Re: mountroot
2011/6/29 Warren Block wbl...@wonkity.com On Wed, 29 Jun 2011, Dick Hoogendijk wrote: Op 29-6-2011 21:19, Trond Endrestøl schreef: On Wed, 29 Jun 2011 21:18+0200, Dick Hoogendijk wrote: Op 29-6-2011 21:15, Trond Endrestøl schreef: On Wed, 29 Jun 2011 20:42+0200, Dick Hoogendijk wrote: I'm a little desperade. I installed a mirrored ZFS freebsd system in a VM the other day and all went well. Now I did the same procedure on a real systrem with two drives and I can't get the system to boot properly. Everytime it halts at the mountroot prompt. If I manually put zfs:zroot at the prompt the system boots to the login screen. I checked the /etc/rc.conf and the /boot/loader.conf for syntax errors but all seems well. What on earth can be the cause of this behaviour? What do I check? Help? Have you specified a bootfs? E.g.: zpool set bootfs=zroot zroot Yes, I did. And just did it again. Please post your /boot/loader.conf. And did it again (zpool set bootfs=zroot zroot) ; rebooted and finally the system boots up. So, problem solved. Posts arfe being fetched. Thanks. If it's a timeout problem, there's kern.cam.boot_delay=1. +1 for kern.cam.boot_delay=1. I had the same problem and that setting fixed it for me. ___ 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
Squeezebox Server 7.6 failed to load: YAML::Syck
I'm running FreeBSD 8.2-RC3 and use Squeezebox Server (v7.5.1, r30836, Tue Jun 1 07:00:00 MDT 2010). Squeezebox Server 7.6 has experimental native support for UPnP media renderers (of which I have many) and I wanted to try it out. I downloaded the latest tarball and when I run slimserver.pl I get the following error: The following modules failed to load: YAML::Syck *** NOTE: If you're running some unsupported Linux/Unix platform, please use the buildme.sh script located here: http://svn.slimdevices.com/repos/slim/7.6/trunk/vendor/CPAN/ If 7.6 is outdated by the time you read this, Replace 7.6 with the major version You should never need to do this if you're on Windows or Mac OSX. If the installers don't work for you, ask for help and/or report a bug. of Squeezebox Server you are running. *** Exiting.. --- The version of Perl seems to be OK: sodserve# perl -v This is perl, v5.10.1 (*) built for amd64-freebsd When I install /usr/ports/textproc/p5-YAML-Syck it's supposedly already installed: === Installing for p5-YAML-Syck-1.17 === p5-YAML-Syck-1.17 depends on file: /usr/local/bin/perl5.10.1 - found === Generating temporary packing list === Checking if textproc/p5-YAML-Syck already installed === p5-YAML-Syck-1.17 is already installed cpan install gives me: YAML::Syck is up to date (1.17) I'm not really sure what else to check. Can someone point me in the right direction? Carl ___ 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: ROOT on ZFS with MBR partitions
How long are you waiting? What are you booting from? On Sat, Feb 26, 2011 at 7:54 PM, Slawomir Wojtczak verma...@gmx.com wrote: Anything interesting happening during your install? I would say no, everything seems smooth until I try to boot it. ___ 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 ___ 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: Bridge Interface Members
Yes. You overwrite your first ifconfig_bridge0 setting with the second one. These are shell variable initializations, not executable statements. There are various ways to fix the problem. Try this for example: replace the second ifconfig_bridge0 line with: ipv4_addrs_bridge0=10.0.1.2/24 Doh! Of course, thanks. Rookie mistake. Carl ___ 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
Bridge Interface Members
