I’m running TrueNas scale ElectricEel-24.10.2
I was unaware that this was marked as fixed in an upcoming version. I’ll also similarly upgrade a few weeks after GA. Hopefully, this resolves the issue!
Not sure I can see this fix… where did you get this info?
Booo, it looks like them moved it to 25.10 HERE.
At this stage the issue is still not triaged… it might be a bug fix or a new feature that is needed.
I guess I am unclear as to how ix uses Jira, in my world a fix version identified means it is fixed. I though the triage log was block to external users. 
Its a target date for a fix… but its really only confirmed with a github pull request.
I think I’m running into a similar issue here. My first couple TrueCloud backups of a small dataset (125 GiB) took a number of minutes. But now the currently running task says it’s going to take 46 days (!) to finish, and that number goes up every time I check it.
Looks like that pull request referenced in the Jira ticket was merged at least.
Yes, Pull Request is: NAS-134067 / 25.10 / Add restic cache_path by themylogin · Pull Request #16247 · truenas/middleware · GitHub
The issue is that Restic needs a cache enabled to deal with lots of small files.
The first place we put the fix in is the next major release 25.10 (Goldeye).
It can be validated in a Goldeye nightly. Once validated, we can look to see whether it can be backported to the current Fangtooth release.
In the meantime, I’d recommend running the backup tasks less frequently: nightly or weekly . Let us know if it really takes 46 days, or that is just a bad estimate.
I believe the 46 day ETA—the backup had been running for at least 7 hours at that point. In fact the ETA kept climbing every time I checked it, it got up to 52 days before I aborted the job 
The dataset I’m backing up includes an Immich install, and Immich seems to like creating a bunch of directories, and the backup task is dog slow going through each directory one by one. I apparently had a bunch of leftover empty directories hanging around (about 50k of them). I’ve deleted them and now the ETA is stable around 13 hours to go through the remaining 12k non-empty directories, approximately none of which have any changes to be backed up. My CPU, RAM and network are barely being used at all too so I’m not sure what the bottleneck is.