gchristensen changed the topic of #nixos-chat to: NixOS but much less topical || https://logs.nix.samueldr.com/nixos-chat
<Diogenes-> Eyep. Course I have to run home-assistant in a container
<Diogenes-> Since the python packages for zigbee support aren't there
das_j has quit [Remote host closed the connection]
das_j has joined #nixos-chat
<infinisil> Oh I have a great idea for youtube recommendations
<infinisil> Let me set a maximum view count
<infinisil> Everything over that maximum doesn't get in my recommendations
<infinisil> This enforces discovery of newer and less known things
<colemickens> I have a variety of things - hard drives, ergo keyboards, headsets, nvidia sli bridges, etc. I don't know what to do with this near-junk. Craigslist free doesn't even work.
<joepie91> colemickens: what locale?
<colemickens> Any ideas? I heard somewhere in Japan there's an app where you just take a picture and then everything else is done for you, you just drop the item off somewhere and basically get money deposited. I wish I had that, even if it were just for free stuff. I hate to throw this stuff out, and I feel like if i take it to goodwill... idk, I just don't know if that's a good place for it
<colemickens> Seattle
<joepie91> colemickens: if you don't need money for it, I'd recommend looking for a local Really Really Free Market
<colemickens> For a while I was using "Buy Nothing" groups on FB, but I got off FB and won
<colemickens> dont want to go back just for this.
<colemickens> I'll check that out.
<joepie91> or a giveaway store, if there happens to be one (though RRFMs are more common in the US afaik)
drakonis has quit [Quit: WeeChat 2.7]
endformationage has quit [Quit: WeeChat 2.6]
deadloko has quit [Quit: WeeChat 2.6]
MichaelRaskin has quit [Quit: MichaelRaskin]
avn has quit [Ping timeout: 260 seconds]
<sphalerite> does anyone know of a haskell-esque programming language for embedded devices?
<sphalerite> probably something that compiles to C or C++, for maximum versatility…
<Remosi> not quite what you're looking for, but, rust? :)
<sphalerite> hm, I guess that's a lot nicer than C++.
<tokudan[m]> I'm looking for a tool to monitor multiple hosts both by executing commands local to the monitoring system and by having commands run on those hosts. are there any easy to use tools around or do I have to resort to the allmighty nagios?
drakonis has joined #nixos-chat
<thefloweringash> Maybe check out sensu-go
<tokudan[m]> hmm... sensu-go sounds more complicated than nagios
<tokudan[m]> i guess i'll dive into nagios then
<gchristensen> oh dear
<joepie91> samueldr: you wrote the NixOS wiki page on running it in LXC - is that still working?
<joepie91> because someone here at the assembly is trying to get it to work and it's breaking in.... interesting ways :P
<gchristensen> tokudan[m]: why is that the specific architecture you need?
<tokudan[m]> gchristensen: the "architecture" we have is: monitoring of ~10 linux systems from outside of the network, as we want to catch errors only visible from the outside
<tokudan[m]> installing an agent on each of the systems doesn't work that well
<gchristensen> ahh
<tokudan[m]> so for the local checks on the system i'll just setup nagios to login through ssh and run those checks
<gchristensen> I'm just thinking it might be worthwhile to use something more modern
<tokudan[m]> for example?
<gchristensen> prometheus
<tokudan[m]> doesn't that require a webserver on every monitored host?
<gchristensen> well on every system which is scraped yes, but your HTTP server which is scraped could be on an external system and running those commands you want to run and exporting many host's data from that one point
<tokudan[m]> so you'd have the http server that runs those external commands basically stand-alone while prometheus itself could be run anywhere... but I'd have to run it on the same system anyway, so that additional component only makes adminitration more complex
<gchristensen> that is a totally fine choice
<gchristensen> for me, I would take the simplicity of those 2 systems over the complexity of Nagios' early-2000's software architecture
<gchristensen> s/simplicity/complexity/
<tokudan[m]> i'm just going to make a poc and see if it works and I can easily put the config into git. that's pretty much the most important point for it.
waleee-cl has joined #nixos-chat
<gchristensen> right on
<aanderse> on topic of nagios: https://github.com/NixOS/nixpkgs/pull/76551
<{^_^}> #76551 (by symphorien, 3 days ago, open): nagios: various improvements
<aanderse> yay an abandoned module found a new maintainer!
drakonis has quit [Ping timeout: 240 seconds]
__monty__ has joined #nixos-chat
tilpner_ is now known as tilpner
<tokudan[m]> oh, that's good to know. aaron : how did you see that it was abandoned before that PR?