Trying to configure a bridge interface with two member interfaces, fxp0 and re0. Configuring the interface from scratch manually works fine but when I add config entries to rc.conf the two member interfaces aren't added at boot. Bridge0 is created it just doesn't have any members. From the serial console I can manually add the two member interfaces and everything is fine but obviously I'd like it to work without manual intervention. Any ideas? Here's my rc.conf entries: cloned_interfaces=bridge0 ifconfig_bridge0=addm fxp0 addm re0 ifconfig_fxp0=up ifconfig_re0=up ifconfig_bridge0=inet 10.0.1.2 netmask 255.255.255.0 up Any ideas? Troubleshooting is bit of a pain as I'm booting zfs root from a USB stick and there's a 5 minute (yes, 5 minutes!) delay at the BTX loader before the boot loader menu is displayed. I haven't figured out what's causing that but it makes tweaking and rebooting a slow process! ___ 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: Date of a FreeBSD installation
I'd suggest looking at the Btimes of top level directories stat -f %SB %N /* Or how about just / as this ~15 minutes earlier than most of the remaining top level directories sodserve# stat -f %SB %N /* Jan 9 04:54:21 2011 /COPYRIGHT Jan 9 04:54:21 2011 /bin Jan 9 04:54:21 2011 /boot Dec 31 18:59:59 1969 /dev Jan 9 04:54:21 2011 /etc Jan 9 04:54:21 2011 /lib Jan 9 04:54:21 2011 /libexec Jan 9 04:54:21 2011 /media Jan 9 04:54:21 2011 /mnt Jan 9 04:54:21 2011 /proc Jan 9 04:54:21 2011 /rescue Jan 9 04:54:21 2011 /root Jan 9 04:54:21 2011 /sbin Jan 9 04:54:21 2011 /sys Jan 9 04:48:39 2011 /tmp Jan 9 04:48:45 2011 /usr Jan 9 04:49:39 2011 /var sodserve# stat -f %SB %N / Jan 9 04:39:59 2011 / ___ 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 + GPT with root on memory stick and mirrored SATA drives
snip echo -en \n\nNow run these two commands to make the changes live, and reboot zfs set mountpoint=legacy $zpool/be/$nroot zpool set bootfs=$zpool/be/$nroot $zpool\n\n Thanks for the input krad. It would be nice to easily switch back and forth but aren't you still stuck if everything blows up on that first reboot? In order to switch back to the known working dataset you've got to get to a fixit prompt to set the correct bootfs property right? ___ 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
ZFS + GPT with root on memory stick and mirrored SATA drives
Posting the below for input. The bulk of this is from a guide that Morgan Wesström posted to this list. Some of it is taken from the root on ZFS wiki entries on freebsd.org. Some from a pjd post here: http://blogs.freebsdish.org/pjd/2010/08/06/from-sysinstall-to-zfs-only-configuration/ And then there's this that Svein Skogen posted to the list: I usually (today) set up something similar. I sysinstall FreeBSD onto a CF card with the one-big-root method, then create a zpool (on spinning-metal-storage) where I create the usr, tmp, var fs'es, tar|tar the originals over and fix the mountpoint info on the zfs'es. Then I add swap on a zvol (since I don't know how to properly use a kernel dump, I don't need swap to store it). I'm setting up a new home server and I always agonize over partitioning. So the steps below install the base system with zfs root on a usb stick and /tmp /usr /var and swap on mirrored sata drives. I've tested these steps and everything works but before I press on with actually configuring and using the server, does anybody have any input on whether I should or shouldn't do it this way? ZFS best practices suggests that having elements of the root filesystem on different pools is a bad idea. So that might be strike 1. Memory Stick / /bin /boot /dev /etc /lib /libexec /media /mnt /proc /rescue /root /sbin /sys -- /usr/src/sys Hard disk zpool --- /tmp /usr /var swap on zvol Separate zfs datasets - /tmp /usr /usr/home /usr/local /usr/obj /usr/ports /usr/ports/distfiles /usr/ports/packages /usr/src /var /var/log /var/audit /var/tmp Install Procedure (Mostly by Morgan Wesström) - Select your country and keyboard layout. Enter the Fixit environment and use the live filesystem on your DVD. Your usb memory stick will most likely be da0 but you can (and should) check it with camcontrol devlist before you continue. Create a new GPT partitioning scheme: # gpart create -s gpt da0 Create a 64KiB partition for the zfs bootcode starting at LBA 1920: # gpart add -b 1920 -s 128 -t freebsd-boot da0 Create a zfs partition spanning the remainder of the usb memory stick and give it a label we can refer to: # gpart add -t freebsd-zfs -l FreeBSDonUSB da0 (The starting LBA for the first partition is there to align