Does anyone run high performance MariaDB on their FreeBSD servers? My org is looking for a new home for our data.

i got a opentelemetry collector problem. a filelog receiver has logs coming in but service_name says "unknown_service". so how do i set it? looked everywhere and there's NO docs saying

i need my node.js daemon to be started before caddy starts. so should caddy's rc.d REQUIRE: mynodejsapp or should my node.js app BEFORE: caddy?

there a command in base that will output a hash of a file?

like sha1sum -c <file>

the man sha1sum has a bunch of options of different types to try :)

tyvm

how aggressively should i use vars in rc.d scripts? like should i have myapp_pidfile="path/to.pid", or ${name}_pidfile=...?

what works for yoU?

i personaly do the first option until i get to a point where i see a pattern

over optimization and variables / configuration usually causes headaches and if a headache happens.. without understanding how you got to that point.. makes it difficult to debug

like knowing 2 + 2 + 2 = 6 is easier to understand than 2 x 3 = 6

so there's no style guide it's just a free for all?

sorry, arrived late, what?

how aggressively should i use vars in rc.d scripts? like should i have myapp_pidfile="path/to.pid", or ${name}_pidfile=...?

thanks for letting me in on it. sorry to interupt when i haven't needed to touch rc.d scripting yet/

np

In pedestrian scripting I only use vars when I want it to update while it's going

i need my node.js daemon to be started before caddy starts. so should caddy's rc.d REQUIRE: mynodejsapp or should my node.js app BEFORE: caddy?

tykling can you recommend how to use certgrinder on hetzner freebsd host with jails, where other jails use port 22/80 for web, and as i read in the doc you use those exact ports in certgrinder server/ client for getting certs, i only have one ipv4 but one ipv6 per jail with hostnames where the ipv4 shares with vnet/epairs

hi all

adilix, if you have a FreeBSD question, you can ask it directly. Then wait in channel until someone who can answer sees it.

i know it thx for the info

gtg cy

probably a tired question now... is wifi 802.11ac working on fBSD without wifibox or containerisation?

Puzfire: hm either use dns-01 or make whatever is using port 80 forward requests for the ACME url location to the certgrinder server jail

ah yes think creating txt records in dns for the certs will be easyest to implement for me

i need my node.js daemon to be started before caddy starts. so should caddy's rc.d REQUIRE: mynodejsapp or should my node.js app BEFORE: caddy?

tykling for DNS-01 do i need certgrinderd and client on same jail to use this? its unclear when i read documentation. unless i missed it

maybe some usecase docs would be nice, for ppl not using something like this in day 2 day work ;)

i need my node.js daemon to be started before caddy starts. so should caddy's rc.d REQUIRE: mynodejsapp or should my node.js app BEFORE: caddy?

polyex: The two have the same effect

you sure?

are you asking, because neither works?

Does caddy need to start before the node.js daemon or does the node.js daemon require that caddy is running? Answer that question and you will have your answer.

The effect is the same in both cases. The difference is which file gets modified. Do you associate the change with caddy or with node.js daemon?

node.js needs to start first

Hiya! I've got a jail which exhasted somaxconn and now the pids inside the jail don't seem to be killable which has me quite perplexed. I cannot seem to stop the jail no matter how hard I try

from host going sudo service jail stop jailname doesn't work?

nope that's hung. jexecing into the jail and kill -9 ing the process also does nothing x_x

jexec -lU root jailname ...

huh

sudo service jail stop?

to try to stop all jails

and anything in the jail console log in /var/log of the jail host? anything in /var/log/messages of jail host?

gist.github.com/rtyler/b111233dc364011dc7dc7d6858d6d256  is the host messages file, the jail doesn't have anything in it. The host is also barely responsive but things are just fine inside the jail

Title: host_messages.log · GitHub

quite odd behavior

any resource limits set anywhere?

gist.github.com/rtyler/b111233dc364…d6858d6d256#file-jail_processes-txt the jail processes are really trying to shut down :D but those nginx process don't die

Title: host_messages.log · GitHub

nginx.org/en/docs/freebsd_tuning.html the limit of 192 was reached, but I'm surprised this 14.0-RELEASE machine is basically crumbling seemingly because of this

Title: Tuning FreeBSD for the highload

I do also have bhyve VMs on this machine, and some were under high load. It makes me think that the somaxconn exhaustion could be a side effect not a cause.

of course, on the host commands in my shell are not able to return to the shell after executing them, oof

reboot?

hah, I think that's going to end up being my only option

Looks like most of your processes are swapped out

hm that would be consistent with some of the behavior I was seeing, but I already punted it

do you have any nfs or zfs mounts?