has anyone had good luck with using freedns's free domains for IRC bouncers btw?

nvm

noob here. How different is fbsd from nbsd and obsd? Significantly or minimally?

dukester: very

the nice thing is pretty much all BSD share code with each other

we use nmake from netbsg and a lot of drivers from openbsd

it is bmake, maybe?

oenbsd doesnt use non free licenses and blobs so youll never have zfs in openbsd, you have it in netbsd but its still rather broken

however openbsd may take forever to get things going but ocne it does it's pretty rock solid

Thanks for the brief overview!

angry_vincent, cpet: it is normally called bmake, as in bsdmake, nmake is from microsoft and is a bit diffrent from posix make

right

How do I reinstall ALL packages on the machine?

I thought it's "pkg upgrade -f" but it isn't

okay got it

The pkg db has been corrupted and thinks no package is installed...

another question, does FreeBSD with UFS use RAM as cache for Disk?

Like Linux does

Farooq[lp]: yes

accelerat0r, ty

Farooq[lp]: they are not " technically equivalent" tho

accelerat0r, where can I read about the technical details?

Farooq[lp]: maybe thats barried in some weird docs but I remmebered reading somethin about that in that book from marshall kirk a whileeeeee ago

hmm

okie dokie

Farooq[lp]: da design and implementation of the fbsd OS if Im not mistaken

maybe read 4.4BSD book :P

as stated: "The Design and Implementation of the FreeBSD Operationg System" (second Edition) Section 7.4 Buffer Management (page 347 onward), seem like a good start. (ISBN-13: 978-0-321-96897-5 )

psycorama: that's insanely accurated to say ! I cannot even remember exactly where! but sure there is something related

i just used the occasion to (for once) use my copy.

I don't think there is any modern OS, that doesn't have some disk cache in RAM, as free RAM is just wasted, so you want to use about everything that isn't currently used for something else for caches

heya, anyone know how to get rid of linuxlator?

cant remove /compat/linux/ due to device busy errors.

find what is using it

there is fstat and also procstat

/clear

nvm, that was just the skeletons left by mounted compat devices. Umounting them got rid of em

now which kernel modules that are not conveniently named "lin*" are linux dependencies, too?

i dont want to reboot just now

hey so, i saw in the FreeBSD journal that i should use ZFS even if i'm on a single disk and take advantage of its self-healing capabilities (as seen here freebsdfoundation.org/our-work/jour…ion/laptop-desktop/we-get-letters-7), why? is ZFS way better in its self-healing capabilities than UFS?

dunno much about ZFS. But there are tons more capabilities than just self healing

ZFS doesn't have significantly better self-healing, but it doesn't require fsck'ing at boot to be consistent after a hard shutdown

copies=2 on single disk;)

even without copies=2, BEs are such a solid win

reinhilde: what's the advantage of that?

kevans_: what's a BE?

boot environment

i'd say see bectl(8) but the manpage is probably actually shit at explaining from the perspective of someone unfamiliarr

How can I find out which program is responsible to add an IPv6 address to me lo0?

fe80::1 gets assigned to lo0 which is the IPv6 gateway in this network

and then IPv6 stops working

Also another IPv6 addr I don't want gets assigned to the primary ethernet, which is vtnet0 on this machine

fe80::1 should be localhost, it shouldn't be the network's gateway

each block on zfs is checksummed, so zfs can detect on read if some file got corrupted on disk, without copies it can't heal that, but at least you detect corruption early

Reinhilde, hmmmm?

wow

how can it be?

nope. on this provider, fe80::1 is the IPv6 gateway

What do I tell them?

hm

pretty sure fe80::1 is the link-local gateway, so lo0 having fe80::1 is normal, but you shouldn't have default routed via fe80::1%lo0, but whatever is your main interface

kevans_: can confirm, "Boot envionments allow system changes to be tested safely, as they are selectable directly from the boot loader" says nothing, but the FreeBSD wikii did lead me to vermaden.wordpress.com/wp-content/u…nvironments-reloaded-2018-11-15.pdf which starts with "zfs BE is a bootable clone/snapshot of the working system"

i'm looking into that for you

Reinhilde, tell me how do you look so I will do it meself next time :)

lemme search the RFCs

that is exactly what i'm doing actually

i am trying to find a source for my claim in the RFCs, because I may have lied

:)))

liar

LIAR

:P

does netstat -rWn really show default fe80::1%lo0 […] lo0 for you?

ah, fe80::1 is actually supposed to be an admin assigned token according to the modified EUI64 scheme in question...

only 0::1 itself s actually necessarily loopback.

i can see somewhat why a bunch of nerds i know use ZFS

it sounds interesting

In RFC6874: "such as sending to link-local

addresses under prefix fe80::/10"

geckoware: that's really good feedback, thanks

and it seems that i can manage a bunch of jails easier by using the snapshots features, which is good, i like low maintenance

any pastebin?