the partitions to the flash memory's erase block size. This is particularly important for the main zfs partition. The main partition above will start at exactly 1MiB (LBA 2048) which will align it to any erase block size used today. This alignment is also of great importance if you use this guide to install FreeBSD to one of the newer harddrives using 4096 byte sectors.) Install the protective MBR to LBA 0 and the zfs bootcode to the first partition: # gpart bootcode -b /dist/boot/pmbr -p /dist/boot/gptzfsboot -i 1 da0 Create /boot/zfs (for zpool.cache) and load the zfs kernel modules: # mkdir /boot/zfs # kldload /dist/boot/kernel/opensolaris.ko # kldload /dist/boot/kernel/zfs.ko Create a zfs pool and set its bootfs property: # zpool create zrootusb /dev/gpt/FreeBSDonUSB # zpool set bootfs=zrootusb zrootusb Switch to fletcher4 checksums and turn off access time modifications: # zfs set checksum=fletcher4 zrootusb # zfs set atime=off zrootusb Create zfs mirrored data pool on SATA disks # zpool create zdata mirror /dev/ad4 /dev/ad6 # zfs set canmount=off zdata # zfs set mountpoint=/zrootusb zdata # zfs set checksum=fletcher4 zdata # zfs create zdata/tmp # zfs create zdata/usr # zfs create zdata/usr/home # zfs create zdata/usr/local # zfs create zdata/usr/obj # zfs create zdata/usr/ports # zfs create zdata/usr/ports/distfiles # zfs create zdata/usr/ports/packages # zfs create zdata/usr/src # zfs create zdata/var # zfs create zdata/var/log # zfs create zdata/var/audit # zfs create zdata/var/tmp Create swap zvol on zdata pool # zfs create -V 5G zdata/swap # zfs set org.freebsd:swap=on zdata/swap # zfs set checksum=off zdata/swap Extract at a minimum, base and the generic kernel: # cd /dist/8.1-RELEASE/base # DESTDIR=/zrootusb ./install.sh # cd ../kernels # DESTDIR=/zrootusb ./install.sh generic Delete the empty, default kernel directory and move the generic kernel into its place: # rmdir /zrootusb/boot/kernel # mv /zrootusb/boot/GENERIC /zrootusb/boot/kernel Make sure the zfs modules are loaded at boot: # cat /zrootusb/boot/loader.conf zfs_load=YES vfs.root.mountfrom=zfs:zrootusb kern.cam.boot_delay=1 ^d Create /etc/rc.conf. Adjust and add to your own needs: # cat /zrootusb/etc/rc.conf hostname=sodserve sshd_enable=YES zfs_enable=YES ^d Setup your time zone: # cp /zrootusb/usr/share/zoneinfo/EST5EDT /zrootusb/etc/localtime Create an empty fstab to avoid startup warnings: # touch /zrootusb/etc/fstab Set the root password in the new environment: # cd / # chroot /zrootusb /bin/sh # passwd root #
GPT/ZFS/USB mountroot prompt
I followed a gpt/zfs on USB stick guide for putting a base 8.1-RELEASE amd64 onto a 4GB sandisk USB memory stick. All went fairly well and the system will boot but fails to mount the root file system and dumps me a the mountroot prompt. Entering zfs:zrootusb at the prompt works and the system finishes booting. In /boot/loader.conf I've got: zfs_load=YES vfs.root.mountfrom=zfs:zrootusb zrootusb mountpoint is set to legacy and /etc/fstab exists but is empty, per the guide. Any ideas? Thanks, Carl ___ 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: GPT/ZFS/USB mountroot prompt
On Wed, Dec 1, 2010 at 12:07 AM, Carl Chave c...@chave.us wrote: I followed a gpt/zfs on USB stick guide for putting a base 8.1-RELEASE amd64 onto a 4GB sandisk USB memory stick. All went fairly well and the system will boot but fails to mount the root file system and dumps me a the mountroot prompt. Entering zfs:zrootusb at the prompt works and the system finishes booting. In /boot/loader.conf I've got: zfs_load=YES vfs.root.mountfrom=zfs:zrootusb zrootusb mountpoint is set to legacy and /etc/fstab exists but is empty, per the guide. Any ideas? Thanks, Carl adding kern.cam.boot_delay=1 to /boot/loader.conf seems to have fixed it. ___ 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
Bridging Gigabit and Fast Ethernet Interfaces
