Re: OpenBSD 4.4-release installation hangs on large disk (x86)
On 1/12/2008, at 9:24 AM, Chris wrote:
2) in an earlier message you indicated that there was some kind of
RAID on this system, I think it is safe to say that it is a BIOS-
assisted
software RAID, which COULD be causing you problems if it is still
configured in the BIOS. And even if it isn't causing this problem,
it WILL bite you in the future when the BIOS decides top copy
something
from one drive to the other...
Sorry about my previous post but there is no RAID.
So, start by disabling the BIOS "RAID" do-hickey thing and unplug
the second drive. I suspect you will have no problems then.
I have tried pulling the plug on the second disk (wd1) with no luck.
Once you convince yourself the size of the drive is not directly
an issue, I'd suggest starting over with a more sane partitioning
plan. Just because you have a cheap 500G disk doesn't mean you
need to allocate all or most of it. For one, the bigger the disk,
the longer it takes to fsck after you trip over the power cord.
I actually have two "cheap" 500GB; so that makes it 1 terabyte. I
don't mind fsck doing its thing when I trip over the cable. But I
really do need a large /data1 partition to rsync stuff over from other
places. Here's the last disk partition I tried (and it didn't work): /
10g, swap 3g, /tmp 20g, /home 100g, /var/ 50g, /usr/ 50g. It was
sitting there for 30 minutes after the last file set was installed
(xserv44.tgz) and then I turned it off.
Thanks for any further help.
I've got a couple of Dell SC440s with the same sort of set-up, and
no problems here.
Two 500Gb drives, wd0 and wd1:
# df -h
Filesystem SizeUsed Avail Capacity Mounted on
/dev/wd0a 84.4M 36.6M 43.6M46%/
/dev/wd0h 1008M2.7M955M 0%/home
/dev/wd0i 436G121G294G29%/public
/dev/wd1a 458G116G320G27%/public2
/dev/wd0d 1008M2.0K958M 0%/tmp
/dev/wd0g 9.8G1.9G7.4G21%/usr
/dev/wd0e 9.8G 31.2M9.3G 0%/var
I wanted one large dumping ground on wd0, and another on wd1,
and had no trouble creating them (I cannot recall any lengthy delay;
not the same as saying there wasn't one, but I don't remember it.)
And yes, fsck does take a long time.
Thanks.
dmesg (sd0 is an 8Gb USB memstick):
OpenBSD 4.4 (GENERIC) #0: Tue Nov 11 10:18:14 NZDT 2008
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel(R) Pentium(R) D CPU 3.00GHz ("GenuineIntel" 686-class) 3 GHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,
CFLUSH,DS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,SBF,SSE3,MWAIT,DS-
CPL,EST,CNXT-ID,CX16,xTPR
real mem = 1071722496 (1022MB)
avail mem = 1027870720 (980MB)
mainbus0 at root
bios0 at mainbus0: AT/286+ BIOS, date 07/03/07, BIOS32 rev. 0 @
0xffe90, SMBIOS rev. 2.3 @ 0xf0450 (63 entries)
bios0: vendor Dell Inc. version "1.4.1" date 07/03/2007
bios0: Dell Inc. PowerEdge SC440
apm0 at bios0: Power Management spec V1.2
apm0: AC on, battery charge unknown
acpi at bios0 function 0x0 not configured
pcibios0 at bios0: rev 2.1 @ 0xf/0x1
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xfed10/256 (14 entries)
pcibios0: PCI Interrupt Router at 000:31:0 ("Intel 82801GH LPC" rev
0x00)
pcibios0: PCI bus #5 is the last bus
bios0: ROM list: 0xc/0x9000 0xc9000/0x2000! 0xcb000/0x1000
cpu0 at mainbus0
cpu0: Enhanced SpeedStep disabled by BIOS
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel E7230 Host" rev 0x00
ppb0 at pci0 dev 1 function 0 "Intel E7230 PCIE" rev 0x00: irq 11
pci1 at ppb0 bus 1
ppb1 at pci0 dev 28 function 0 "Intel 82801GB PCIE" rev 0x01: irq 11
pci2 at ppb1 bus 2
