What can you do with a $5/mo Linode server?

Everything currently running in screens on my Linode server

  • Notifier (port 8086, reachable through annoying Apache config proxying in the main PN conf file). A Node/Express app that started as a Slack notification thing; other homebrew apps I fool with can POST to it while they do their thing. Notifier takes those POSTed messages and sends them to Slack, from the Slackbot. (Because it’s handy and there, it is now also holding/hosting some other random stuff: a settable/gettable CSS hex color variable for the jokey iPad app/video I instagrammed the other day, and some photos and the .kml file for Every Bench in Southwood. When/if that bench project is done I’ll have to move those files somewhere smarter.)

  • Bckgrnd-Tgthr (port 3009, separate simple (?) Apache conf file routes to it). The backend server for bckgrnd-tgthr.porknachos.com

  • The Icecast service. (I forget how i routed this but I know I wrote it up when I started it.) I could’ve detached this but then I’d have forgotten about it. Leaving it as a separate screen will (maybe) remind me that it’s around.

  • A tiny Node app that merely triggers a rebuild of the various agencynews.org pages I built a few weeks ago as a proof-of-concept. If memory serves it’s literally just a setInterval to trigger a series of GETs to webhooks at Netlify every 12 hours. It probably could’ve just been a cron job.

  • The backend for Howard quotes; the database query-er and an Express route to serve the front-end silliness. On port 3001, with a separate Apache conf file to redirect that route to the Node app.

  • LogAPI (port 3333, again, a separate Apache conf file). A sort-of self-reporter thing that I stopped using, really. Every day it sorts through certain unix logs and produces a couple of numbers: How many IPs blocked per day, for the last month, that sort of thing. It generates a route I can query to get those numbers. For a while, my dashboard project had a little component that queried for that info and built a chart. I scrapped it.

all the active subdomains on porknachos.com, according to my DNS provider

(note-to-self: c.f. /var/www/*)

  • agency.porknachos.com – was a test-bed for the agencynews.org project mentioned above. Now merely says ‘Hello.’ Not sure why I left it there.

  • bckgrnd-tgthr.porknachos.com – a doofus-y music thing I’ve mentioned before.

  • code.porknachos.com – an implementation of Gitea; I did it just to do it; I think I’m using it for one or two projects but I shouldn’t be; the guy who inspired me to try it subsequently wrote up a post about how it was overkill for his needs and he nearly lost a ton of code, messing with the underlying SQL db one afternoon.

  • logapi.porknachos.com – this is the endpoint for the logging thing described above. pretty much nonsense.

  • nextcloud.porknachos.com – a Dropbox-like service, mostly. As with the Gitea thing, it’s running on the SQL db, about which I’m a little shaky, so I probably shouldn’t rely on this as a first-class, leave-Dropbox-behind sort of thing. But I am, mostly. (Dropbox still running on my desktop Mac until I feel more confident about this).

  • radio.porknachos.com – what it sounds like. The front-end that connects back to the Icecast server mentioned above.

…all that and the self-hosted Wordpress on the main domain, where I intend to post all of this info when I’m done typing it up.

(oh and also a dashboard thing that might not be 100% secure so I’m not mentioning it. Ironically, it’s the one thing I do actually use every day.)

To Do

If I get super-motivated, I’ll take all the other little apps and services I’ve got running on Heroku, Glitch, et al, and move them over to Linode, too. No more waiting for the micropub endpoint, indieauth service, or Stripe backend to “warm up” from hibernation.

This entry was posted on May 13, 2019 with tags