I need to send a paragraph from RFC4007

my plans are to use a FreeBSD thing into a router/web server/email/homelab thing, something which i'd feel more at ease if i separated them in jails

if i separated the different services*

as far as I've understood, having two fe80::1 is fine if their scopes are well defined

my usual reco when someone needs a pastebin

but: for RFCs you can cite section and subsections

like i did with that rfc-editor.org link

Reinhilde, right. RFC4007, section 6

and then?

ah

First paragraph

right

so uh

So it should work even though fe80::1 is assigned to lo0

FreeBSD should differ between fe80::1%lo0 and the other one

does your provider run anything listening on fe80::1%yourLinkToThem that you could hit to find out if you are even connected to their thing

yeah, as already said, fe80:: is link local, so recheck netstat if default really goes via lo0 for you

wait... is your ethernet or wifi (or pppoe) to your provider set accept_rtadv ?

assuming they use router advertisements

termbin.com/i6fqa netstat says default goes through fe80::1%vtnet0

Reinhilde, it's a VPS. I've configured it statically

ah so it's a VM

yup

VPS providers are famously poorly managed

just, all of them

there is another thing

so... does pinging fe80::1%vtnet0 return anythng

wait a moment

yeah, that netstat seems fine, assuming vtnet0 is upstream internet for that VM

it is. IPv4 works

kevans_: wiki.freebsd.org/BootEnvironment also tells me how i can set it up, but not necessarily why i'd use it

shit, i wrote that one, to

too

ok, maybe we'll write a better introduction for the wiki page, then we can see about the bectl(8) intro

okay. after a reboot, "ping6 google.com" works as does "ping6 fe80::1%vtnet0"

but it always stops working after ~10 minutes

let's see if this will happen now, too

wait you changed nick from kevans_ lol oops

... it's my understanding that route -6 show works on freebsd

wait hold on, regdomain FCC3 and APAC = DFS for some freqs? WTF

Reinhilde, nope

FreeBSD's route hasn't got show

oh sorry

it has got show

but it doesn't show the routing table

it shows routing for a specific target

13m uptime and still working

i use it when i only care about a specific target

22m uptime and still working

how did the problem get solved? :/

okay now it has stopped working. netstat still looks fine

ping6 to fe80::1%vtnet0 still works

route -6 shows the same as gateway

traceroute?

you mean traceroute6?

doing that now

nothing till hop 11

nothing till hop 41

...

what's next? I guess post it on the mailing list?

okay I have posted me problem on freebsd-net mailing list

Farooq[lp]: ah, might be good to switch to plaintext mail for mailing lists, if you can

that's kinda painful to read via the archives

indeed, unless you have built-in html decoder :D

Could be worse. Maybe there's a pure json emailer out there.

... email client that only handles 6 bit ASCII

My laptop has 2 HDD: (a)internal=>Linux (b)external/USB. I want to install fbsd on (b) from a USB stick. The disk partition section defaults to HDD (a). How do I tell install to use (b).

alternate_nick dukester_

dukester_: so pick to install on (likely) da0

i wonder if installer hides USB drives

rtprio: how?

nxjoseph: that might be so! I guess I need to manually chop up the drive then.

might be, if you have never given a chance to choose a disk and it just defaults to ada0

nxjoseph: that's what it did! and that drive is my internal Libux drive

I see.

nxjoseph: tell the boys to NOT hide any drives and give the choice where to install. :)

do you boot from that usb drive? it shouldn't let you see the drive you're booted off of

just a guess, i don't know how installer behaves

nxjoseph: I get into BIOS and select the drive I want to boot from.

I mean, you flashed FreeBSD iso to the usb drive, right?

and you want to install freebsd onto that drive

that might not be possible for safety

nxjoseph: but right now I'm installing FROM: a USB stick TO: external USB HDD

oh i see

I'll just partition manually and copy the fbsd recommended partitions

i hope that it's in the handbook somewhere though