if_bridge(4) says: The if_bridge driver currently supports only Ethernet and Ethernet-like (e.g., 802.11) network devices, with exactly the same interface MTU size as the bridge device. Am I correct to assume then that I can bridge a gigabit interface and a fast ethernet interface and that one of the negatives of doing this is that Jumbo frames couldn't be used on the gigabit side? I've got an Atom based server with an onboard gigabit nic and only one PCI slot. The server sits physically close to my 10/100 switch that hangs off my firewall. I was thinking of putting a 10/100 nic into the single PCI slot and running that to the 10/100 switch for internet access and then running cable across the room from the gigabit interface to a gigabit switch on my workbench. Wired gigabit clients on the bench would then have the benefit of gigabit access to the server for doing backups but also still have internet access via the server's bridge interface right? Is there a reason I wouldn't want to do it this way? Thanks, Carl ___ 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: pxe LiveCD setup
Linux systems with netboot can boot off http (using wget) and it's much better. Twice as fast for the Clonezilla load, but System Rescue went from three minutes to only 18 seconds. Etherboot/gPXE is interesting also. It will boot from http. One of my grub4dos menu entries is a gPXE floppy image with the generic UNDI driver though I haven't really used it much. http://www.etherboot.org/wiki/start ___ 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: pxe LiveCD setup
With grub4dos title FreeBSD pxe keep chainloader --raw (pd)/images/freebsd/boot/pxeboot it loads pxeboot, but then: netboot: couldn't probe pxenet0 pxe_open: netif_open() failed ... can't load 'kernel' OK So it loads pxeboot, but then pxeboot can't use the pxenet0 device. Your goal in the pxeboot/nfs/livefs was to avoid having to transfer the large livefs iso? I won't be much help solving your problem above, just curious where you're going with your setup. ___ 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: pxe LiveCD setup
On Sat, Jul 3, 2010 at 3:07 PM, Warren Block wbl...@wonkity.com wrote: Is there a quick way to set up a PXE boot menu for booting into a number of ISO images? There's net/pxe, but it looks like only part of the solution. Ideally, there'd just be a minimal setup with a directory of ISO files and a built-in loader that lets the user choose which ISO to boot. ___ I've had a lot of luck with grub4dos. At work I use it to present a menu to the PXE client. I've had most success booting .iso files by having grub4dos memory map them, so having a fair amount of ram is helpful. I've used it to boot damn small linux, puppy linux, Dell diagnostic cd .iso, dban iso, spinrite .iso etc. See the grub4dos section of this forum for good info: http://www.boot-land.net/forums/index.php?showforum=66 ___ 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: pxe LiveCD setup
After a very cursory setup, it works! I took notes and will write it up in a bit. The only complaint I have so far is the speed of download via tftp. A 236M ISO took two minutes to load, or about 2M per second. A full CD takes a long, long time. Is that typical, or maybe just the poor Ethernet on this Aspire One D250? I haven't timed them so I can't say for sure. The biggest file I currently use is a windows PE .iso and it does take a bit to transfer. I'll breakout the stopwatch next week and see. My DHCP/tftp server is a Sunfire V240 with Solaris 10. I was having a horrible time with the default tftp server and switched to tftpd-hpa which helped a lot, especially with being able to remap \ to /. Floppy images work well also. I've got a Freedos boot disk floppy image with 3com's universal PXE ethernet driver for dumping and restoring ghost images. Works out of the box with every PXE client I've tried, no need to have custom boot disks with different nic drivers. One of the tweaks I did with the grldr file was to hex edit a section so it goes straight to the menu instead of cycling though the mac address variations. Then I added the company logo to the background of the menu... and ... nobody sees it but me! ___ 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: [#24529544] Re: Network sort of stops working with the em (intel) driver under load.
