search

TheNewNormal / kube-cluster-osx

Local development multi node Kubernetes Cluster for macOS made very simple
Apache License 2.0
298 stars 26 forks source link

Setup -> Enable shared NFS user home folder not available #91

Closed pcm32 closed 8 years ago

pcm32 commented 8 years ago

Hi there,

I recently reinstalled completely to be on the newest release. The NFS mount is not triggered any longer, and I don't get the menu item that allows to trigger the NFS mount of the /Users/myself on the VMs. This was a major feature for my use case of kube-cluster. Would it be possible to please fix/reinstate this?

Thanks! Pablo

rimusz commented 8 years ago

hey @pcm32, we took it of as lot's of guys are having problems of booting VMs with NFS enabled by default. But sure, I can make an option to enable/disable NFS as other apps have. Will see if I can find a spare time over weekend. In the mean time you can use kube-solo app which has the option to trigger NFS on/off

pcm32 commented 8 years ago

Thanks @rimusz, that would be great!

Actually, I just realised that using the previous settings left by the previous version on my /etc/exports, I can restart my nfsd and just call it from inside kubernetes through the NFS PersistentVolume... so no rush for me actually! But might serve others of course.

Just for the reference of other interested users, make sure that your /etc/exports has the following line:

/Users/<your_username> -network 192.168.64.0 -mask 255.255.255.0 -alldirs -maproot=root:wheel

and then restart your nfs daemon:

sudo nfsd restart

and then use a Persistent Volume that connects to this:

kind: PersistentVolume
apiVersion: v1
metadata:
  name: pv-nfs
  labels:
    type: nfs
spec:
  capacity:
    storage: 30Gi
  accessModes:
    - ReadWriteMany
  persistentVolumeReclaimPolicy: Retained
  nfs:
    path: /Users/<your_username>/desired/path/to/mount
    server: 192.168.64.1

Put that on pv_nfs.yml and create the PV with kubectl create -f pv_nfs.yml.

Then create a persistent volume claim that can use this, and then call that persistent volume claim in your pod/job/rc.

rimusz commented 8 years ago

all this /Users/<your_username> -network 192.168.64.0 -mask 255.255.255.0 -alldirs -maproot=root:wheel will be taken care automatically by corectld.

Nice example of NFS use you gave, might bump it to docs at some point. Maybe you wish to help me there?

AntonioMeireles commented 8 years ago

just one note, for the wider audience... safer, more resilient to instead of server: 192.168.64.1 have server: corectld as albeit 192.168.64.1 is generally the default vmnet's gateway it's not always as users are able to change its subnet.

pcm32 commented 8 years ago

Sure, I'm happy to add it as documentation through a PR. Where do you want it? There doesn't seem to be a wiki. Somewhere in the readme.md? as a separate md at a particular directory linking from the readme?

Let me know!

AntonioMeireles commented 8 years ago

fwiw - examples folder perhaps... (with a README mention)

rimusz commented 8 years ago

yep, examples/pv folder would be fine

rimusz commented 8 years ago

@pcm32 here we go https://github.com/TheNewNormal/kube-cluster-osx/releases/tag/v0.4.8