Couldn’t reach that bucket — double-check the endpoint, bucket name and key. (Invalid URL) confirmed all information is correct
7 replies
tomspark wrote:
That error's coming from Duplicati talking to Backblaze, not from SparkBox itself. Most common cause: if you picked the "S3 Compatible" destination type, the endpoint field needs to be just the plain address Backblaze gave you (like s3.us-west-004.backblazeb2.com) with nothing before or after it - no https://, no bucket name tacked on. If that's already clean, try switching the destination type to the native "B2 Cloud Storage" option instead - you just paste your Account ID and Application Key there, no URL to get wrong, and it's usually painless. Let me know if it still won't connect.
tomspark wrote:
Thanks for the screenshot, that's exactly what I needed to see - and I owe you a correction, this isn't Duplicati at all, my last reply was aimed at the wrong tool. This is our own new Off-site Backup screen, and I can see the real problem: it's a bug on our end. Try putting https:// right in front of the address, so it reads https://s3.us-east-005.backblazeb2.com - even though our own example text doesn't show it that way. Everything else you typed looks correct. Let me know if it connects once you add that.
m2elite wrote:
still same error Couldn’t reach that bucket — double-check the endpoint, bucket name and key. (Invalid URL)
tomspark wrote:
Sorry for sending you on a wild goose chase there — that didn't help because it couldn't have: the page itself deletes anything you type before "s3." in that box before it's even sent anywhere, so your fix never had a chance to land. This is 100% a bug on our end, not your bucket or your info. There's nothing more to try typing into that box — I've flagged it to get a real fix shipped, and I'll post right here the moment it's sorted. Sorry for the runaround.
tomspark wrote:
Fixed! v1.6.369 is rolling out now. Once your box picks it up (it updates on its own, or grab it from the Updates page), go back to Off-site Backup and hit Test connection again — the address works in any form now, with or without the https:// part. Thanks for your patience on this one, and for the screenshot that nailed it. You were set up exactly right the whole time; the bug was entirely on our side. If anything else snags during your first backup run, post here and I'm on it.