On Tue, Jun 1, 2010 at 3:34 PM, mark rowlands rowlands.m...@gmail.com wrote: Ticket subject: Re: Network sort of stops working with the em (intel) driver under load. Ticket number: 24529544 Ticket link: https://secure.mpcustomer.com/ticket.php?ticket=24529544 Ticket body: On Mon, May 31, 2010 at 9:30 PM, mark rowlands rowlands.m...@gmail.com wrote: Newly built and cvsupped system,with GENERIC kernel, when copying large amounts of data over a gigabit link via scp the network will hang after a couple of gig. I can then no longer login via ssh. If I leave it be, after about 12-24 hours I can then login again. (A reboot of course fixes the issue immediately...) . snip... I can now confirm that a similar copy to the vr card does not cause the same problem... so it would appear more likely that it might be an em related problem. and after replacing the em0 with a realtek gigabit card, the problem also does not occur. (the vr0 was only 100mb) So it would seem there is something hinky with at least the em (82540) card . ___ 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 ___ 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: Network sort of stops working with the em (intel) driver under load.
On Mon, May 31, 2010 at 3:30 PM, mark rowlands rowlands.m...@gmail.com wrote: Newly built and cvsupped system,with GENERIC kernel, when copying large amounts of data over a gigabit link via scp the network will hang after a couple of gig. I can then no longer login via ssh. If I leave it be, after about 12-24 hours I can then login again. (A reboot of course fixes the issue immediately...) . I've had this behavior also with Intel Pro 1000 cards using the em driver. I picked up a lot of 4 of them off ebay and never figured out if it was a driver problem or faulty hardware. I went back to FreeNAS on that same box (7.2-RELEASE-p4) and it seems ok except my Netgear gigabit switch also died so I'm currently plugged into a 100Mbps switch. ___ 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: Media streaming
Sorry, yes I'm using Fuppes on FreeNAS now. FreeNAS was using Mediatomb on previous releases which I also used. On Thu, May 27, 2010 at 3:22 PM, Dale Scott dalesc...@shaw.ca wrote: ... So I am search of a media server that can transcode from native format to MPEG2 on the fly that does not require a GUI. I'm experimenting with FreeNAS (FreeBSD? NanoBSD?), which uses Fuppes (Free uPnP Entertainment Services) for on-the-fly transcoding (I'm streaming to a plasma tv from the FreeNAS box via an Xbox 360). It runs headless, although there's a webgui for configuration (FreeNAS and Fupples). Just mentioning in case any of the technology could be applied to your situation. Dale ___ 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 ___ 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: Media streaming
As for ushare and multiple streams Carl, I played something on the xbox360 upstairs then tried the Philips streamium downstairs, it could not see the media server at all, was that what you were getting ? I will test it again tomorrow as this streamium bit of kit is quite old and fairly flakey in its operation. Thanks for testing Graeme. It's been probably two years now since I last used ushare, but, it was the easiest and most reliable of the bunch for doing what I wanted, which was just to browse my directory structure - none of the annoying automatic categorization and cluttered menus. The fatal flaw was not being able to stream to two players at once (not talking about synchronizing here, just different streams to multiple players). I can't remember exactly what happened though either the second player would error out when attempting to start a track or it just didn't see the server as you described. Maddening and I never had the opportunity to ask anyone to test it to see if it was my specific setup or what... ___ 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: Media streaming
Graeme, No answer from me unfortunately, but, possibly mediatomb? I had a question for you about ushare though - I use it to use it with my SMCWAA-G media players and it came the closest to operating like I wanted, BUT, I wasn't never able to stream audio to more than one player at a time. I never figured out if it was a ushare problem or something else. Are you able to stream to more than one device at a time? ___ 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: Wireless networking question
