I can’t believe what I am seeing. I am backing up around a TB of mixed data (documents, source code, photos, books, and more) and it’s maxing out my Gigabit connection at 90-120 MB/sec. How is it possible for kopia to be so much faster than pretty much any other backup tool I have tried/used (and that’s many!)?
I am really shocked. How has been reliability with restores for others?
Yeah - Kopia is by far the fasted backup tool I’ve ever used. It’s technology is stable, however you’ll have to make sure, that the underlying hardware runs flawlessly - as you’d have to with any other backup solution as well.
I have been using Kopia at work for cross-server deduplication, which requires one big repository. Our biggest one currently sports 98 TB of deduped data from the last 3 or 4 years.
Kopia has its shortcomings when it comes to user management and UI clients, but as a backup/archiving solution I’d totally recommend Kopia.
Wow that’s quite a large backup. Have you ever had any issues with restores?
We’re running quartely automated restore tests and none of those failed for the last 2 1/2 years.
As everyone should. Restore-tests are crucial.
Could you give an overview of restore test procedure, please?
Depends on your need, really. If your repo is not too big you simply could run kopia snapshot verify and have it to verify up to 100% of your snapshots. E.g. for our rather big 90T repo a snapshot verify of 5% was done in 15 hrs, so I’d probably not run a 100% very often.
Then we are doing some other teste, where we generate a 10MB file filled with random noise and have it backed up on our regular daily runs. Next day, we have it restored automatically and check the SHA512 checksums against the original one.
Gotcha. Thanks.
For anyone else following this: snapshot verify | Kopia
I found this reply from Jarek very helpful
It explains the difference between snapshot and content verification, and illustrates how to verify a subset if you have a large repo.
Edit: just realised the docs cover it as well.