The lost posts.

The lost posts.

I broke it again. Unexpected this time to just full on delete the whole thing. It all started with bloat brought on by me failing to manage my docker build cache. That thing was 96% of my storage somehow. I will have to keep an eye on it.

So I lost some posts there, see, it turns out that docker uses lots of storage for building things and if you don't manually clean that up... dumb stuff happens. I managed to restore an old backup from just after I broke the blog the last time. The real worry today is getting Mantilogs back online. Got this back up while waiting for files to copy.

What an odd day. You know this week I wanted to start making a little halloween light. I will probably take a little time to do that. Sketch it out and figure out how I will have to wire it and if I need any wire to hold the clay together in what parts... lots to figure out there.

Anyway I might not have much if any time for Yarrdachi today thanks to this going bang. Better now than later though, I will have to implement a more rigid backup schedule on databases even in development from here on. I think I will make a backup server and have an entry in my control panel that tells me when the last backups from where and when are logged with the longest since last log at the top.

I guess with my experience breaking this blog so many times I know how to recover quickly? Probably should stop breaking it though.

I think my next project after this is to build my own smart watch and make a scheduling API on the dev server where I can keep a easily editable schedule with notifications being pushed to the watch in order to help me take breaks and remember to do crap.

I do seem to need to get more on top of my time management, I have been very focused lately and forgetting other things. Can't get stuck in a channel. Crap to do.

This transfer is taking a little while. I really, really need to get on that backup system. I should make backups with timestamps to an external server and maybe then save that to some cloud file host with redundancies I don't have access to(yet).

Alright, so I lost a few posts about how I built a chunk of the back end and how I worked out some things on the frontend. I will have to start writing documentation as well and including that in the backups. I may release a large chunk of this project as open source software if anyone wanted it.

The issues with the blog are just me messing around in production, I did make backups but not often enough. Hmm...

Welp. I got some fixing to do, transfer will be done eventually and I want to get some more stuff fixed with this time rather than write about it. Just glad this is back up.

Also lost some images for the posts back then, I will fix this later as it's not critical... well mostly. I don't think I have the images anywhere but I will have a look around. Must remember to back up the uploads folder with the DB.

Edit: Bahahaha, Postgres just updated to version 14, gotta redo this with a version 13 container and then make a backup of that and port it to 14. I spent all that time debugging nothing, it was just a version mismatch.

Some days are wasted, some days you learn things. I learned some things about my backup system today and that is that it is weak and I have not put enough though into it because I thought I could do other things first. This is foolish. I should also get on writing documentation sooner rather than later. It's at the point where I will regret it if I don't do it now. Project has become a thing, no longer just something to learn new workflows and technologies on my own. Kind of a way to put the last year's progress into something with at least a little value to people. Hopefully people will play this. If not I did really learn a lot so far.

I also learned that I should probably check my version before doing a pull with a latest tag. Fuckin' idiot.

Oh boy...