the same behavior with the helm-logs plugin

I am trying to migrate the existing tiller installation from kube-system namespace to tiller namespace.

I created a backup like this
helm backup -l OWNER=TILLER -t kube-system dev
and then trying to restore like this
helm backup --file dev.tgz dev --restore -t tiller

However, I get this error:

the namespace from the provided object "kube-system" does not match the namespace "tiller". You must pass '--namespace=kube-system' to perform this operation.

Does this mean that the source and the destination namespace has to be the same?

Hi, I ran into this on Windows 10 using git-bash for Azure's AKS

$ helm backup svfb --tiller-namespace svfb
2018/11/29 16:00:20 No Auth Provider found for name "azure"
Error: plugin "backup" exited with error

How do I enable debug? This message coming from Kube or Helm is vague.

I use --kubeconfig to manage multiple kubernetes clusters.

but however, looks like the backup plugin not support this flag

$ helm --kubeconfig=~/.kube/config  backup -h
Error: unknown flag: --kubeconfig
Usage:
  backup [flags] NAMESPACE

Flags:
      --file string               file name to use (.tgz file). If not provided - will use <namespace>.tgz
  -h, --help                      help for backup
  -l, --label string              label to select tiller resources by (default "OWNER=TILLER")
  -r, --restore                   restore instead of backup
      --tiller-namespace string   namespace of Tiller (default "kube-system")

Error: plugin "backup" exited with error

Thanks for the plugins

hey there, love your tools for export and import you made.
it would make more sense to have one plugin which does both.
maybe call it backup 🙂

Hello,

On my local, I have microk8s installed via conjure up, and it does not have following file. I think it is not created for minikube as well.

/home/shipahi/.kube/config

I am getting following error.

helm backup default
2018/12/31 01:13:21 getting tiller storage
2018/12/31 01:13:21 stat /home/shipahi/.kube/config: no such file or directory

any idea or improvement suggestions on this?
If you need credentials, maybe we can supply auth token value?