00:00:31 <polyex> jmnbtslsQE any idea how kevent would handle tailing 1000 files at once?
00:05:09 <mason> polyex: Better than tailing logs would be a monitoring system.
00:05:35 <polyex> mason how does that work other than tailing files itself?
00:06:10 <mason> polyex: Doesn't matter. If you have a thousand sources of data, then a dashboard where you can whack the red thing to see more is way better than the vain hope you'll see one of the thousand logs showing an error with your eyeballs.
00:06:52 <polyex> mason ya i get you but you're ignoring what i'm asking
00:07:06 <mason> polyex: Absolutely not.
00:07:33 <polyex> "doesn't matter"
00:07:35 <polyex> genius
00:11:13 <jmnbtslsQE> polyex:  what do you mean about how it would handle? performance-wise?
00:11:42 <polyex> jmnbtslsQE is kevent pretty efficient? like could you tail 1000 files without much system overhead or would it fall over?
00:11:59 <polyex> like how heavy is it to 'watch' a file for changes under kevent
00:13:13 <jmnbtslsQE> that number of files in and of itself would not be a problem at all, and you could probably use the simpler "legacy" interface (select) instead if you want
00:13:32 <polyex> what about watching every file on the FS?
00:15:13 <jmnbtslsQE> i don't have the expertise to speculate about that. i would guess that it would eventually become a problem depending on how much activity there is
00:15:27 <mason> https://docs.freebsd.org/en/books/handbook/audit/
00:15:29 <VimDiesel> Title: Chapter 19. Security Event Auditing | FreeBSD Documentation Portal
00:15:35 <jmnbtslsQE> on a typical idle system without a major workload, it would not be a problem
00:17:33 <polyex> damn that's cool
00:18:35 <polyex> ty mason
00:19:10 <mason> polyex: It's notionally equivalent to Linux. I've not driven a custom endpoint myself, so I can't speak to how trivial it'd be to set up.
00:19:39 <mason> polyex: FWIW, I took your initial question as a desire to actually tail a large number of files.
00:19:48 <polyex> ya
00:19:56 <polyex> was just curious how efficient that is
00:21:40 <mason> Tail itself doesn't appear to use kevent: https://cgit.freebsd.org/src/tree/usr.bin/tail/
00:21:42 <VimDiesel> Title: tail « usr.bin - src - FreeBSD source tree
00:23:04 <polyex> why not i wonder
00:24:29 <V_PauAmma_V> Because "tail -f" came earlier?
00:31:51 <dh|2> I see kqueue / kevent being used in the tail source.
00:33:18 <jmnbtslsQE> yeah, the follow function in forward.c
00:41:56 <mason> Oh, blah. I stopped at read.c once I saw it looping over getc.
00:42:28 <mason> I only keep a source tree on my laptop which is off at the moment, so I didn't have grep handy.
00:58:35 <polyex> i have a reproducible fluent bit seg fault bug i can't find a solution to. any help greatly appreciated: https://termbin.com/ntly
01:12:55 <llua> have you tried reporting to the fluent community?
01:13:04 <polyex> i dont got github
01:13:30 <llua> and the issue isn't important enough to make one?
01:13:42 <polyex> EULA
01:14:15 <llua> weird way of spelling yes
01:14:35 <polyex> legal agreements to make accounts
01:44:32 <voy4g3r2> polyex: you could always find a mailing list or a discord.
01:45:22 <voy4g3r2> and the logging questions, and "tailing" multiple files.. there are services like prometheus that acould do a "dashboard" of outputs or you can have an aggregation of files to a cetnral place with tools like splunk.. but if your barrier to entry is EULA.. that may just be moot points
01:46:34 <ketas> i have empty github
02:12:55 <voy4g3r2> ketas: yes, the terms of service tie in the EULA if you have an account
02:15:22 <ketas> well they do have terms, but why?
02:15:29 <ketas> and terms can suck
02:18:08 <voy4g3r2> well i can give a cynical answer
02:18:33 <voy4g3r2> it is like why are there warns on palstic bags, please do not put over head as it may suffocate you
02:19:37 <ketas> does github? :p
02:19:51 <ketas> work like plastic bag
02:20:09 <ketas> they have done weird shit indeed
02:20:31 <voy4g3r2> well you could just install git and have your own repo
02:20:42 <voy4g3r2> and an account is not requried to clone repos
02:21:36 <ketas> unsure why i have account even
02:21:56 <ketas> recently i moaned under some acme.sh bug
02:22:11 <ketas> it's still there btw
02:22:23 <ketas> acme.sh port will not work
02:22:43 <ketas> i should notice maintainer maybe
02:23:25 <ketas> i had to patch the tr-related code in :p
02:24:14 <voy4g3r2> so then this goes to the situation, do you NOT have a github account and deal with the bug.. or you just suck it up and get the account to scratch an itch
02:25:39 <ketas> i know the most nono thing github/ms did was to call everything people pushed there theirs and they feed that into their weird ai
02:25:50 <ketas> but isn't anyone doing that
02:25:58 <ketas> what's "free"
02:26:58 <voy4g3r2> nothing stops any company from doing something similar with different platforms
02:27:15 <voy4g3r2> it is a license that allows people to download, modify and use the code anyway they want..