More info: I found the following in the output of pciconf -vl: no...@pci0:2:0:0: class=0x028000 card=0x10891a3b chip=0x002b168c rev=0x01 hdr=0x00 vendor = 'Atheros Communications Inc.' class = network From here: http://www.pcidatabase.com/vendor_details.php?id=174 0x002b is Atheros AR9285 Wireless LAN 802.11 a/b/g/n Controller ___ 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: Write attempt to file in ZFS snapshot dir causes panic
Thanks for testing it out krad. I went ahead and submitted the PR: http://www.freebsd.org/cgi/query-pr.cgi?pr=139806 ___ 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: Strange performance issues
Chris, Don't know, but, I will paste in a portion of the ZFS admin guide from SUN: Because these statistics are cumulative since boot, bandwidth might appear low if the pool is relatively idle. You can request a more accurate view of current bandwidth usage by specifying an interval. For example: # zpool iostat tank 2 Later it says about the -v option: You can use the same set of options (interval and count) when examining virtual device statistics. http://dlc.sun.com/pdf/819-5461/819-5461.pdf Page 95 ___ 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
Write attempt to file in ZFS snapshot dir causes panic
I'm new to FreeBSD. Been experimenting with 8.0-RC1 and zfs in a VM. Really haven't even installed it yet, just getting familiar with zfs usage from the fixit environment. I experienced some strange behavior and was wondering if this would warrant a bug report: 1. load zfs from bootloader prompt and then boot. 2. enter fixit environment. 3. import zpool (in this case a 2 disk vmware mirror named sodpool) 4. cd to previously created snapshot at sodpool/test/myfs/.zfs/snapshot/one 5. attempting to create a new file here results in: Fixit# echo hello hello.txt cannot create hello.txt: Read-only file system 6. That seems like the desired response. Next, attempt to modify a file that already exists in the snapshot: Fixit# echo hello test.txt panic: dirtying snapshot! I know I'm not supposed to be modifying a snapshot file, but a panic doesn't seem like the best response to this situation. I'm using DVD iso named 8.0-RC1-amd64-dvd1.iso This bug report has the same panic but the scenario is different: http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/138764 Is there someplace else I should check for existing issues? ___ 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: Serial console trouble: loader and login works, but no kernel messages
Did you try booting with the keyboard disconnected from the FreeBSD machine? Perhaps the vidconsole is favored when a keyboard is detected? On a linux box I had, I would get serial output from Grub, lose it during kernel load and then get a login once the OS was up, much like what you describe. I had to add a kernel argument to my Grub config so the kernel would output to the serial port. Did you look here: http://www.freebsd.org/doc/en/articles/console-server/freebsd.html I think 7.2 might be what you are missing but I can't check it myself. On Sun, Aug 23, 2009 at 3:13 PM, Thomas Backmanseren...@exscape.org wrote: On Aug 23, 2009, at 20:25, Tim Judd wrote: On 8/23/09, Thomas Backman seren...@exscape.org wrote: First off: Not subscribed to this list, please make sure to Cc me if you don't reply directly. :) Anyway, I finally got my null modem cable, and plugged in in between a machine running 8.0-BETA2 and one running WinXP using Hyperterminal. My settings: /boot/loader.conf: boot_multicons=YES boot_serial=YES comconsole_speed=115200 console=comconsole,vidconsole /etc/ttys: # Serial terminals # The 'dialup' keyword identifies dialin lines to login, fingerd etc. ttyu0 /usr/libexec/getty std.115200 vt100 on secure /boot.config (which is read properly): -Dh -S115200 Anything wrong in the above? Hyperterminal is set to 115200 bps, 8 bits, no parity, 1 stop bit, and no flow control (if that's the correct translation to English). On the serial console, I go from the screen with the FreeBSD logo, with single-user options etc. (which works fine), and then nothing, until a login tty pops up (which also works fine). The main, if not only, reason I want a serial console is to be able to use it for single user mode, DDB, and so on. All kernel messages, and all rc messages are seen only on the graphics card; the serial console receives nothing but the /boot.config: - Dh ..., the logo screen, and then the login screen, during startup and *nothing* at all during shutdown. Also, I'm able to login and use the system both via the serial console and via the