When I set up the Synology server MrsPatmore, I set it up to be a target of the Synology Time Backup service running on the Synology server Elsie. Time Backup is supposed to copy over versions of files every time they change, so you can go back to an older version even if it has been overwritten on the source server.
Time Backup contains a feature called Smart Recycle, which will supposedly delete older files if the space is needed on the target machine for new ones.
Set it and forget it. Sounded good to me.
Not so fast. Today I logged on to MrsPatmore and noticed the disk was 100% full.
That’s not supposed to happen, right? All the machine is doing is being the target for Time Backup. I dug around and found out the Time Backup will not delete older files if they are less than three months old.
I’m going to get rid of Time Backup. Too bad. It’s a bad implementation of a good idea.
I deleted the Time Backup task on the source machine. That didn’t delete the files from the target. I used the GUI to do that. I got a strange progress message:
65 TB! What’s that about? This array only has room for 16 TB.
I am on the same boat here. To make it worse I had to eliminate the entire volume where I used to put all backups in, and it stopped all services and everyone was kicked out of the NAS. please let us know if you have found an alternative backup solution
I’ve not found — or looked for — any alternate solution that runs on the NAS box. I’m using Vice-Versa on a server to achieve the same effect.
Jim
Time backup’s “smart recycle” doesn’t get rid of old versions based on space considerations. Per the documentation, it gets rid of old backups on a time-based algorithm.
Also, it uses hard links, so the size reported is the size of all the backed up files multiplied by the number of times they’ve been backed up (a new hard link is created each time). But each version of each file is really only stored once in the backup.
Hey *,
65TB because time backup uses hardlinks. According to here: http://www.admin-magazine.com/Articles/Using-rsync-for-Backups/%28offset%29/2
The filesystem “thinks” that same versions of files (in different backups) are linked to the same inode (and are consuming only the space of the file once, although they are present in each version). so the inode is only deleted if the last refence (hard link) to that inode is deleted. So you have to multiple delete the same inode, if there are different hard links, that multiplies the amount of space that has to be deleted (it not a bug, its a feature).
HTH, flomow