02:29:46 <ketas> now we have situation where almost all code is in github
02:29:50 <ketas> kind of sad too
02:31:55 <voy4g3r2> there is gitlab
02:32:20 <SponiX> We should all just trust our code to Microsoft - right?
02:32:27 <SponiX> what could possibly go wrong
02:37:05 <ketas> first product that won't suck is vacuum cleaner
02:37:23 <ketas> actually ms is funny company
02:37:31 <ketas> has history and so on
02:38:03 <V_PauAmma_V> "Nothing sucks like an Electrolux!"
02:39:43 <ketas> they actually did that?
02:39:48 <voy4g3r2> oh yeah
02:40:09 <voy4g3r2> i think they were mentioned in an episode of madmen
02:40:22 <voy4g3r2> the ads shown in madmen were real..
02:40:25 <ketas> i hope they didn't use it for other products
02:41:49 <voy4g3r2> a real good podcast about microsoft history: https://www.acquired.fm/episodes/microsoft
02:41:51 <VimDiesel> Title: Microsoft: The Complete History and Strategy
02:42:46 <ketas> that is 4.5h long!
02:43:07 <voy4g3r2> yeah
02:43:37 <voy4g3r2> they make some real good episodes about company history.. they do awesome research on it
02:44:19 <voy4g3r2> and sometimes people.. they did an epsidoe just on taylor swift and how she became the star she is and the business around her. fascinating
02:44:35 <voy4g3r2> but i'll stop, this is not hte focus on this channel and i need to write up a presentation for work
02:45:13 <ketas> indeed good
02:45:56 <ketas> but i think those will work for some trip of other work
02:46:08 <ketas> due extremely long
07:12:16 <polyex> i have a reproducible fluent bit seg fault bug i can't find a solution to. any help greatly appreciated: https://termbin.com/ntly
07:15:37 <mason> polyex: Might be worth grabbing a ktrace to see where in the code it's choking.
12:48:32 <Bebef> Hi everyone
13:02:29 <Bebef> I habe a routing issue. net.inet.ip.forwarding=1 and routes are correct, pf is off. Still, it seems that no packets are forwarded, but silently dropped. Any idea how I could debug this (i.e., get the deceision why packets were dropped)?
13:05:34 <hsnovel> "Ich habe" tell me you are german without telling me you are german
13:06:45 <hsnovel> oh there wasn't ich either. I got too carried away when I read "habe"
13:07:21 <Bebef> b and v ALSO are right next to each other on the keyboard. </hint>
13:07:43 <hsnovel> oh yeah, ur right
13:07:48 <hsnovel> fat fingering huh
13:08:37 <Bebef> Happens, especially when you have to type on a notebook keyboard rather than a reasonable one ;)
13:09:11 <hsnovel> yeah it sucks
13:16:41 <Bebef> Okay, OPNsense seems to be very broken. I just reverted the config to DHCP and everything works (although it doesn't do anything different). Will take this to OPNsense then.
13:20:01 <hsnovel> I am no OPNsense expert or anything but seems like you have to configure static ip address (which I assume was the thing you were using since you said that you switched to DHCP) Can you check this link? docs.opnsense.org/manual/dhcp.html
13:20:32 <hsnovel> they have a section on setting static ip. Seems like you have to change a configuration on somewhere. I am not sure if you have done it or not but if you didn't you might want to give it a try
13:23:37 <Bebef> Thanks for the pointer. However, that's DHCP server. I configured the interface from DHCP to static.
13:24:02 <Bebef> Regardless, it works now and I won't touch it. Won't be my responsibility for long anyway ;)
13:24:27 <hsnovel> haha lol
13:24:36 <hsnovel> I'm glad than ;)
14:11:16 <crest> looks like the FreeBSD 14.0 (amd64) PkgBase mirrors are all broken right now
14:11:23 <crest> the FreeBSD-acct package is missing
14:11:45 <crest> (at least that's the first one that pkg install -r base stumbles over
14:11:47 <crest> )
14:22:56 <concussious> im trying to "/usr/bin/time -h make -j4 buildworld buildkernel kernconf=JIBSD". everything seemingly builds fine, but the output says "GENERIC completed"
14:24:44 <concussious> I have placed this kernconf in sys/amd64/conf/JIBSD, and i used "ident JIBSD" and "makeoptions KERNEL=JIBSD" in there
14:26:19 <concussious> sorry for the stupid question, but how do i figure out what's going on here?
14:38:19 <jmnbtslsQE> concussious: i think the problem is KERNCONF instead of kernconf
14:39:04 <jmnbtslsQE> it should say toward the top of the output which one it's using
14:40:24 <lastshell> Hi, I'
14:41:37 <lastshell> Hi, I'm fairly new to bsd world, I have a freebsd currently at home, trying to setup permanent ip alias I followed this and try with and without inet prefix but not able to make it work, my first ip is using DHCP is that the problem ?
