I set up a Truenas in my mothers home pure for making a backup in the 3-2-1 principle. Few months ago I made this a replication pull-task on the backup-computer for safety reasons. Up till then it was a push task for my main Truenas. The replication task on the backup computer performs well due to the fact that the task shows a ‘success’ sign. But when I look at the size of my userdata from Nextcloud it is 30Gb or so smaller in the backup computer than in my home Truenas. After looking in the Userdata map (backup computer)I see that this map is not updated since months. So no real ‘success’ I think. Are my settings OK? How can I troubleshoot this problem? My target in this all is making a 100% copy of my Truenas data mnt that is day refreshed every day, and only updating the changed files, removing the removed files and coping the new files.
success on the task only means the last snapshot it considered got sent. it will still skip a dataset that isnt in the source/recursive set.
dest userdata not updating for months is the tell. you are probably replicating the parent or a different child than the nextcloud dataset.
on both boxes:
zfs list -o name,used,referenced,written,creation -r tank
zfs list -t snap | grep -i userdata | tail
match newest snap name source vs dest. if dest newest snap is months old, the pull task isnt selecting that dataset (recursive off, naming schema, or source path is the pool not the dataset).
also confirm the task is pull FROM the main nas, source dataset exact path, recursive on, naming schema matches the periodic snapshot task. 30g missing with a green check is almost always a different dataset than you think.