ppb2 at pci0 dev 28 function 4 "Intel 82801G PCIE" rev 0x01: irq 11
pci3 at ppb2 bus 3
em0 at pci3 dev 0 function 0 "Intel PRO/1000 PT (82572EI)" rev 0x06:
irq 11, address 00:15:17:3d:38:11
ppb3 at pci0 dev 28 function 5 "Intel 82801G PCIE" rev 0x01: irq 10
pci4 at ppb3 bus 4
bge0 at pci4 dev 0 function 0 "Broadcom BCM5754" rev 0x02,
BCM5754/5787 A2 (0xb002): irq 10, address 00:1d:09:09:94:17
brgphy0 at bge0 phy 1: BCM5787 10/100/1000baseT PHY, rev. 0
uhci0 at pci0 dev 29 function 0 "Intel 82801GB USB" rev 0x01: irq 9
uhci1 at pci0 dev 29 function 1 "Intel 82801GB USB" rev 0x01: irq 5
uhci2 at pci0 dev 29 function 2 "Intel 82801GB USB" rev 0x01: irq 3
uhci3 at pci0 dev 29 function 3 "Intel 82801GB USB" rev 0x01: irq 10
ehci0 at pci0 dev 29 function 7 "Intel 82801GB USB" rev 0x01: irq 9
usb0 at ehci0: USB revision 2.0
uhub0 at usb0 "Intel EHCI root hub" rev 2.00/1.00 addr 1
ppb4 at pci0 dev 30 function 0 "Intel 82801BA Hub-to-PCI" rev 0xe1
pci5 at ppb4 bus 5
vga1 at pci5 dev 7 function 0 "ATI ES1000" rev 0x02
wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
drm at vga1 unsupported
ichpcib0 at pci0 dev 31 function 0 "Intel 82801GB LPC" rev 0x01: PM
disabled
pciide0 at pci0 dev 31 function 1 "I
Re: OpenBSD 4.4-release installation hangs on large disk (x86)
> Just because you have a cheap 500G disk doesn't mean you > need to allocate all or most of it. For one, the bigger the disk, > the longer it takes to fsck after you trip over the power cord. Wait for fsck? So OpenBSD doesn't have background fsck? :-(
cannot nfs mount an mfs
I'm trying to export an mfs filesystem, but it seems not to work. fstab reads /dev/wd0b /tmp mfs rw,nosuid,-s=524288 0 0 exports reads /tmp-maproot=root -alldirs clientname The client can mount but any I/O results in # mkdir /tmp/shared # mount -t nfs 192.168.1.10:/tmp /tmp/shared # ls /tmp/shared ls: /tmp/shared: Input/output error If I try to export a subdirectory, and mount that, I get a file protection error on the mount attempt. Is this possible, or is there simply no solution? --John
Re: bioctl and RAID0
Softraid will not print anything. It will mark a disk offline and if
the discipline does not support redundancy it will mark the volume
offline as well. bioctl will tell you what is going on.
On Sun, Nov 30, 2008 at 01:26:24PM -0800, Manuel Ravasio wrote:
> Another bioctl related question, right out of curiosity.
>
> What happens when one or more disks in a RAID fail?
> I mean, I suppose some kind of error messages will be logged and/or sent to
> console.
> I also imagine bioctl softraid? will show useful messages.
>
> Can anyone point me to some documentation explaining the various possible
> error messages and their meaning?
>
> Thank you all once more.
> Manuel
>
>
>
> --
> "On two occasions I have been asked [by members of Parliament!], `Pray,
> Mr. Babbage, if you put into the machine wrong figures, will the right
> answers come out?' I am not able rightly to apprehend the kind of
> confusion of ideas that could provoke such a question."
> -- Charles Babbage
>
>
>
> - Original Message
> > From: Marco Peereboom <[EMAIL PROTECTED]>
> > To: Manuel Ravasio <[EMAIL PROTECTED]>
> > Cc: openbsd
> > Sent: Saturday, November 29, 2008 11:05:00 PM
> > Subject: Re: bioctl and RAID0
> >
> > Bah that is a bug though; the disk should not be knocked offline for an
> > out of bounds read/write. I'll fix this.
> >
> > Thanks for the report.
> >
> > Your mistake is not to fdisk and disklabel the brand new disk that you
> > created. See softraid(4) for examples.