<aanderse> sugi: last real commit (by someone other than me) was 2016
<aanderse> and every time i tried to get anyone to review PRs for nagios they said they don't use it anymore and had moved on
<gchristensen> that is another thing. with something like nagios, you might end up owning the module
<aanderse> sorry, last real commit was 2014
<aanderse> zabbix seems to be a nice middleground
<aanderse> there are a few people still "stuck" on zabbix
<aanderse> zabbix isn't the new hotness, but a very well established product that actually has architecture to it
<aanderse> thats what i ended up replacing nagios with at work
<aanderse> newer products seemed like a good choice.... but not for where i work
<aanderse> nixos attracts way more people who want cutting edge technology than it attracts people trying to keep the lights on ;-)
<tokudan[m]> yeah, i know that feeling. got a similar issue with our freifunk network which i'm part of :)
<__monty__> The nixpkgs policy on "old software" doesn't help with that, aanderse
<aanderse> __monty__: true... but i was going to remove nagios at some point if someone hadn't stepped up to maintain it. leaving old broken software in nixpkgs is a nightmare for the distros reputation
<aanderse> what a better way to scream to potential users "if you use this distro you will have to do everything yourself"
<gchristensen> ah finally, the rain has transitioned in to freezing rain
<tokudan[m]> aaron: does zabbix need an agent on the hosts or is it possible to run checks over ssh?
<aanderse> i'm pretty sure it can do over ssh, but it is really geared toward the zabbix agent
<tokudan[m]> then it's probably not really suited for me :)
<thefloweringash> is installing an agent completely out of the question?
<aanderse> i get it. i had a huge uphill battle to get zabbix accepted at work in large part because of a client.
<aanderse> 1000+ clients and we had no deployment solution
<aanderse> that isn't an easy sell
<gchristensen> woof
<aanderse> the windows teams were considering a deployment solution, so i used zabbix as a way to push them into a deployment solution faster
<aanderse> people ended up seeing that as a win
<gchristensen> nice politicking :)
<aanderse> and as far as the linux side... i apt-get+yum'd my way through 100+ linux servers -_-
<__monty__> Why the luddite approach to deployment?
<aanderse> because we don't have a deployment tool or strategy at work on the linux side and people had been apprehensive about adopting one...
<aanderse> until i pushed nixos+nixops
<gchristensen> :D :D :D
<aanderse> and i have zero intention of pushing any other deployment tool or strategy for debian and redhat
<aanderse> because every time shit hits the fan and we need to update 100+ boxes
<aanderse> i say "nix boxes are done" in about 10 minutes
<aanderse> and then we break up the rest of the servers and plan 3 days
<aanderse> it pushes my nix agenda nicely :)
<aanderse> i don't think anyone is particularly motivated to come up with a deployment strategy for debian/redhat. not sure why 🤷‍♂️
<aanderse> hmm... either still some funkyness in the acme module, or my configs
rardiol has quit [Ping timeout: 260 seconds]
rardiol has joined #nixos-chat
<tokudan[m]> thefloweringash: I just don't want to bother having to figure out how to install an agent when every single system has a different type of linux on it
<thefloweringash> are the commands that need to be run for the checks already present on the machines?
<tokudan[m]> thefloweringash: yes, simple shell scripts will do everything required. the intended checks are just the type of "any systemd services failed?" and "is process X still running"?
<aanderse> oh that reminds me... i need to continue on my zabbix agent 2 pr which has full systemd support :)
MichaelRaskin has joined #nixos-chat
<samueldr> joepie91: no idea, stopped using it about 2 months after I wrote that; I didn't like relying on a non-nixos system to host nixos stuff
<samueldr> (that was an experiement mainly)
waleee-cl has quit [Quit: Connection closed for inactivity]
avn has joined #nixos-chat
drakonis has joined #nixos-chat
mtjmullen has joined #nixos-chat
drakonis has quit [Quit: WeeChat 2.7]
drakonis has joined #nixos-chat
drakonis has quit [Read error: Connection reset by peer]
drakonis has joined #nixos-chat
__monty__ has quit [Quit: leaving]
drakonis has quit [Ping timeout: 240 seconds]