ivy: I am now convinced my IPv6 issues are routing related. I am also running radvd (since before yesterday)>

ivy: now reading freebsdfoundation.org/wp-content/uploads/2022/08/sato_IPv6.pdf and discovering sysctl net.inet6.ip6.rfc6204w3

Hello

Can someone indicate a meta AI chatbot which consolidate all other chatbots?

Currently I use chatGPT to help me for my questions but I feel its answers are not helpful enough

I wanted to use Claude but it doesn't support phone numbers in Iran :((

dvl: i don't see how the problem you were describing earlier could be related to rfc6204w3, all that does is let a router autoconfigure addresses - surely you aren't using slaac to assign addresses to your routers internally!

it mostly comes up when you have a crappy internet connection that requires slaac for internet access so you need to enable that on your edge router

ivy: I know of slaac and I know it's associated with IPv6. My hosts are all manually assigned IPv6 addresses.

i find myself feeling like we could do with a C++ "BSD Template Library" in base... i vaguely remember reading something like this on one of the lists recently

I rebooted my router today to discover that, if ng0 (external pppoe interface ) doesn't come up fast, I have no ipv6 default route

how does one handle this?

how are you adding the route? i usually do it from a post-up script in mpd and haven't had any issues like that

ivy: that sounds like the thing that I do not know about yet

set iface up-script /etc/whatever

then in /etc/whatever: iface=$1; route add -inet6 2000::/3 -interface $iface

awesome, let me try that a few times

the static route in rc.conf above seems to only work if the interface is up

which probably makes sense

it seems somewhat redundant to set $ipv6_defaultrouter and also have an $ipv6_static_routes for ::/0

but with the mpd script, you shouldn't need either of those

hi! anybody managed to change the soft limit of the nofile of a podman container using ocijail?

something is wrong with the command

something is wrong with the command

"/usr/sbin/daemon -S -l daemon -s debug -T gitea -u git -p /var/run/gitea.pid -l daemon -s warning /usr/bin/env -i 'GITEA_WORK_DIR=/usr/local/share/gitea' 'GITEA_CUSTOM=/usr/local/etc/gitea' 'HOME=/usr/local/git' 'PATH=/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin:/root/bin' 'USER=git' /usr/local/sbin/gitea web"

How can I debug the daemon?

I also asked chatGPT and copilot

momken: in these cases I do it manually in a terminal:

`su -m $USER`

`env ... /bin/sh`

`export -p`

`/usr/local/sbin/gitea=web`

usually its a permissions thing wrong somewhere

dch Thanks for responding. Should I try to run `/usr/local/sbin/gitea=web` as git user?

momken: you need to do the commands *exactly* as the rc.d script

also, try adding `rc_debug=YES` and `rc_startmsgs=YES` to rc.conf

these tell you what the commands are

a common problem is that you try to set the user in both rc.d and in daemon invocation

then its insufficient permissions

dch After adding `rc_debug=YES` and `rc_startmsgs=YES` to rc.conf I get this:

# service gitea status

gitea is not running.

# service gitea status

gitea is not running.

# service gitea status

`/usr/local/etc/rc.d/gitea: DEBUG: pid file (/var/run/gitea.pid): not readable.`

`gitea is not running.`

dch I also tried these commands and they worked.

root@gitea-manual-2:~ # su -m git

% env 'GITEA_WORK_DIR=/usr/local/share/gitea' 'GITEA_CUSTOM=/usr/local/etc/gitea' 'HOME=/usr/local/git' 'PATH=/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin:/root/bin' 'USER=git'

PATH=/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin:/root/bin

HOME=/usr/local/git

GITEA_CUSTOM=/usr/local/etc/gitea

GITEA_WORK_DIR=/usr/local/share/gitea

LOGNAME=root

SHELL=/bin/csh

USER=git

BLOCKSIZE=K

MAIL=/var/mail/root

MM_CHARSET=UTF-8

PWD=/root

TERM=xterm-256color

LANG=C.UTF-8

LC_ALL=en_US.UTF-8

HOSTTYPE=FreeBSD

Something is wrong with `daemon` and I can't figure it out

momken: please use a paste site for more than one-liners. servers might kick you for flooding

ivy: rebooted the gateway. Seems fine now. I will modify nothing for a few days and see:

does anyone knows why my freebsd system is saying it is upgrading to 14.2-release-p2 when it is already on that version

ibb.co/RGdZzb84 (screenshot for reference)

polyduekes: not me, sorry, no help.

because the corresponding lines in INDEX-PRESENT and INDEX-NEW differ

what options are there for unit testing C++ in base? is it just ATF? for some reason i thought we had googletest...

freebsd-update's fetch_filer_uptodate() determines a need to update if it doesn't match the newest version's file

ATF == Arm Trusted Firmware?

sysutils/atf

TMA (Too Many Acronyms)

(not integration testing, i know about kyua but i think that's overkill for this)

i wish /usr/share/mk had more documentation, or examples, or something

ring0_starr: cat /var/db/freebsd-update/tINDEX.present "INDEX-ALL|812d71cc52bbaae6368c0df1304bcf5410ab5bea2cc73b0f2f79b173cf59bc59 INDEX-NEW|d71a7f70daf30fc465fba7a18547ff216a446d47317d5fcf49fed26461b0ecb1 INDEX-OLD|83987ba6574a2212ca9737df3b37e07c2565c33f86b1674fc0eca62c934bc808" mine index files and besides, were ext2fs or tarfs even touched for the p2 patch?

ah, googletest is hiding in /usr/include/private

depends if you built a custom kernel ..?

i just git pull /usr/src (14.2 branch) and then make kernel because the source update and the freebsd-update binaries were different but i don't see any reason for ext2fs and tarfs to change because of that

what even is tarfs? is that like initramfs?

"The tarfs driver implements a read-only filesystem backed by a tar(5)

file. Currently, only POSIX archives, optionally compressed with

zstd(1), are supported." from the tarfs manpage

ah i forgot about newlines on irc

how's it get used in the base system though...

no idea lol

it would get built if it's specified in MODULES_OVERRIDE

no,i didn't touch that

Hello again

hello

I am sure something is wrong with this command:

root@gitea-manual-2:~ # /usr/sbin/daemon -S -l daemon -s debug -T gitea -u git -p /var/run/gitea.pid -l daemon -s warning /usr/bin/env -i 'GITEA_WORK_DIR=/usr/local/share/gitea' 'GITEA_CUSTOM=/usr/local/etc/gitea' 'HOME=/usr/local/git' 'PATH=/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin:/root/bin' 'USER=git'

/usr/local/sbin/gitea web >> /var/log/gitea/gitea_debug.log

> By default, a buildkernel will rebuild not only the kernel, but also every single kernel mod-

ule that comes with it

Is it available to run gitea without starting it as a service?

which it did in my case, i can send my kernel config if you want

I'm bored. Is there any idea of ??????'self hosting' for me? I have a server with 1 vcpu and 1 gb ram running freebsd

radhitya 1 vcpu and 1 gb ram is very small for sth serious. Run a vpn on it if you need

radhitya If you can host your services at home (i.e. have fiber internet and consistent power supply) I suggest you install TrueNAS on a spare computer, add some HDDs to it and self host many services on it like NextCloud and Gitea (or Gitlab CE if you have enough RAM)

I've done with them. Nextcloud is just not suitable for me. I still prefer scp/rsync + crontab for backup and downloading a file, momken

Currently services are sitting in my server: nsd, unbound, dovecot, opensmtpd, and nginx.

TIL freebsd-update --currently-running

this is the only channel which never survives a host change

I think it's the +r