> >
> > On Sat, Nov 29, 2008 at 04:34:52AM -0800, Manuel Ravasio wrote:
> > > Hello list.
> > > i386 PC with 3 PATA disks.
> > > - a 60g Maxtor attached to motherboard's IDE controller
> > > - two 160g Maxtor attached to a Promise FastTrak TX2 PCI controller
> > >
> > > During install all 3 disks are correctly recognized and fully assigned to
> > OpenBSD.
> > > Both 160g disks have a single partition (a) spanning all disk length
> > > (from
> > sector 63), RAID type.
> > >
> > > When I run
> > > # bioctl -c 0 -l /dev/wd1a,/dev/wd2a softraid0
> > > the device is (AFAICT) correctly created:
> > >
> > > # bioctl softraid0
> > > Volume Status Size Device
> > > softraid0 0 Online 327843063808 sd0 RAID0
> > > 0 Online 163921531904 0:0.0 noencl
> > > 1 Online 163921531904 0:1.0 noencl
> > > #
> > >
> > > On device sd0 there is a single 4.2BSD partition spanning all the disk;
> > > that's
> > ok with me.
> > >
> > > # disklabel sd0
> > > # /dev/rsd0c:
> > > type: SCSI
> > > disk: SCSI disk
> > > label: SR RAID 0
> > > flags:
> > > bytes/sector: 512
> > > sectors/track: 63
> > > tracks/cylinder: 255
> > > sectors/cylinder: 16065
> > > cylinders: 39857
> > > total sectors: 640318485
> > > rpm: 3600
> > > interleave: 1
> > > trackskew: 0
> > > cylinderskew: 0
> > > headswitch: 0 # microseconds
> > > track-to-track seek: 0 # microseconds
> > > drivedata: 0
> > >
> > > 16 partitions:
> > > #size offset fstype [fsize bsize cpg]
> > > a:6403184850 4.2BSD 2048 163841
> > > c:6403184850 unused 0 0
> > > #
> > >
> > > Then I try to create a new filesystem on the partition and I receive an
> > > error
> > message:
> > >
> > > # newfs /dev/rsd0a
> > > newfs: wtfs: write error on block 640318484: Input/output error
> > > #
> > >
> > > Now bioctl shows one disk as "Offline".
> > >
> > > # bioctl softraid0
> > > Volume Status Size Device
> > > softraid0 0 Offline 327843063808 sd0 RAID0
> > > 0 Offline 163921531904 0:0.0 noencl
> > > 1 Online 163921531904 0:1.0 noencl
> > > #
> > >
> > > What does this mean?
> > > Why is the disk offline?
> > > What am I doing wrong?
> > >
> > > By the way, when I try to delete both sd0 and softraid0 the machine goes
> > > to a
> > "ddb>" prompt.
> > >
> > >
> > > Thank you all,
> > > bye,
> > > Manuel
> > >
> > >
> > > PS: During installation and during boot I received a few "interface CRC
> > > error"
> > messages from wd0.
> > > What des this mean exactly?
> > > The disk itself is quite old, 6 years at the very least. Do these
> > > messages
> > mean it is going to die soon?
> > >
> > > Thanks again,
> > > M.
> > >
> > > ==
> > > dmesg right after install
> > > ==
> > >
> > > OpenBSD 4.4 (GENERIC) #1021: Tue Aug 12 17:16:55 MDT 2008
> > >[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
> > > cpu0: AMD-K7(tm) Processor ("AuthenticAMD" 686-class, 512KB L2 cache) 550
> > > MHz
> > > cpu0: FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,MMX
> > > real mem = 267939840 (255MB)
> > > avail mem = 250646528 (239MB)
> > > mainbus0 at root
> > > bios0 at mainbus0: AT/286+ BIOS, date 08/02/00, BIOS32 rev. 0 @ 0xfd9c0,
> > SMBIOS rev. 2.3 @ 0xf04f0 (3
Re: bioctl and RAID0 -- RAID1+0 and 0+1?
