got myself another page fault, this time I have a core file

there's no possible way dhclient has a page fault bug right?

rwp I use HashiCorp vault for secrets management

there’s an open source fork called openbao that isn’t yet in ports tree

AMA it’s great even at small scale

nmz: it's diverged from upstream, so who knows

I wanted to put FreeBSD-14.1-RELEASE-arm64-aarch64-RPI.img onto a 32GB microSD card, but it seems the image is too big, how do I manually downsize it?

the image is not too big, wut

zilti: the image should be about 1gb no?

are you sure you don't have a faulty sd card? sdcards are usually shit

r0ni: it is 5GB as a .img file. The sd card is definitely fine. dd just keeps writing, it did indeed fill the complete microSD, then complain that there is no space left

nmz:

Sorry for the mis-ping. Not sure what happened there

zilti: can't you use the rpi-imager?

nmz: I guess I could. I didn't know of its existence. Thanks!

if that doesn't work then ¯\_(ツ)_/¯

rpi's boot in a strange way, you have to create a special partition then set it as bootable then write the os on another partition

I have audio out I want on pcm0, but the microphone is usb, so it's pcm5. But since if you set hw.snd.default_unit to 0 results in audio coming out hdmi as wanted, but the mic default still sits at pcm4. I've been all over mixer(8) with no ability to set the mic independant of the default_unit for audio out.

zilti: scammers can modify a microSD card controller to falsify the card volume. When you plug the microSD card into your system, it appears as one size, but once you start writing data to the drive, you'll find that you run out of space early

freshports.org/sysutils/f3 could be used to test the card

not to mention some SDCards are write once and then they become read only

I installed a FreeBSD desktop, and as usual, configuring Xorg drivers renders Xorg to fail to start, but not configuring it makes it work fine...

According to pciconf, my Nvidia card is "vgapci0@pci0:1:0:0: class=0x030000 rev=0xa2 hdr=0x00 vendor=0x10de device=0x13b0 subvendor=0x17aa subdevice=0x222e", and configured the driver as Identifier "Card0", Driver "nvidia", and BusID "pci0:1:0:0". Should be correct...

I don't get why it isn't correct.

remiliascarlet: why are you trying to specify busid at all, have several gpus?

yuripv: This laptop has 2 GPU's. An Intel and an Nvidia.

Even if I comment everything out except for "Section" and "EndSection", it really seems to be tripping over `Section "Device"`.

for my sake, assume yes because I have never gotten X (without config) to work with my hardware, on a hp proliant 580 gen9 with 3 nVidia Quadro K6000, and a builtin, it just says no screens found

In my case it's "[01:59:24.245] (II) DAEMON: Running: /usr/local/bin/X -nolisten tcp -background none -seat seat0 vt9 -auth /var/run/sddm/xauth_tKIzuz -noreset -displayfd 15", "[01:59:24.256] (EE) DAEMON: Failed to read display number from pipe", "[01:59:24.256] (WW) DAEMON: Attempt 3 starting the Display server on vt 9 failed", and "[01:59:24.256] (EE) DAEMON: Could not start Display server on

vt 9"

dch, I look forward to asking you about openbao some time in the future when we both have time to talk about it. I am definitely interested in how it works. Thanks!

im looking into jails atm what ip addresses should one use? im not very knowledgable on ip addresses, is there like a range of addresses one can safely use without causing trouble?

OwlWizard: it depends on the purpose of the jail. For example, I have a unifi controller that I want to have an address on my LAN, so it's 192.168.etc.etc. But for my caddy jail and jails that it reverse proxies to, i have them in the 10/8 space, and those are on the loopback interface, inaccessible outside the host unless i have pf forward packets to them

OwlWizard: There's more info regarding ranges here: datatracker.ietf.org/doc/html/rfc1918#section-3

ill have a look at a

hit enter too early

but yea im learning jails atm and the guide doesnt tell me anything about ip-address, so atm its a test jail but im thinking about making one for installing ports into

also can i assume that the host can execute apps inside of a jail?

I was hoping someone could take a look at an issue I'm having with my interface bridging. I have a dual 10G SFP+ card in bridge mode and it works setting up manually, but I can't seem to have the rc.conf settings bring it up during boot. dpaste.org/9kf38

Also it seems that bsd.to is down so I used a different pastebin :/

I was wrong at the top of my paste. An IP address gets assigned to bridge0 but the member interfaces are not added.

I suspect its a timing issue, the commands are being ran too fast?

ah, I figured it out. My two `ifconfig_bridge0` variables were overwriting eachother. I need to combine the ifconfig command into one like `ifconfig_bridge0="inet 192.168.80.12 netmask 255.255.255.0 addm ix0 addm ix1 up"

👍

has Samba recently added some sort of ZFS block cloning feature? i just copy-pasted 3 10GB files on a Windows SMB client / FreeBSD server and they copied instantly

Do I still need nfs to pxe boot FreeBSD 14.1? I just want to pxe boot the installer. :/

ivy, Probably block_cloning is enabled and probably Samba is using it now by default. It's a new feature that recently has rolled through into release.

rwp: yeah, i saw block cloning when it was introduced (mostly because of that one bug :-) but wasn't aware Samba was using it

pretty nice though, even over 10GbE it would usually take long enough to copy that much data that i'd log in and do via ssh instead

Both the source and destination would need to be in the same zfs pool though. 3x 10GB = 30GB ... How much ram do you have for file system buffer cache? It might have all gone into ram.

nah, the pool is only 8x 7200rpm HDDs, there's no way it could have read that much data in a couple of seconds

at most it does ~1GB/s and usually slower

Some commands are aggressive about using the newest features and will opportunistically do all of the fancy stuff if it is available. Sometimes that is good to give testing and to find bugs. Sometimes not so good when it leads to data corruption.

Was both the source and destination on the same pool? It can't do block cloning if it is in different pools.

yes

different filesystems, same pool

Then it would be a candidate for it. Could be super fast in that case.

is there a way to check if a file got block cloned? i guess not as it happens at a fairly low level

truenas.com/community/threads/truen…with-additional-smb-features.113649 says they do block cloning over SMB but doesn't say if they use Samba or not

presumably they do though

I imagine that if you ran a command under truss that it would be visible in the copy options? I don't know.

i was thinking after the fact, i don't care enough to truss smbd :-d

That's something I imagine is available to deduce from some zfs list options but off the top of my head I do not know. Maybe someone else will know?

hi, good evening,