graphics card/ keyboard... Is this supposed to be? I'm not complaining, I just got the impression it was one or the other. Any advice on how to get the kernel/rc messages etc. to the serial console (only or as well)? Regards, Thomas Do you use the VGA/vidconsole at all? A serial-only device (think soekris, ALIX/WRAP boards) that has no VGA will have different requirements than a serial-only device will. Your loader.conf statements are different than mine in the definition that you have more than I do to enable serial. My loader.conf just has one statement: console=comconsole - to feed ALL bootloaders, kernel probing, rc startup on the serial device. /etc/ttys defines the login lines. Though trial and error, I found when you use a dual-setup: comconsole,vidconsole, the first one (comconsole) will get rc output, and vidconsole won't. Of course, you're on 8.0 and I don't run BETAs. So the 8.0 BETA might still be having com port oddities, plus I noticed your ttys line is ttyu0, not ttyd0. Did 8.0 change the serial line device? To enable a serial-only device in my setups: /boot/loader.conf: console=comconsole /boot.config: -D /etc/ttys: # enable serial line, cons25 or vt100, depending if I'm originating from a bsd or windows box. Enabling dual-setups should be just the loader.conf change to dual console. HTH (Sorry for the lack of inline replies.) I do have a graphics card, and ideally I'd like to be able to use both, but serial has higher priority (with serial access, I can use minicom on another *nix box and essentially ssh into DDB, and stuff like that - right now I have to borrow a monitor, and write info down manually if needed, turning my head back and forth). I've tried lots of combinations of console=, including simply 'console=comconsole' and/or combinations of that and -D, -h- -Dh and -P in /boot.config. The extra lines in loader.conf are from the handbook, which says they're needed to use comconsole_speed. It seems they do the same thing as -D and -h, though. Oh, and re: /etc/ttys: Yup, it's ttyuX when using uart(4) which seems to be the default now. Actually, since my last buildworld half an hour ago I'm on 9.0-CURRENT. ;) Also, I made sure to set flags to 0x10 for the serial port as per the handbook (although I did it using loader.conf, not the kernel config); before the change, dmesg didn't mention any flags, but it now does. Didn't help squat, though. Though trial and error, I found when you use a dual-setup: comconsole,vidconsole, the first one (comconsole) will get rc output, and vidconsole won't. This doesn't mirror my experience; comconsole and comconsole,vidconsole appears to be just the same for me. I've never gotten anything except the boot loader and a login prompt over to the serial line - at least not at
Re: 7.1 System Crashing
Could also be a PSU going out - unable to maintain the correct voltages. On Tue, Apr 14, 2009 at 3:23 PM, APseudoUtopia apseudouto...@gmail.com wrote: My FreeBSD 7.1-RELEASE-p3 system has gone down a few times in the last two days. I'm trying to figure out why, but there is nothing in /var/log/messages or dmesg about the incident. Here's the output from the `last` command: User ttyp0 1.2.3.4 Tue Apr 14 19:02 still logged in User ttyp0 1.2.3.4 Tue Apr 14 18:52 - 19:02 (00:09) reboot ~ Tue Apr 14 18:52 User ttyp0 1.2.3.4 Tue Apr 14 18:42 - crash (00:09) User ttyp0 1.2.3.4 Tue Apr 14 16:06 - 18:42 (02:36) User ttyp0 1.2.3.4 Tue Apr 14 11:23 - 12:53 (01:29) reboot ~ Tue Apr 14 07:44 User ttyp0 1.2.3.4 Mon Apr 13 20:01 - 22:58 (02:57) User ttyp0 1.2.3.4 Mon Apr 13 19:56 - 20:01 (00:04) User ttyp0 1.2.3.4 Mon Apr 13 19:31 - 19:56 (00:25) User ttyp0 1.2.3.4 Sun Apr 12 15:02 - 16:15 (01:12) User ttyp0 1.2.3.4 Sun Apr 12 14:48 - 15:02 (00:14) reboot ~ Sun Apr 12 14:48 shutdown ~ Sun Apr 12 14:45 User ttyp0 1.2.3.4 Sun Apr 12 14:24 - shutdown (00:21) User ttyp0 1.2.3.4 Sat Apr 11 13:03 - 17:00 (03:56) wtmp begins Wed Apr 1 21:21:28 UTC 2009 As you can see, the system reboot several times, and crashed the most recent time. Only one of the shutdown/reboots listed is valid - the furthest one. I have no idea why it is rebooting/crashing. It has been completely stable until about 1-2 days ago. If anyone can direct me to post the output to any commands or any log messages, I'd be more than willing to do so in order to help anyone diagnose my problem. Thank you for your time. ___ 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 ___ 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: from very early this morning...
