|
How can Samba bring down the system? I noticed some file copies hang. The client eventually gave up and said the server disappeared. I tried restarting the Samba service and that failed. I tried to kill / kill -9 / killall it, but it wouldn't die. # ps aux | grep smb code:
I had never seen this with Samba 3.6 (which ran fine for months). I'm getting a "stuck" process with Samba 4.1 every few days. I tried rebooting and it failed. It goes through the shutdown process but never reboots. I have to manually hold the Power button to kill the system completely. How do I recover from a stuck process like that? FreeBSD 9.3, Samba 4.1.17.
|
# ? Apr 22, 2015 17:47 |
|
|
# ? May 14, 2024 08:36 |
|
Well, after wondering why my Logwatch settings never seemed to work right, I just found out the correct path for the configuration file on FreeBSD isn't mentioned anywhere. Logwatch on FreeBSD uses /usr/local/etc/logwatch/logwatch.conf. I just tried a Google search for that; quote:No results found for "/usr/local/etc/logwatch/logwatch.conf". The man pages and HOWTO say it uses /etc/logwatch. That directory is completely ignored by Logwatch on FreeBSD. When you search for Logwatch "guides", all of them seem to tell you to modify /usr/local/etc/logwatch/defaults/logwatch.conf, which is a big no-no. That file is over-written with an update, anyway. I had to look into the source of /usr/local/sbin/logwatch.pl to see where it was looking. Am I missing something with this? Why cannot I find correct documentation for Logwatch on FreeBSD? Why would their own documentation on it be wrong? "man logwatch" seems to suggest that you use /etc/logwatch. /usr/local/share/doc/logwatch/README and /usr/local/share/doc/logwatch/HOWTO-Customize-LogWatch both say to use "/etc/logwatch/conf/logwatch.conf" I guess another bug report is on my ToDo list...
|
# ? May 5, 2015 18:17 |
|
Xenomorph posted:Well, after wondering why my Logwatch settings never seemed to work right, I just found out the correct path for the configuration file on FreeBSD isn't mentioned anywhere. It's not a bug. The docs may be a bug, because they seem to have just packaged the default documentation, but as a general rule: if the binary is in /usr/local (because you installed it from ports or packages), the config file is in /usr/local, as is the init script
|
# ? May 5, 2015 18:28 |
|
In other news, I finally, finally got NFSv4 with sec=krb5 working. Apart from all the services and config setup that has to be right, it helps to compile a kernel with GSSAPI. Without it, everything appears to work*, but reading/writing files always fails with I/O errors. * I did finally notice an error in the messages log that sent me in the right direction.
|
# ? May 6, 2015 04:13 |
|
Is there anyway to get /dev/gpt to update without a reboot or deleting a partition? I wanted to rename some "/dev/mfisyspdXdX" disks and partitions to things like "gpt/zfs0" or "gpt/swap1". It seems that I either have to delete & remake the partition, or reboot the server to get the changes to stick. There was a bug report opened in 2011 saying that it requires a reboot to update /dev/gpt: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=162690
|
# ? May 12, 2015 20:19 |
|
10.0 is done right and I should "update-freebsd -r 10.1-RELEASE upgrade" ? Isn't there a difference between 10.even and 10.odd?
|
# ? May 14, 2015 17:07 |
|
freebsd-update upgrade -r 10.1-RELEASE is the syntax, but yes 10.1 is just the next version of 10. Remember to install updates after! 10.1 is up to -p9 or something.
