Adventures in IT

In this post, I'll list the #fediverse accounts that I follow and that could be interesting for #military people who want to start using the #fediverse. I'll group them by topics.

Military

Name & description Fediverse
The War Zone @defensenews@botsin.space
Theatrum Belli @theatrumbelli@botsin.space
Military Strategy Magazine @milstrategy@botsin.space
Janes Intel @janesintel@botsin.space
Defense News @defensenews@botsin.space
Army Recognition @armyrecognition@botsin.space
NATO @NATO@botsin.space
Canadian Armed Forces @canadianforces@botsin.space

Politics & News

Name & description Fediverse
Canadian Global Affairs Institute @caglobalaffairs@bots.franssen.xyz
Foreign Policy @foreignpolicy@bots.franssen.xyz
Courrier International @courrierinternational@mastodon.social
The Atlantic @theatlantic@newsbots.eu

Technology

Name & description Fediverse
Interesting Engineering @intengineering@bots.franssen.xyz
Cults 3D printing @cults3d@bots.franssen.xyz
Tech News Bot @technewsbot@assortedflotsam.com
Matrix.org (messaging) @matrix@mastodon.matrix.org
The Tor Project @torproject@mastodon.social

If you have more interesting accounts; let me know @jeroen@social.franssen.xyz!

Here I list the bots (listed by twitter handles and chronologically) that I'm running from one of my SBC. The script I use is https://gitlab.com/jeancf/twoot Disclaimer: these mirror are unofficial and will be removed once the owners have their own #fediverse capable social media presence.

Self hosted on bots.franssen.xyz

Name & description Twitter Fediverse
Respectfull Memes link link
System76 creators of hardware & #PopOS link link
The War Zone: blog about #military (US oriented) link link
The Register link link
Canadian Global Affairs Institute link link
Vivaldi Browser link link
Instant Karma link link
Cults 3D (3D printing) link link
Problem Solved (funny posts) link link
Foreign Policy link link

Hosted elsewhere

First: thanks to these site administrators to host them

Botsin.space

Name & description Twitter Fediverse
Theatrum Belli: French blog about #military link link
JozefPrusa: CEO of Prusa printers link link
Mil Strategy Magazine link link
Prusa3D: creators of my Prusa Mk3S printer link link
Intengineering link link
ArmyRecognition: blog about #military link link
Textfiles: Jason Scott, Internet Archivist link link
PicardTips: management tips from Captain Picard link link
Darwin awards: funny posting about people acting not really intelligently link link
NATO: The account of NATO. I was really surprised that nobody had already mirrored it. link link
Defense_News link link
janesintel Janes OpenSource Intelligence link link
canadianforces: Canadian Armed Forces link link

Newsbots.eu

Name & description Twitter Fediverse
TheAtlantic link link

I've started using #Fediverse apps almost two years ago. It started with #Matrix for replacing #Whatsapp & #Telegram and then extended to finding a replacement for #Twitter. I have to admit that this has proven to be a bit more difficult to accomplish.

I started with a #Friendica installation, then moved on to #Mastodon, which was really to heavy for self-hosting on a small platform. I then discovered #Pleroma and this has been my go to solution since then.

So the server part is covered, there is a good ecosystem for mobile apps and if you use the excellent #Soapbox (https://soapbox.pub) addon for Pleroma, the interface is excellent for mobile use.

What was lacking however was content. @Fedifollows@mastodon.online is a great way to start but doesn't cover everything. I use Twitter mainly to follow news, blogs etc and getting this content on the #fediverse to be a challenge. There are numerous sites that mirror Twitter accounts but not always to ones personal taste. I started then experimenting with self mirroring twitter accounts that I liked. You need two things to accomplish this: – the mirror software (I use https://gitlab.com/jeancf/twoot) – hosting (of course). There are some good bot hosting #Mastodon instances on the net, such as https://botsin.space, https://newsbots.eu etc but not all my twitter requirements were met.

I gently asked for accounts and actually got some but other request where denied because of the nature of the Twitter account or limited in performance (throttled). I completly understand the reasoning behind this, but it also didn't really help me in my quest. There is also something to be said for not concentrating all the bots on one instance (I feel that there should be a dedicated search engine for this purpose but that's probably for another time).

In a next step, I decided then to create my own bot dedicated Pleroma instance (https://bots.franssen.xyz).

This way, I decide myself whom to add and follow and get direct results. The workload is pretty limited as the site uses #Yunohost that is supporting SSO with Pleroma. I'm however curious about storage requirements but as this instance is only hosting his own images, it should be manageable.

My endgoal is to arrive at a situation where all the accounts I find important, have a place on the Fediverse and my timeline is filled with only known accounts...

I have another post that lists the bots that I currently “power” here

Running a docker #Syncthing relay (which I might describe in another post) gives interesting stats about the usage of your relay... but it's in JSON so no scraping with #Prometheus for visualisation in #Grafana.

The data

You can view your data by browsing to the IP of your docker, http://DOCKERIP:22070/status This gives the following data { "bytesProxied": 395065732238, "goArch": "amd64", "goMaxProcs": 4, "goNumRoutine": 557, "goOS": "linux", "goVersion": "go1.9.3", "kbps10s1m5m15m30m60m": [ 0, 27, 162, 1302, 697, 2428 ], "numActiveSessions": 10, "numConnections": 257, "numPendingSessionKeys": 0, "numProxies": 19, "options": { "global-rate": 0, "message-timeout": 60, "network-timeout": 120, "per-session-rate": 0, "ping-interval": 60, "pools": [ "https://relays.syncthing.net/endpoint" ], "provided-by": "XXXX" }, "startTime": "2020-09-02T18:40:24.868619628+02:00", "uptimeSeconds": 767343 } Especially the sessions and kbps data is interesting to follow.

Telegraf

How to get this data in #Influx? inputs.file input to the rescue. Simply add this to your telegraf.conf file (probably in /etc/telegraf). We'll output the JSON to /tmp as this data is not very relevant and to limit write/read actions on your storage (such as my #DietPi emmc)

[[inputs.file]] files = ["/tmp/syncthing"] name_override = "syncthing_relay" data_format = "json"

Create the JSON

Then, make a simple curl call to get the file on your system so we can read it. This needs to be added to a cron script afterwards. Do crontab -e to edit your cron profile and add:

*/1 * * * * curl http://DOCKERIP:22070/status -o /tmp/syncthing >/dev/null 2>&1

The file gets updated every minute by cron, telegraf reads it and stores in Influx. Next up... Visualisation in Grafana.

Grafana Dashboard for Syncthing Relay

Just make a new dashboard and import this JSON. You'll have to adapt your influx source to make it work.

You can download the JSON [here] (https://franssen.xyz/gitea/jeroen/public_snips/src/branch/master/grafana_syncthing_relay)