Has anyone else been using Wasabi? Are there any gotchas I haven’t run into yet? I realize that they themselves aren’t backed up to other regions, which is presumably why they’re cheaper than S3. I don’t think I’d recommend Wasabi to my users because of the plethora of choices, though when I know more I could probably set up an easy enough ‘do this exactly and don’t touch anything else’ sort of how-to for using it with Arq (for the subset who would obey the ‘don’t touch’ part). It would also be good for file sharing with someone else, since you could restrict them to a particular file heirarchy (‘bucket’). That’s useful even for one person, since you can give each client unique credentials if someone steals your laptop you can clobber that access key without having to change all the others. You can create users and groups, tailor access permissions, and a bunch more I haven’t looked at yet. I still need to learn a lot about using Wasabi/S3. It should be possible for someone to port Cyberduck and Cryptomator since they’re open source, but that’s not going on my list in foreseeable future. If anyone knows of good iOS apps for generically accessing files on S3, especially with open source or audited encryption, I’m all eyes. The web interface also might work better with smaller files, but it wouldn’t be encrypted. iOS options for connecting to S3 seem to be nil just now, but I luckily bought iOS Transmit when it was still available, so in principle that ought to all work until the app breaks. Something like Panic’s Transmit, or CyberDuck + Cryptomator. Having a nice generic cloud available also opens up having my own fully encrypted cloud service, at least for the desktops. I could probably do the entire backup that way, then have Arq adopt it, but I’d have to test that a lot first the separation/archive seems less fraught. If I segregate older unchanging files, I can back those up with Arq to a hard drive, which gives me the encryption, compression and deduplication, then take the drive to a fast connection and upload the backup to the cloud with any third party client.
#CYBERDUCK WASABI HOW TO#
One of the things I’ve discovered I like a lot about the Arq / other cloud client model is that I’ve figured out how to quick start my backup so it doesn’t take a year to get it all up there. The linux folder got there in about 8 minutes over a GB connection, and I threw a bunch of other things at it with no trouble. I downloaded a demo of ExpanDrive (no fuse!), used a Wasabi help file to get it connected, and that worked great. I’ve never had Fuse not cause weird problems on a working system. Did Show Package Contents, peeked at the Frameworks, discovered that it uses osxfuse, and tossed it. The transfer hung after 5 minutes, and I couldn’t find a way to cleanly cancel the operation. As my first test, I told it to to copy a folder with eight linux installer isos, 17GB total. Upload though the web interface was unreliable at best. I’ve been playing with the 30 day demo, and on the whole I think I’ll stick with it though there have been a couple of minor issues. They clone the Amazon S3 model, so many clients work easily with it, and the wasabi help is fairly good so far. Wasabi has a choice of server locations - west coast, east coast, europe. B2 simply isn’t far enough away from my primary disaster scenario - major PNW earthquake or volcano. Now I’m back to the ‘which cloud?’ choice. I wish it could select files based on modification dates and other metadata, but I can live with file paths as long as I can see the hierarchy as I select things. Multiple backup sets and other useful features. Sane file selection so it’s easy to prioritize what gets backed up on a slow connection and gradually add to it.
#CYBERDUCK WASABI FULL#
I played more with a full Arq demo at work, and it’s so much nicer. It’s certainly worth trying out, but keep a close eye on it throughout the demo period (and after) to make sure it’s doing what it should.
I expect that it will eventually be a good choice for many people, but for now I’d call it a beta release. The developer never replied to the report I sent about the problems I was having (I didn’t really expect a reply, it’s a small shop). At some point it just stalled, and was only sending out a few bites every couple of minutes. In about 9 days, it managed to transfer 2.5 GB out of 100 GB of it. And it did, until I added my Aperture library which is the primary thing that needs to be backed up. I hoped that if I limited it to my main system boot drive, it could at least manage that cleanly.
I gave up on it on my fileserver pretty quickly because ACB gets a D- at selecting what you want to backup or not backup. It’s not ready for primetime, at least for me. Unfortunately I had to give up completely on Arq Cloud Backup.
0 kommentar(er)