|
# ? May 15, 2015 04:46 |
|
Neat. It looks like two of the bugs reports that I submitted apparently made it into yesterday's 4.1.18 samba41 update. https://svnweb.freebsd.org/ports/head/net/samba41/Makefile?r1=386869&r2=386868&pathrev=386869 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=197320 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194046
|
# ? May 21, 2015 17:53 |
Despite having run FreeBSD since 5.0-RELEASE, I'm always looking to optimize my system(s) and I recently discovered two small tools that make a lot of difference, namely iocage and iohyve that I didn't see mentioned in the thread, so I thought I would. iocage is a classic jail management tool, with the exception that it's written from the ground up to make use of ZFS and libvirt, and it's ready for the change in jail configuration that's supposedly coming in FreeBSD 11. On top of this, it's very easy to use and supports templates, cloning, forking, quotas, and thin-provisioning and all sorts of other very useful things you can make use of for trouble-shooting, testing and upgrading whatever you're running in the jail(s). iohyve is a tiny piece of software to put on top of bhyve, the ESXi-like bare-metal hypervisor FreeBSD have been working on for a while and which is now a very impressive piece of software. It makes it possible to create virtualized OS' that run on top of ZFS and use libvirt with a very simple syntax.
|
|
# ? Jun 20, 2015 10:02 |
|
you guys should check out hardenedbsd, security-enhanced fork of freebsd developed by some very smart people; they recently implemented aslr and stack randomization, and they're working on implementing PIE for all base executables iirc
|
# ? Aug 31, 2015 17:56 |
|
reddit liker posted:you guys should check out hardenedbsd, security-enhanced fork of freebsd developed by some very smart people; they recently implemented aslr and stack randomization, and they're working on implementing PIE for all base executables iirc Why not just use OpenBSD?
|
# ? Aug 31, 2015 18:33 |
|
MrDoDo posted:Why not just use OpenBSD? because some people would rather use a freebsd fork? also: quote:Over the July 4th weekend, we implemented randomization of the VDSO (Virtual Dynamic Shared Object). The VDSO is a spot in memory that is shared between the kernel and userland memory. It contains the signal trampoline and time-related code (like gettimeofday(2)). Even though the amount of code is small in the VDSO, it could still theoretically be used to generate ROP gadgets. Removing that piece of determinism makes generating ROP gadgets based on code in the VDSO more difficult. Randomizing the VDSO was the last piece of the address space to randomize. additionally hardenedbsd is pushing their discoveries and implementations upstream to freebsd
|
# ? Aug 31, 2015 19:04 |
|
Anybody been playing with the new and shiny bhyve capabilities?
|
# ? Oct 5, 2015 17:33 |
|
Oh, neat. Nothing I remotely need, but that's sort of secondary. In other news, kerberized nfs4 (with a windows 2003r2 server I have no control over as the kerberos server) is completely mystifying. It stopped working overnight and nothing I've done has worked, short of removing sec=krb5 from the mount options. Oh well, there's always next week.
|
# ? Oct 8, 2015 16:13 |
|
Spotted down the road from my house:
|
# ? Oct 8, 2015 16:33 |
OWLS! posted:Anybody been playing with the new and shiny bhyve capabilities? Tangentially related to the above image, I know FreeBSD gets used in core routers (JunOS), TVs (Panasonic), Playstation 3 and 4, the Mars rovers (specifically the FreeBSD netstack) and many other surprising places - along with some not-so-surprising like a lot of mail servers on the web - but I've never heard of any cars running it. It'd be kinda neat if it was used there as well.
|
|
# ? Oct 8, 2015 18:41 |
|
D. Ebdrup posted:I've been playing around with bhyve a lot, but didn't know this. Since you posted, I've been playing around with it and it's actually really neat. The netstack gets used everywhere, but that's not a great indication of usage. Windows also used it for a long time. JunOS is also pulled from comparatively ancient versions of FreeBSD, even recent versions. FreeBSD leaves a lot to be desired in realtime applications, but there's a realtime patchset and canbus drivers, so it's possible, but nobody's talking about it if so, without even getting into certifying it to work in given scenarios. QNX and VxWorks are still very much kings of that (embedded realtime OS) hill.
|
# ? Oct 8, 2015 19:35 |
|
Doing a big upgrade on my FreeBSD box soon, mostly to get more space on the boot device. Anyone have a guide they like to use for something like this? Is it better to just do a "fresh" install on the new drive and then move /usr /etc separately?
|
# ? Dec 7, 2015 19:45 |
|
roadhead posted:Doing a big upgrade on my FreeBSD box soon, mostly to get more space on the boot device. Anyone have a guide they like to use for something like this? /etc /home /usr/local/etc SSH keys Jails running pkg_info to get a list of installed ports Since I'm lazy I think I will just keep the existing drive just in case I missed something.
|
# ? Dec 10, 2015 02:13 |
|
I'm getting some horrific kernel panics in 9.3 when I pull a HDD. I tested it in the past (9.1/9.2 days), and it worked fine. I'd pull a drive, the kernel would log that a drive was disconnected. I'd plug the drive back in, and the system continued to purr like nothing happened in the first place. Now I'm actually trying to replace drives, and I'm having a problem. My plan was to pull a small drive (3TB), replace it with a larger drive (6TB), then let ZFS rebuild. It takes about 30 hours to resilver for each replaced drive, and I have 12 drives to replace. When I pull an old drive, the system has an immediate kernel panic and reboots. No error messages. Nothing logged to /var/log. Just a dump and a reboot. Obviously I don't want or need this thing uncleanly restarting itself over and over. It didn't do this before. I haven't pulled a drive from it since it had 9.1, though. PowerEdge R720xd PERC H310 HBA mfi driver 9.3 "-30" update. Anyone know why it would kernel panic when it loses a drive??
|
# ? Dec 10, 2015 18:42 |
|
Bluecobra posted:I'm in the same boat, I've been putting off upgrading my FreeBSD 9.0 box for some time now. Now that the installer makes ZFS root disks easy, I plan on blowing the whole thing out and starting over. Off the top of my head I plan on backing up the following: I'm moving from a 32 gb IDE ssd to a new 512 gb Samsung 850 Evo because /usr and /var are always drat near full.
|
# ? Dec 11, 2015 15:34 |
|
evol262 posted:The netstack gets used everywhere, but that's not a great indication of usage. Windows also used it for a long time. JunOS is also pulled from comparatively ancient versions of FreeBSD, even recent versions. Just a note, Junos 15.1 upgrades the underlying OS from 6.1 to FreeBSD 10. http://www.juniper.net/techpubs/en_US/junos15.1/information-products/topic-collections/release-notes/15.1/topic-83366.html#jd0e3809 Only for select platforms running Intel CPUs though.
|
# ? Dec 12, 2015 20:27 |
|
OWLS! posted:Anybody been playing with the new and shiny bhyve capabilities? i know my friend lattera absolutely loves bhyve, so it might be worth picking his brain if you're on twitter. he's also on the binrev irc channel, irc.binrev.net #binrev
|
# ? Dec 14, 2015 15:11 |
|
roadhead posted:I'm moving from a 32 gb IDE ssd to a new 512 gb Samsung 850 Evo because /usr and /var are always drat near full. i can think of a much better reason to do that and it doesn't involve disk space
|
# ? Dec 14, 2015 15:21 |
|
protip: when installing via console make sure your /etc/boot.conf containscode:
and not code:
|
# ? Feb 21, 2016 03:25 |
|
The Third Man posted:protip: when installing via console make sure your /etc/boot.conf contains can you open a bug report on this please?
|
# ? Feb 21, 2016 19:02 |
|
roadhead posted:I'm moving from a 32 gb IDE ssd to a new 512 gb Samsung 850 Evo because /usr and /var are always drat near full. OK I finally did this, made all the new partitions, dump|restored the data one partition at a time. Reboot, change the boot order in the BIOS, so far so good. The one thing both "guides" I was following fail to mention, and that actually was sort of important, was flags for the /tmp - apparently there is a t attribute you need set for that to work properly. Only noticed when I went to start up TMUX code:
|
# ? Feb 29, 2016 20:31 |
|
can anyone who isn't poo poo at PF like me tell me why traffic on port 80 is getting to my host but not through the NAT to the nginx jail i have running? /etc/pf.conf: code:
code:
code:
RISCy Business fucked around with this message at 18:12 on May 5, 2016 |
# ? May 5, 2016 18:06 |
|
Is net.inet.ip.forwarding=1 ? I have something similar (that works) set up at work, so I can take a look tomorrow. In other news I've almost finished writing an nss module that forwards passwd and group queries over TCP , so you can get identical username:uid and group:gid mappings. I'd love it if the AD domain at work upgraded to a functional level that includes working uid/uid fields, but in the meantime etc.
|
# ? May 6, 2016 00:21 |
|
Computer viking posted:Is net.inet.ip.forwarding=1 ? code:
|
# ? May 6, 2016 00:23 |
|
Do you have the sysctl set for ip forwarding? net.inet.ip.forwarding Does the host machine have an ip address on lo1? Does the jail know to hit up that host ip address as it's gateway?
|
# ? May 6, 2016 00:31 |
|
EvilMoFo posted:Does the jail know to hit up that host ip address as it's gateway? yes to the other two but can you elaborate on this? i'm new to jails and PF specifically, this is a pet project of mine and i'm trying to learn more.
|
# ? May 6, 2016 00:35 |
|
Try "route add default 10.0.0.x" in the jail, where x is the IP of the host. If that helps, you can set defaultrouter="10.0.0.x" in the jail's rc.conf. Also, read this for more details: http://www.freebsd.no/doc/handbook/network-routing.html Computer viking fucked around with this message at 10:37 on May 6, 2016 |
# ? May 6, 2016 10:00 |
|
nvm, fixed it
RISCy Business fucked around with this message at 16:38 on May 6, 2016 |
# ? May 6, 2016 15:46 |
|
Oh right. Maybe you need to pass port 80 from any to your external IP?
|
# ? May 6, 2016 15:56 |
|
I'm running freeNAS and I've got a few thing installed manually in jails because freenas plugins never work/update correctly. I updated freeNAS and now portmaster in the jails is bitching about UNAME_r and OSVERSION do not agree on major version number. What's the procedure to upgrade them?
poverty goat fucked around with this message at 21:24 on May 20, 2016 |
# ? May 20, 2016 21:05 |
|
I just bought a new computer and plan on making my old one into a freeNAS box. I'm told that FreeNAS forcefully uses the entire HD that it's installed to, regardless of the size, 1GB or 100TB. Is there any benefit to having more space in that install HD or would I be just fine with a 16gb SATA DOM?
|
# ? Sep 10, 2016 19:33 |
|
YouTuber posted:I just bought a new computer and plan on making my old one into a freeNAS box. I'm told that FreeNAS forcefully uses the entire HD that it's installed to, regardless of the size, 1GB or 100TB. Is there any benefit to having more space in that install HD or would I be just fine with a 16gb SATA DOM? don't use a hard drive at all, get a pair of usb drives.
|
# ? Sep 10, 2016 22:03 |
|
thebigcow posted:don't use a hard drive at all, get a pair of usb drives. I've done this and I agree. But in my case the Kimgston USB drives I used keep crapping out on me so antidotal evidence suggests that you spring for quality drives, 16gb was more then enough in my use case.
|
# ? Sep 21, 2016 05:05 |
|
|
# ? May 14, 2024 08:36 |
|
Nystral posted:antidotal evidence Unrelated, I have an NFSv4 problem. Granted, solving a problem with NFSv4+krb5 seems to be like regex ("now you have two problems"), but I think I've got it very nearly sorted. I have a FreeBSD 10.3 fileserver, a FreeBSD 10.3 client, base kerberos (so heimdal), and a windows server 2003 R2 domain as the kerberos server. Following this it's actually working fine ... for some hours, and then the mounts completely die on the client: All IO to the NFS mount including umount gets stuck permanently in rpccon, and the only real way out is to reboot the client. It works fine for a while again after rebooting. Of course, there isn't a single word about this in the logs on either machine. At a random guess, there is a kerberos ticket timing out in there somewhere ... but which one, and how do I renew it?
|
# ? Sep 21, 2016 08:15 |