Correct Ephemeral Permissions on Migration

[benofben]

No description provided.

[benofben]

In the event of a node failure, the system should handle removing the failed node from the cluster and replacing it with a new instance. It should then be added to the cluster.

• If a backup is available, recover from the backup
• If no backup is available, bring up a new node

In the event of unscheduled reboots or scheduled reboots where a snapshot could not be completed before the node went down, anti-entropy repair would be conducted automatically.

[benofben]

If a host is using ephemeral and gets migrated, the ephemeral comes up with bad permissions and DSE fails to start.

[benofben]

Recovery is probably going to be handled as part of DMC. We still need to fix the ephemeral restart issue.

[benofben]

This is still an issue. Should probably add something to the node.sh to resolve this.