ZFS replication tasks in TrueNAS can fail with stuck resume tokens when network interruptions occur during transfers. The issue requires command-line intervention to resolve: first identify and kill hanging netcat processes using ps/htop, then clear the resume token with 'zfs receive -A'. After clearing the token, replication can resume from where it stopped without resending all data. The problem occurs when the listening process on the receiving system doesn't timeout properly after connection drops, though it's difficult to reproduce consistently.

β€’6m watch time

Sort: