witten @ witten @lemmy.world

Posts

0

Comments

314

Joined

2 yr. ago

How many examples do you think are needed before people stop doing drugs?

Unlike Twitter, Mastodon seems to rely much more on stable hashtags for discovery. That's probably because there's no "the algorithm" like on Twitter. So with Mastodon you very well can subscribe to hashtags and expect them to get reused over time.

A used Pixel can be had in that range.

Not sure if it's exactly what you're looking for, but here's a "primer" on Podman for Docker users: https://lemmy.world/post/213870

TIL!! I was snookered. Thanks so much for sharing this.

[Comment content deleted by author because apparently I was spouting lies!]

Intolerance of intolerance is essential to a free society. https://en.m.wikipedia.org/wiki/Paradox_of_tolerance

For starters: encryption at rest, block-level deduplication across backups, cloud storage, database support, container support, etc etc.

I will humbly submit that perhaps you are watching the wrong films.

Ah, maybe I need to do the auto reboot too. Thanks.

I have the same, but just as another data point: mine regularly locks up and needs to be reset.. Maybe made worse by big data transfers. Are you running stock firmware or OpenWRT or something?

I'm not sure I understand the question. By "data" do you mean "configuration"? If you've got multiple devs working on a project (or even if you don't), IMO your Docker Compose or Podman configuration should be in source control. That will allow multiple devs to all collaborate on the config, do code reviews, etc. Then, you can use whatever your deployment method is to effect those changes on your server(s)... manually run Ansible, automatically run CI-triggered deployment, whatever.

Yeah, basically. If you want it richer like commercial oat milk and/or you want it not to separate in coffee, you'll also need to blend in some neutral oil plus xanthan gum after straining.

Try making your own oat milk and put in as little sugar as you want! It's pretty easy.

For the cron dumps, you could plug the cron job into a monitoring service (Healthchecks, etc.) so you'd at least know when it fails.

Yeah, the main "innovation" is that it streams directly from MySQL/MariaDB to your encrypted Borg repository without hitting disk.

borgmatic dev here. What I do is run borgmatic locally on each server that needs to get backed up. That's a whole lot easier IMO than setting up network filesystems / rclone or tunnels or screwing around with database dumps yourself, and potentially more reliable. So in your case, I'd run borgmatic on the VPS and then have it connect locally to your MariaDB database using borgmatic's native filesystem support. And then if you also backup the local files with that same VPS instance of borgmatic as well, there's nothing to "merge."

I'd generally recommend one Borg repository per source server / instance of borgmatic.

Lastly, my raspberry uses rclone to push to S3 and I don’t want the keys to be accessible on the VPS’s, that’s why I’m trying to have borgmatic only on my raspberry.

You could always have borgmatic backup to a local Borg repository on the VPS, and then run rclone on your trusted server to copy that repository to S3. Personally I'd probably just put the S3 keys on the VPS and lock it down so that I trust its security, but you do you. 😀

Also.. and here's the fun part.. aluminum cans are lined with plastic.

Why are you not a fan? I'm genuinely curious, as I'm vaguely evaluating such software.

Exactly.