You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on May 14, 2020. It is now read-only.
For operational reasons, you might want to move a zfs dataset to another node.
The actual dataset movement is straightforward, thanks to tools like zrepl, sanoid etc. I'm also not worried about having to shut down any pod using it first.
But I'm not sure how best to update k8s with the new info. Is it possible to update the existing pv and pvc to reference the new node? Is it possible to create a new pvc which "adopts" an existing zfs dataset?
I guess at worst, the process might be:
create a new pvc
need to be able to force it onto the right node
this in turn may require a temporary pod with node affinity
delete the temporary pod but keep the pvc
find the zfs dataset it has just created
delete it, and in its place sync across the source zfs dataset
start the "real" application pod with the new pvc
The text was updated successfully, but these errors were encountered:
For operational reasons, you might want to move a zfs dataset to another node.
The actual dataset movement is straightforward, thanks to tools like zrepl, sanoid etc. I'm also not worried about having to shut down any pod using it first.
But I'm not sure how best to update k8s with the new info. Is it possible to update the existing pv and pvc to reference the new node? Is it possible to create a new pvc which "adopts" an existing zfs dataset?
I guess at worst, the process might be:
The text was updated successfully, but these errors were encountered: