Re: IPFW transparent VS dummynet rules
On Sun, Jan 8, 2012 at 1:00 PM, Ian Smith smi...@nimnet.asn.au wrote: On Sat, 7 Jan 2012, budsz wrote: [..] keyword instead of an explicit address. The search terminates if this rule matches. Note particularly the last sentence. You'll have to do your dummynet piping first, if it is to apply also to forwarded packets. (sysctl) net.inet.ip.fw.one_pass: 1 When set, the packet exiting from the dummynet pipe or from ng_ipfw(4) node is not passed though the firewall again. Other- wise, after an action, the packet is reinjected into the firewall at the next rule. It seems that you may have one_pass set to 1. Set to 0, packets will continue through the ruleset on exit from pipe/s, so to your fwd rule. cheers, Ian Thank you very much, lazy to read ipfw(8) :) pipe pipe_nr Pass packet to a dummynet ``pipe'' (for bandwidth limitation, delay, etc.). See the TRAFFIC SHAPER (DUMMYNET) CONFIGURATION Section for further information. The search terminates; however, on exit from the pipe and if the sysctl(8) variable net.inet.ip.fw.one_pass is not set, the packet is passed again to the firewall code starting from the next rule. -- budsz No problem. However it's considered good form to also copy responses cc'd back to the two lists this thread appears on, for the archives. Not that I need the credit, but it shows that the advice was useful, and that other list members need not also respond, thinking it unresolved. cheers, Ian OK,thank you for reminding me :) TIA -- budsz ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
anybody close to Bristol, UK, wants to give a talk on FreeBSD and numerical analysis to UG students?
Introduction: I'll be teaching computer based modelling to year 1 mechanical engineering students. The unit is based around Matlab, which is not ideal, in my opinion, but is beyond my control. The unit is pretty low level - I have to start from loops and conditional statements, but ultimately I want them to be able to tackle numerical solution of algebraic and diff. equations and a bit of graphics. I want to complement Matlab by several lectures giving students a broader view of numerical computing and related subjects. For example, I'll probably talk about vector vs raster graphics and related software, precision of floating point calculations, intro to latex, importance of standards in software, etc. What I'm looking for: I'd like to have one lecture on FreeBSD and what it can do for numerical analysis. I'm looking for somebody who can come to Bristol on a Tuesday between 31-FEB-2012 and 20-MAR-2012 and give a 50 min lecture from 1400 to 1450 to about 120-150 students. The exact details of the talk are not that important. Some of them would've heard of linux, probably not of FreeBSD. Some of them would've used macs, but unlikely any software beyound MS office. The talk can just raise the students' awareness that numerical analysis tools available via FreeBSD ports are an alternative to Matlab. I'll pay the travel expences (have to double check with the finance office) but cannot pay for the talk itself. If you are interested, or have another idea, please get in touch directly. Thanks Anton -- Anton Shterenlikht Room 2.6, Queen's Building Mech Eng Dept Bristol University University Walk, Bristol BS8 1TR, UK Tel: +44 (0)117 331 5944 Fax: +44 (0)117 929 4423 ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re[9]: high load system do not take all CPU time
Also I notice next: in case of overload 'ping localhost' or any IP this router has get timeouts about 50-100ms, pinging any external host on LAN or Internet get normal results: 5ms LAN, 40ms Internet. I do not think this issue related to re0 interface or its driver. This is related to kernel and its structures. -- С уважением, Коньков mailto:kes-...@yandex.ru ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re[10]: high load system do not take all CPU time
КЕ Also I notice next:
КЕ in case of overload 'ping localhost' or any IP this router has get
КЕ timeouts about 50-100ms, pinging any external host on LAN or Internet
КЕ get normal results: 5ms LAN, 40ms Internet.
КЕ I do not think this issue related to re0 interface or its driver.
КЕ This is related to kernel and its structures.
details:
which system queues or buffers can cause such bad results for
localhost pinging?
# ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1): 56 data bytes
64 bytes from 127.0.0.1: icmp_seq=0 ttl=64 time=122.377 ms
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=53.025 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=36.214 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=85.151 ms
64 bytes from 127.0.0.1: icmp_seq=4 ttl=64 time=105.704 ms
64 bytes from 127.0.0.1: icmp_seq=5 ttl=64 time=1.145 ms
64 bytes from 127.0.0.1: icmp_seq=6 ttl=64 time=26.240 ms
64 bytes from 127.0.0.1: icmp_seq=7 ttl=64 time=37.532 ms
64 bytes from 127.0.0.1: icmp_seq=8 ttl=64 time=20.161 ms
64 bytes from 127.0.0.1: icmp_seq=9 ttl=64 time=7.876 ms
64 bytes from 127.0.0.1: icmp_seq=10 ttl=64 time=36.441 ms
64 bytes from 127.0.0.1: icmp_seq=11 ttl=64 time=0.840 ms
64 bytes from 127.0.0.1: icmp_seq=12 ttl=64 time=45.483 ms
64 bytes from 127.0.0.1: icmp_seq=13 ttl=64 time=29.629 ms
64 bytes from 127.0.0.1: icmp_seq=14 ttl=64 time=0.025 ms
64 bytes from 127.0.0.1: icmp_seq=15 ttl=64 time=86.228 ms
64 bytes from 127.0.0.1: icmp_seq=16 ttl=64 time=141.489 ms
64 bytes from 127.0.0.1: icmp_seq=17 ttl=64 time=118.011 ms
64 bytes from 127.0.0.1: icmp_seq=18 ttl=64 time=14.077 ms
64 bytes from 127.0.0.1: icmp_seq=19 ttl=64 time=0.599 ms
64 bytes from 127.0.0.1: icmp_seq=20 ttl=64 time=0.041 ms
64 bytes from 127.0.0.1: icmp_seq=21 ttl=64 time=59.191 ms
64 bytes from 127.0.0.1: icmp_seq=22 ttl=64 time=36.222 ms
64 bytes from 127.0.0.1: icmp_seq=23 ttl=64 time=3.278 ms
64 bytes from 127.0.0.1: icmp_seq=24 ttl=64 time=153.970 ms
64 bytes from 127.0.0.1: icmp_seq=25 ttl=64 time=71.832 ms
64 bytes from 127.0.0.1: icmp_seq=26 ttl=64 time=0.740 ms
64 bytes from 127.0.0.1: icmp_seq=27 ttl=64 time=22.389 ms
64 bytes from 127.0.0.1: icmp_seq=28 ttl=64 time=6.637 ms
64 bytes from 127.0.0.1: icmp_seq=29 ttl=64 time=2.888 ms
64 bytes from 127.0.0.1: icmp_seq=30 ttl=64 time=27.595 ms
64 bytes from 127.0.0.1: icmp_seq=31 ttl=64 time=59.914 ms
64 bytes from 127.0.0.1: icmp_seq=32 ttl=64 time=8.892 ms
^C
--- 127.0.0.1 ping statistics ---
33 packets transmitted, 33 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.025/43.086/153.970/43.818 ms
last pid: 34214; load averages: 4.02, 4.13, 4.38 up 8+20:50:08
17:18:01
276 processes: 6 running, 251 sleeping, 16 waiting, 3 lock
CPU 0: 16.5% user, 0.0% nice, 12.2% system, 48.9% interrupt, 22.3% idle
CPU 1: 12.2% user, 0.0% nice, 13.7% system, 60.4% interrupt, 13.7% idle
CPU 2: 8.6% user, 0.0% nice, 8.6% system, 68.3% interrupt, 14.4% idle
CPU 3: 10.8% user, 0.0% nice, 4.3% system, 72.7% interrupt, 12.2% idle
Mem: 638M Active, 2804M Inact, 313M Wired, 135M Cache, 112M Buf, 8736K Free
Swap: 4096M Total, 16M Used, 4080M Free
PID USERNAME PRI NICE SIZERES STATE C TIME WCPU COMMAND
12 root -72- 0K 160K *per-i 3 67.7H 95.26% {swi1: netisr 3}
12 root -72- 0K 160K CPU11 46.5H 84.62% {swi1: netisr 1}
12 root -72- 0K 160K *per-i 2 24.5H 36.57% {swi1: netisr 2}
11 root 155 ki31 0K32K RUN 0 139.3H 34.81% {idle: cpu0}
12 root -72- 0K 160K *per-i 3 19.6H 32.57% {swi1: netisr 0}
11 root 155 ki31 0K32K RUN 1 141.2H 19.48% {idle: cpu1}
11 root 155 ki31 0K32K CPU22 140.2H 17.43% {idle: cpu2}
11 root 155 ki31 0K32K CPU33 141.5H 15.92% {idle: cpu3}
12 root -92- 0K 160K WAIT0 26.4H 13.09% {irq256: re0}
93929 root240 15392K 5616K select 0 86:57 6.88% snmpd
13 root -16- 0K32K sleep 2 599:49 4.83% {ng_queue1}
13 root -16- 0K32K sleep 2 600:32 4.20% {ng_queue0}
13 root -16- 0K32K sleep 3 600:57 3.86% {ng_queue3}
13 root -16- 0K32K sleep 0 600:05 3.76% {ng_queue2}
34145 cacti 270 12000K 3096K select 0 0:00 1.22% snmpwalk
34185 cacti 520 32256K 16604K nanslp 2 0:00 0.93% php
86746 root200 139M 57632K select 2 17:09 0.29% {mpd5}
86746 root200 139M 57632K select 0 0:00 0.29% {mpd5}
86746 root200 139M 57632K select 2 0:00 0.29% {mpd5}
32865 freeradius 20 -20 151M 123M usem3 1:31 0.24% {radiusd}
32865 freeradius 20 -20 151M 123M usem2 1:31 0.24% {radiusd}
32865 freeradius 20 -20 151M 123M usem0 1:25 0.24% {radiusd}
1 usersLoad 4.88 4.34 4.45 Jan 8 17:18
Mem:KBREALVIRTUAL
Re: Installing FreeBSD ver. 8.2
On Sat 2012-01-07 15:05:55 UTC-0800, leeoliveshackelf...@surewest.net (leeoliveshackelf...@surewest.net) wrote: (5) What device driver must be installed for the sound board to be able to receive a m.i.d.i. over u.s.b. signal? This signal would be generated by a musician's keyboard, and would control a music synthesizer application, to be installed. I could find no mention of this topic in the Handbook. There are USB to MIDI in/out hardware devices available. Last I looked they were selling for about US$25 on eBay. I bought one about two years ago and use it in Ubuntu Linux. I don't think I ever tested if it worked in FreeBSD but I suspect it would. I also have a Casio WK3300 keyboard with USB output. I don't think it was supported by FreeBSD, but Ubuntu Linux (10.04 Lucid) recognised it. The sound card you use is irrelevant as to whether you can use MIDI over USB. In fact MIDI can be used for non-audio applications, for example lighting rigs. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: Installing FreeBSD ver. 8.2
On Sun, Jan 08, 2012 at 12:32:25AM +0100, Polytropon wrote: On Sat, 7 Jan 2012 15:05:55 -0800 (PST), leeoliveshackelf...@surewest.net wrote: (1) Does anyone know how to get FreeBSD to read the motherboard name? This name, on an xw4400, starts with HP followed by a eleven digits, and is given in Windows XP as Full Computer Name on the Computer Name tab of the System Properties window. Among other purposes, this name is used by Novell network operating system to distinguish hosts on a subnet. The OS provides the output of dmesg and maybe the output of pciconf -lv, as well as the sysctl value dev.acpi.0.%desc which may contain the required information. However, I'm sure there is a program in the ports collection that can be used to obtain that kind of information. Try: dmesg | grep HP sysctl -a | grep HP pciconf -lv | less and see if there's such a number mentioned. Maybe you can also use acpidump to retrieve that information from the ACPI datasets. (2) I cannot get the find command to locate files that I believe were installed at the time of sysinstall. If I understand the Handbook correctly, when one runs find from the / directory, it is supposed to inspect all directories and subdirectories of all partitions, which it is not doing. What concept am I missing here? It would be easier to answer if you could provide the find command line you've been running. :-) See man find for more information. Basically, find / -name namespec -type f should be sufficient to access all partitions currently mounted to search for namespec specified regular files. (3) I thought that I would obtain a better understanding of the file system by running man heir. This command runs fine under sh. When I switch to my preferred shell, which is bash, I type, and receive echo on the screen, man hei. As soon as I depress r, the entire previously entered command echos to the screen, starting where the r should have appeared. In checking the bash manual, it says that this response is correct for control-r. I could not find non-shift-r to be called out as a command. Am I doing something wrong? Is this a hardware bug? Is this a software bug? Is there something that needs to be defined or undefined in a configuration file? No, bash's configuration files provided after install should be fine. However, I think you have a typo. The command you're intending to run is man hier (hierarchy). I've tested both csh and bash here, both allow the command to be entered without any interruption. When I type man hei followed by Ctrl+R, I get the following output: (reverse-i-search)`': man hei. (4) Not having very good luck with the find command, I thought I would try to use the locate command. To use this command, one must create a database. On www.us-webmasters.com, I read that this database could be constructed by running the command #usr/libexec/locate.updatedb. The required task is usually executed by the system's night job at 3:00 once a week. The script that will be run is /etc/periodic/weekly/310.locate which you could run manually. It will deal with the correct call of /usr/libexec/locate.updatedb (instead of running it as root!). The thing to run is periodic(8): # periodic weekly That will also update other useful stuff. Regards, -- Frank Contact info: http://www.shute.org.uk/misc/contact.html pgpBnZ7iYcXM2.pgp Description: PGP signature
problem with terminal capabilities when using terminal from within emacs
From within emacs, if I invoke a new multi-term buffer with C-c c, the TERM environment variable is set to eterm-color and a TERMCAP variable is also set and everything works fine. However, if I ssh to a remote host from within this buffer, only the TERM variable remains set and the keybindings don't work. If I manually set the TERMCAP variable nothing changes. Is there a way to ssh to a remote host from within a multi-term buffer and have the terminal capabilities work out? I'm running emacs 23.3 on 8-STABLE. Thanks. -- -- Public key: http://gly.ath.cx/pub_key.txt -- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
OpenBSD disk on FreeBSD.
I have a usb external hard drive from a dead OpenBSD x86 system, and I want to mount the drive on $ uname -a FreeBSD AngkorWat 8.2-RELEASE FreeBSD 8.2-RELEASE #0: Thu Feb 17 02:41:51 UTC 2011 r...@mason.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC amd64 $ fdisk on the FreeBSD system looks like Disk name: da0FDISK Partition Editor DISK Geometry: 20023 cyls/255 heads/63 sectors = 321669495 sectors (157065MB) Offset Size(ST)End Name PType Desc SubtypeFlags 0 63 62- 12 unused0 63 321669432 321669494da0s4 4 OpenBSD FFS 166 = 321669495 3465 321672959- 12 unused0 The OpenBSD fstab is gone, and the backup copy is (of course) on the disk I want to mount. How do I go about mounting this disk on FreeBSD? The following don't work: # mount /dev/da0 mnt mount: /dev/da0 : Invalid argument # mount /dev/da0s4 mnt mount: /dev/da0s4 : Operation not permitted # ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: OpenBSD disk on FreeBSD.
On Sun, 08 Jan 2012 14:42:01 -0500, R. Clayton wrote: The OpenBSD fstab is gone, and the backup copy is (of course) on the disk I want to mount. How do I go about mounting this disk on FreeBSD? The following don't work: # mount /dev/da0 mnt mount: /dev/da0 : Invalid argument # mount /dev/da0s4 mnt mount: /dev/da0s4 : Operation not permitted According to http://www.openbsd.org/faq/faq14.html you can address the partitions on that slice (da0s4) like you would access them on FreeBSD. # mount -o ro -t ufs /dev/da0s4a /mnt This will give you the root partition. Other partitions can be mounted separately (or in sequence, after unmounting /mnt), or they can be mounted into the /mnt tree to their original locations, e. g. # mount -o ro -t ufs /dev/da0s4e /mnt/tmp # mount -o ro -t ufs /dev/da0s4f /mnt/var # mount -o ro -t ufs /dev/da0s4g /mnt/usr # mount -o ro -t ufs /dev/da0s4h /mnt/home I would suggest you _first_ mount each partition individually and see from its content what it has been designated to. If you can find a copy of the original /etc/fstab somewhere on one of the partitions, you can use its content to avoid guessing. :-) -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Mounting (read/write) ext4
Can the upcoming FreeBSD 9 mount ext4 file systems out of the box? Sent from my HTC.___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: sour grapes .. was FreeBSD Kernel Internals Documentation
C. P. Ghost cpgh...@cordula.ws, 2012-01-01 20:30 (+0100): On Sun, Jan 1, 2012 at 7:41 PM, doug d...@fledge.watson.org wrote: That said, FreeBSD has a giant disadvantage in the desktop world. In trying to find if there will be any sort for my current laptop I came across a comment from Robert Noland saying that Xorg is becoming more and more Linux centric. That is a problem the FreeBSD project can not overcome. Did he mean frameworks like evdev(4) and so? http://www.x.org/archive/X11R7.5/doc/man/man4/evdev.4.html That man-page really doesn't explain what has changed. The new thing is that X.org has done away with hald. Stuff like this really ought to be backported to FreeBSD, either directly or by providing more Linuxisms on our side. It also seems rather easy, considering we already have devd which could be used to feed evdev with information about input devices. There's no /technical/ reason why it can't be done. True. He probably also meant stuff like Kernel mode-setting (KMS) and GEM. All this is being worked on in FreeBSD as well: http://wiki.freebsd.org/Intel_GPU -- http://hack.org/mc/ Warning! Plain text e-mail, please. HTML e-mail deleted unread. OpenPGP: 673B 563E 3C78 1BA0 6525 2344 B22E 2C10 E4C9 2FA5 ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: OpenBSD disk on FreeBSD.
On Sun, 08 Jan 2012 16:13:38 -0500 (EST), R. Clayton wrote: Thanks for your reply to my message. you can address the partitions on that slice (da0s4) like you would access them on FreeBSD. mount /dev/da0s4a didn't work. However, looking in dmesg I saw WARNING: R/W mount of /mnt/backups denied. Filesystem is not clean - run fsck If you're intending to copy data FROM the disk, i. e. you want to _recover_ data, do _not_ mount something R/W (because you don't have to). Always use -o ro, at least in early stages. The denial of the mount is justified due to file system errors which are correctly detected (and then corrected by fsck). After running fsck -t ufs /dev/da0s4, mount /dev/da0s4a still didn't work, but # mount /dev/da0s4 mnt did work: # cat mnt/angkor-wat/etc/fstab # fs-spec mount-point type options frequency pass-no /dev/wd0a / ffs rw 1 1 /dev/wd0h /home ffs rw,nodev,nosuid 1 2 /dev/wd0d /tmp ffs rw,nodev,nosuid 1 2 /dev/wd0g /usr ffs rw,nodev 1 2 /dev/wd0e /var ffs rw,nodev,nosuid 1 2 /dev/sd0a /mnt/backups ffs rw,nodev,nosuid 1 2 /dev/sd0d /mnt/storage ffs rw,nodev,nosuid 1 2 /dev/cd0a /cdrom cd9660 ro,noauto 0 0 # However, I can only mount the sd0a (OpenBSD) partition; [...] That is the root partition /, and suffixes d, e, g, h should be available. [...] the sd0d partition is nowhere to be found: $ sudo fsck -t ufs /dev/da0s4d ** /dev/da0s4d Cannot find file system superblock ioctl (GCINFO): Inappropriate ioctl for device fsck_ufs: /dev/da0s4d: can't read disk label $ On the other hand, $ sudo fsck -t ufs /dev/da0s4a ** /dev/da0s4a Cannot find file system superblock ioctl (GCINFO): Inappropriate ioctl for device fsck_ufs: /dev/da0s4a: can't read disk label $ so maybe the partition names are wrong. Can you obtain partitioning information via bsdlabel (disklabel) for the s4 slice to check if FreeBSD can identify the other partitions properly. If you visit the mailing list archives, you will find hints to procedures and software to do a deeper analysis of the structures that need to be accessed, such as disk labels, file system data and other stuff. -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: OpenBSD disk on FreeBSD.
Thanks for your reply to my message. you can address the partitions on that slice (da0s4) like you would access them on FreeBSD. mount /dev/da0s4a didn't work. However, looking in dmesg I saw WARNING: R/W mount of /mnt/backups denied. Filesystem is not clean - run fsck After running fsck -t ufs /dev/da0s4, mount /dev/da0s4a still didn't work, but # mount /dev/da0s4 mnt did work: # cat mnt/angkor-wat/etc/fstab # fs-spec mount-point type options frequency pass-no /dev/wd0a / ffs rw 1 1 /dev/wd0h /home ffs rw,nodev,nosuid 1 2 /dev/wd0d /tmp ffs rw,nodev,nosuid 1 2 /dev/wd0g /usr ffs rw,nodev 1 2 /dev/wd0e /var ffs rw,nodev,nosuid 1 2 /dev/sd0a /mnt/backups ffs rw,nodev,nosuid 1 2 /dev/sd0d /mnt/storage ffs rw,nodev,nosuid 1 2 /dev/cd0a /cdrom cd9660 ro,noauto 0 0 # However, I can only mount the sd0a (OpenBSD) partition; the sd0d partition is nowhere to be found: $ sudo fsck -t ufs /dev/da0s4d ** /dev/da0s4d Cannot find file system superblock ioctl (GCINFO): Inappropriate ioctl for device fsck_ufs: /dev/da0s4d: can't read disk label $ On the other hand, $ sudo fsck -t ufs /dev/da0s4a ** /dev/da0s4a Cannot find file system superblock ioctl (GCINFO): Inappropriate ioctl for device fsck_ufs: /dev/da0s4a: can't read disk label $ so maybe the partition names are wrong. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: ZFS Root Won't Mount - Unknown Filesystem
On Sat, Jan 7, 2012 at 7:44 AM, Drew Tomlinson d...@mykitchentable.net wrote: I'm attempting a new install of 9.0-RC3 amd64. My system has 4 500 GB drives. Using this tutorial as a guide: http://www.aisecure.net/2011/11/28/root-zfs-freebsd9/ When I built my ZFS-root system, I did most of these things, except I had a slightly different setup for the root filesystem itself. What I did is akin to: # zpool import -o altroot=/target -o cachefile=/tmp/zpool.cache zroot # zfs set mountpoint=/ zroot # cp -a /tmp/zpool.cache /target/boot/zfs/zpool.cache I did not manually set mountpoints for anything but the root filesystem. It sounds like manually-set mountpoints are the way most people do it, but you can try to clear the mountpoints by doing: # zfs inherit mountpoint zroot/fs for each filesystem in your root pool. I created a boot and a freebsd-zfs partition on each drive. Then I created a raid1z pool using all 4 drives. I followed the rest of the tutorial exactly and ensured that I copied the zpool.cache to boot/zfs. When I try to boot my new system, it all goes fine up until it's time to mount zfs:zroot. It fails with an error 2 unknown filesystem error. I don't know if this means anything but at the mountfrom prompt, the system will not accept any keyboard input. Same keyboard works fine when booted into LiveCD. I had that same problem with mine for a while, and it turned out that importing with the altroot option implies cachefile=none; until I realized I needed to also specify cachefile=/some/path, I had accidentally ended up with a /boot/zpool.cache that didn't actually reference any zpools. Unfortunately because I can't figure out how to get a LiveCD type environment with sshd running, I can't copy and paste exact error messages or command outputs. I was using PXE/NFS booting to install this machine, so unfortunately I can't help you here. I've searched and the two things that seem to be important are that there's a zpool.cache file and that the zfs partitions are correct. A 'gpart show -l' shows my partitions something close to this: 34 big number ada0 GPT (456G) 34 128 1 null (128K) 162 big number 2 disk0 (456G) What have I done wrong and what do I need to do to get my zfs:zroot pool mounted as root? It sounds like you're almost there! My guess is that the cache file is what is missing/incorrect. Reading over some man pages, make sure you don't do a zpool export before you copy the cache file; exporting the array removes it from the cache and/or deletes the cache file entirely. If you end up with a LiveCD that lets you copy these things, it might help to see # zpool list -o name,altroot,cachefile # zpool status # zfs list -o name,mountpoint,mounted Hope some of this helps. -- Matt Mullins ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: ZFS Root Won't Mount - Unknown Filesystem -- SOLVED
Whoops, I missed this message before posting my reply a few minutes ago. On Sat, Jan 7, 2012 at 9:36 AM, Drew Tomlinson d...@mykitchentable.net wrote: Yes, although I've read that 'zfs set mountpoint=/ zroot' is acceptable as well. I set mine to / after trying to import pool with '-o altroot=/mnt' in LiveCD. When mountpoint was legacy, altroot didn't work right. Opinions on / vs. legacy? Most of the FreeBSD guides seem to think legacy is the way to go, but I much prefer / myself. The main difference is with legacy, one sets a mountpoint option on each filesystem under it (which does, indeed, undermine the altroot facility), whereas / lets the other filesystems inherit their mountpoint from their path in the zpool. / seems to be a little closer to its Solaris usage, which is still the majority of the documentation you'll find on ZFS on the internet (and even in the man pages distributed with FreeBSD). -- Matt Mullins ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: how to force 'device' sources to not compile?
On 01/06/2012 03:39 PM, Коньков Евгений wrote: I have errors while compile kernel === et (all) cc -O2 -pipe -fno-strict-aliasing -Werror -D_KERNEL -DKLD_MODULE -nostdinc -DHAVE_KERNEL_OPTION_HEADERS -include /usr/obj/usr/src/sys/KES_KERN_v9/opt_global.h -I. -I@ -I@/contrib/altq -finline-limit=8000 --param inline-unit-growth=100 --param large-function-growth=1000 -fno-common -g -I/usr/obj/usr/src/sys/KES_KERN_v9 -mno-align-long-strings -mpreferred-stack-boundary=2 -mno-sse -mno-mmx -msoft-float -ffreestanding -fstack-protector -std=iso9899:1999 -fstack-protector -Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual -Wundef -Wno-pointer-sign -fformat-extensions -Wmissing-include-dirs -fdiagnostics-show-option -c /usr/src/sys/modules/et/../../dev/et/if_et.c /usr/src/sys/modules/et/../../dev/et/if_et.c: In function 'et_dma_alloc': /usr/src/sys/modules/et/../../dev/et/if_et.c:782: error: 'ET_RING_ALIGN' undeclared (first use in this function) /usr/src/sys/modules/et/../../dev/et/if_et.c:782: error: (Each undeclared identifier is reported only once /usr/src/sys/modules/et/../../dev/et/if_et.c:782: error: for each function it appears in.) /usr/src/sys/modules/et/../../dev/et/if_et.c:790: error: 'ET_STATUS_ALIGN' undeclared (first use in this function) /usr/src/sys/modules/et/../../dev/et/if_et.c:845: error: 'struct et_softc' has no member named 'sc_rx_mini_tag' /usr/src/sys/modules/et/../../dev/et/if_et.c:854: error: 'struct et_softc' has no member named 'sc_rx_tag' /usr/src/sys/modules/et/../../dev/et/if_et.c:864: error: 'struct et_softc' has no member named 'sc_tx_tag' how to disable 'et' from compiling? This error is from the kernel build process attempting to build the code for a kernel module. Define either MODULES_OVERRIDE or WITHOUT_MODULES as detailed in make.conf(5) to avoid building the if_et module. Note that the syntax of the MODULES_OVERRIDE and WITHOUT_MODULES variables consists of a space-delimited list of directory names found under sys/modules/ and not the canonical module names as found in /boot/kernel/, so some footwork is necessary. -- Fuzzy love, -CyberLeo Technical Administrator CyberLeo.Net Webhosting http://www.CyberLeo.Net cyber...@cyberleo.net Furry Peace! - http://.fur.com/peace/ ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: How To Enable ls Color?
On 1/8/2012 6:14 PM, Polytropon wrote: On Sun, 08 Jan 2012 16:04:17 -0800, Drew Tomlinson wrote: I've installed 9.0-RC3 amd64. I'm trying to enable color output for ls. I've issued the basic 'ls -Gla' but output is not colored. Yet if I can get colorized output by providing color codes (echo ^[[34mhello^[[37m produces a blue hello) at the command line so I know my terminal is capable. Is there some other secret? This is a new install and I'm just trying to set things up. Put setenv LSCOLORS ExGxdxdxCxDxDxBxBxegeg in the csh's initialisation file (typically ~/.cshrc for local use, /etc/csh.cshrc for global effect) and maybe setup an alias: alias ls 'ls -FG' alias ll 'ls -laFG' However, ls should provide colored output even if you don't set the $LSCOLORS variable. It should work with the default terminal emulation (cons25 or cons25l1). From here: http://www.puresimplicity.net/~hemi/freebsd/misc.html CLICOLOR=YES;export CLICOLOR LSCOLORS=ExGxFxdxCxDxDxhbadExEx;export LSCOLORS I last updated that page a while ago...But it still seems to be working for me. :) You shouldn't really have to muck around with term type or anything... Cheers, Josh -- Josh Tolbert h...@puresimplicity.net || http://www.puresimplicity.net/~hemi/ Security is mostly a superstition. It does not exist in nature, nor do the children of men as a whole experience it. Avoiding danger is no safer in the long run than outright exposure. Life is either a daring adventure, or nothing. -- Helen Keller ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