Hi Gary, Just a couple of thoughts, as your setup sounds similar to mine (and a lot of others' I'm sure) - I too recently decided to make a concerted effort to reduce power consumption. I just re-did my file server with FreeNAS and even though I've got tons of hardware laying around I decided to buy the Intel 945GCLF mini-itx board based on the Atom processor, like you find in most netbooks. I put a gigabit NIC in it though as the onboard is 10/100 (but I knew that and already had the NIC). It's running great so far. I'd like to replace my pfSense router/firewall, which is currently powered by an AMD Duron with another mini-itx board that I've had forever, it's one of the Via C3 500 Mhz based boards. It's only got one PCI slot though, which gets me back to the topic at hand. I just changed my network topology when I stood up the new file server. It's now: |-- Wired LAN ADSL Modem -- pfSense | |-- WAP -- WLAN | |-- DMZ (web server) Forgive my artwork. I have my ADSL modem set to bridged ethernet mode which disables all the router/firewall/dhcp features of the modem and just turns it into a media/protocol converter between the phone line and the ethernet cable going to the pfSense box. I use the onboard 10/100 NIC for that PPPoE connection. I've got three more NICs installed to make up the remaining connections. The wired LAN and the WLAN interfaces are bridged. I initially had these as separate networks but most of my media players are wireless and the file server is on the wired side so bridging it was the easiest way (for me!) to get the broadcasts through. The web server is connected directly to the third NIC at the moment and is it's own network. It's still behind the firewall but I can open ports now to it while still protecting the rest of the LAN from the web server if it get's compromised. At least, that's the theory. So that's my setup, don't know if that's the kind of feedback you're looking for but I'd like to hear comments and see what others have going. Carl On Tue, Apr 14, 2009 at 7:23 PM, Steve Bertrand st...@ibctech.ca wrote: Gary Kline wrote: [...big snip...] if i've made any sense so far, great! if not,i'm open for questions. i'm also open for suggestions on how to alter this network configuration. thanks for reading this far. gary It might be simplest to replace my firewall and my server with low-energy-usage i386 computers; is there a better way? What are your requirements for your network ie. are you requiring any fancy trickery, or is this simply trying to NAT a couple of machines behind an ADSL connection? Steve ___ 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 ___ 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: Monitoring geom
From Michael Lucas' Absolute FreeBSD book page 550: FreeBSD can include a status check of your mirrored disks in its daily periodic(8) run. Just add the line daily_status_gmirror_enable=YES to /etc/periodic.conf. Not sure about other raid types beyond mirrors. On 3/5/09, Mister Olli mister.o...@googlemail.com wrote: Hi hi... What is the best way to monitor geom software raids (gmirror gvinum raid5)??? The solution I'm searching for should be a kind of script which checks the status, and drops me an email if something is wrong. I found a nagios plugin, but currently I'm unable to invest enough time to get nagios up and running for the customer... Thanks a lot... greetz Olli ___ 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 ___ 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