This makes sense given the generic nature of block jobs. If mirroring was only for live migration, for example, then we could avoid all this by choosing a single policy. As a generic operation it's nice to have control over error policy.
Stefan
This makes sense given the generic nature of block jobs. If mirroring was only for live migration, for example, then we could avoid all this by choosing a single policy. As a generic operation it's nice to have control over error policy.
Stefan