Is there a feature description for “TrueCloud Backups Tasks” that explains how it differs from “Cloud Sync Tasks” and the use cases where it would be more advantageous to use it?
The standard TrueNAS documentation explains how to use but doesn’t address why would want to use. Poking around, it mostly seems to just be a tighter workflow integration of “Cloud Sync Tasks” with IX-Storj for administration, e.g. setup Storj buckets, and a way of creating a direct link between a snapshot on TNS and the files stored in a Storj bucket. That later part seems like it would be most useful in partial recovery of files rather that disaster recovery when you need to recover an entire server.
In my use case, I’m using Cloud Sync as my last line of backup where I’d expect partial recoveries to come from an on-site backup server rather a cloud server as the cloud server is likely to be more “lossy” than the local backup server because of the practical limitations of my upload data pipe. Does TrueCloud provide any substantial advantage over the current Cloud Sync feature for that use case?
TrueCloud backup tasks are intended and designed to offer users a way to backup their data to Storj in an easy and efficient way. It offers extra features to accomplish the goals of this feature, and offers support for snapshot management and restore.
You can use Cloud sync tasks to back up to Storj if you prefer.
I guess I’m trying to understand the practical details behind “support for snapshot management and restore” since a snapshots in Storj and snapshots in TrueNAS are different implemented in different ways.
It sounds like for snapshots, TrueCloud is just presenting Storj snapshots in the GUI as if they were TrueNAS snapshots (which is great) but that the granularity in the actual backup process to Storj remains at the file rather then block level like I see when using Cloud Sync with B2 (my current service) where my daily TrueNAS snapshot takes 50GB but transmits 300GB to B2.
I’ve done a bit of testing with Storj using Dragonfish but the trial account is limited to 25GB so it’s hard to test with large files. I just spun up a test Electric Eel VM to at least explore the GUI changes.
My biggest frustration in learning TrueNAS with over the last couple of years is that while the “how to use features” is well-documented, the “why” is often a mystery that requires years of personal experience to understand (which by its nature misleading and tends to be rooted is early, personal lessons that may no longer be relevant) or figuring out who you can trust (and annoy with questions) when reading forum postings where, much more often than not, their answers might just as flawed as my own personal experience.
I don’t have any great answers given the anathema that most people (including me) with the necessary combination of knowledge and experience have to writing and maintaining guides that are helpful but don’t put food on the table.
From looking at the implementation it is essentially just a wrapper around restic using the various rclone backends, I think. I assume there is no reason that other object storage providers couldn’t be easily supported?
I wonder if PRs would be accepted to add this functionality? It looks like it is intended to be add with Storj just being the first backend supported.