14:41:49 <lastshell> http://www.unixwerk.eu/bsd/ipalias.html
14:41:50 <VimDiesel> Title: unixwerk: Setting IP Aliases under BSD
14:43:24 <lastshell> it works fine the manual way but I wanted to have it in /etc/rc.conf
14:51:11 <concussious> jmnbtslsQE: that's exactly what it was. thank you so much.
14:58:39 <concussious> lastshell: give this doc a try https://docs.freebsd.org/en/books/handbook/network/#config-network-connection
14:58:41 <VimDiesel> Title: Chapter 7. Network | FreeBSD Documentation Portal
15:02:01 <lastshell> thank you concussious
15:29:53 <crest> lastshell: the manual page explaining it is a bit "hidden" under the name rc.conf
15:30:23 <crest> run `man rc.conf` and search for _alias
15:30:46 <crest> or look it up online under https://man.freebsd.org/rc.conf
15:30:47 <VimDiesel> Title: rc.conf
15:46:24 <lastshell> crest I have this in /etc/rc.conf https://bsd.to/5jES but when I restar the netwotk the 2nd virtual ip is not showing
15:46:25 <VimDiesel> Title: dpaste/5jES (Plain Text)
15:46:46 <lastshell> do I need to put both ips static ?
15:47:26 <crest> combining static aliases and dhcp is a really bad idea even if it's technically possible
15:47:48 <crest> i would recommend using either all static addresses or all dynamic addresses per interface
15:48:07 <crest> at least for stateful things like dhcp
15:48:50 <crest> i would recommend setting the mask to all ones for the alias ips
15:49:34 <crest> e.g. ifconfig_re0_alias0="inet 192.168.0.21/32" # you don't have to use the old address + hex mask notation
15:50:22 <crest> the commands to apply fully will remove your default route so be careful applying them to remote systems
15:51:25 <crest> or at least wrap them like this: nohup sh -c 'service netif restart re0; service routing restart'
15:51:28 <lastshell> https://bsd.to/wh4D
15:51:29 <VimDiesel> Title: dpaste/wh4D (Plain Text)
15:51:46 <crest> is the address already assigned?
15:52:05 <crest> try route -n get -inet 192.168.0.21
15:52:37 <crest> or just ifconfig | less and search for the address
15:53:56 <crest> are you on the same link as the system in question or have some form of out of band access (just reaching over to the old laptop used as lab machine, ipmi serial over lan)?
16:01:07 <lastshell> I had to reboo the router I have fix mac address ip for the freebsd ethernet
16:07:31 <CrtxReavr> lastshell, 'arp -d' is a thing too.
16:22:36 <lastshell> so this worked for me now I have two IPs on my lazy freebsd dns (using dnsmasq) https://bsd.to/EEaV
16:22:37 <VimDiesel> Title: dpaste/EEaV (Plain Text)
16:24:25 <lastshell> thank you all
17:14:47 <entikan> can I use the Makefile from a port to build a project on github? if so, how?
17:14:55 <entikan> from github*
17:15:17 <entikan> I want to try the latest state
17:25:24 <jauntyd> https://docs.freebsd.org/en/books/porters-handbook/makefiles/
17:25:25 <VimDiesel> Title: Chapter 5. Configuring the Makefile | FreeBSD Documentation Portal
17:37:26 <entikan> thanks
18:39:30 <debdrup> https://lists.freebsd.org/archives/freebsd-hackers/2024-April/003195.html (and indeed the whole thread) makes for very interesting reading.
18:39:31 <VimDiesel> Title: Re: Stressing malloc(9)
18:54:01 <enyc> < yamada> anyone running freebsd on the raspberry pi 5 ?
18:54:46 <enyc> could you have freenas/truenas on a rpi5 with sata expander ? etc heh
19:18:00 <polyex> what's the best intrusion detection system for freebsd? basically to know the instant a box has been hacked
19:23:36 <yamada> polyex, ossec !?
19:25:11 <polyex> nice ty. how do you like it?
22:06:16 <debdrup> polyex: freebsd has a whole security event auditing framework cf. https://docs.freebsd.org/en/books/handbook/book/#audit - which can be integrated with dtrace using audit_dtrace(4)
22:06:18 <VimDiesel> Title: FreeBSD Handbook | FreeBSD Documentation Portal
22:06:46 <polyex> debdrup does it do everything ossec does or should i run both or?
22:07:42 <debdrup> it does part of what ossec does, with other parts of freebsd (like the mac framework, veriexec, and other things) doing other parts
22:08:36 <debdrup> if you've already got ossec on infrastructure and you're deploying freebsd, stick with that - if you're deploying from first principles, you're free to pick whichever looks mroe promising
22:09:07 <polyex> ya greenfield. i'll start with freebsd native tech first ty!!
22:09:10 <debdrup> freebsds own systems has the advantage of tight integration in freebsd
22:09:20 <polyex> ya that's what i want
23:17:04 <polyex> jonah kowall come in the channel?
23:19:41 <polyex> i wanna build a go app but i don't know how go works. https://github.com/open-telemetry/opentelemetry-collector. anyone know pls?