This will work although I have not written all the magic to make it pretty. I will at some point make this into an actual raid type so that it is a single create statement instead of several. I do not recommend using it this way. On Sun, Nov 30, 2008 at 01:22:55PM -0800, Manuel Ravasio wrote: > I read that softraid now supports RAID0 and RAID1 only. > I'm thinking of adding two more disks to the i386 pc I wrote about in this > thread. > Would a RAID 1+0 or 0+1 supported in this case? > > I can think of a procedure like this: > - fdisk and disklabel all 4 disks with a single RAID partition > - create a RAID0 sd0/softraid0 device on, say, wd1 and wd2 > - create a RAID0 sd1/softraid1 device on wd3 and wd4 > - disklabel sd0 creating a single RAID partition > - idem with sd1 > - create a RAID1 sd2/softraid2 device on sd0 and sd1 > - disklabel sd2 with a single 4.2BSD partition > - create filesystem, mount, etc etc > > Would something like this work? > Unfortunately I have no suitable hardware available right now. > I'll try tampering with a OpenBSD 4.4 virtual machine as soon as I have a > couple of spare hours. > > Manuel > > > -- > "On two occasions I have been asked [by members of Parliament!], `Pray, > Mr. Babbage, if you put into the machine wrong figures, will the right > answers come out?' I am not able rightly to apprehend the kind of > confusion of ideas that could provoke such a question." > -- Charles Babbage
Re: Samba printing, OpenBSD client to Windows server
Ed Ahlsen-Girard wrote: OK, I've installed Samba, and gotten printcap set such that I printed a straight text fire, but nothing else works now that I tried to print other formats through gv and open-office. Perhaps Samba is not the way to go? Printcap below. #$OpenBSD: printcap,v 1.4 2003/03/28 21:32:30 jmc Exp $ #lp|local line printer:\ #:lp=/dev/lp:sd=/var/spool/output:lf=/var/log/lpd-errs: #rp|remote line printer:\ #:lp=:rm=printhost:rp=lp:sd=/var/spool/output:lf=/var/log/lpd-errs: lp|hpoffice:rp=hpoffice:rm=192.168.1.100:sd=/var/spool/lpd/hpoffice:af=/var/spool/lpd/hpoffice/acct:if=/usr/local/bin/smbprint:mx=0:lp=/dev/null: Personally, I've found it easier to just install an LPR daemon on Windows itself. I had a similar problem of trying to print to a printer connected to a Windows box, and I have yet to find any solution with, eg Samba or otherwise, to print to it over SMB. At least on Windows XP, there is a "Add/Remove Windows Components" in the control panel that allows you to install an LPR daemon (which is not installed by default), and pretty much anything can print to it after that. Maybe the situation will work better if you run Windows 2003 or 2008, but I'm not familiar with that case.
Re: (open)smtpd, the mystery smtpd daemon
2008/11/11 Marc Balmer <[EMAIL PROTECTED]>: > * Mathias Reitinger wrote: >> On 22:46 10 Nov 08, Jason Beaudoin wrote: >> > oh.. is the stuffed puffy (seen in your photos) available for >> > purchase? I threw out my stress-tux, but my speaker needs a >> > replacement toy :P >> >> you can order the Pluffy from wim at >> >> http://www.kd85.com/notforsale.html > > and here: http://shop.msys.ch/product_info.php?cPath=29&products_id=43 > (shipping cost valid in .ch only, all others please inquire) Pluffy appears to no longer be listed at https://https.openbsd.org/cgi-bin/order.eu (it was listed there a year ago). I wonder whether this is intentional? Thanks and regards, --ropers
Re: halt -p does not power off ThinkPad X61 under 4.4
> used halt -p to shutdown the machine and walked away. The next morning i > found that while it appears to have shutdown correctly, the machine did > not power off but instead showed > > syncing disks...done > uchi2: host controller halted This bug has been supposedly fixed in OpenBSD -current. See http://marc.info/?l=openbsd-bugs&m=122788235218793
Re: bioctl and RAID0
Another bioctl related question, right out of curiosity.
What happens when one or more disks in a RAID fail?
I mean, I suppose some kind of error messages will be logged and/or sent to
console.
I also imagine bioctl softraid? will show useful messages.
Can anyone point me to some documentation explaining the various possible error
messages and their meaning?
Thank you all once more.
Manuel
--
"On two occasions I have been asked [by members of Parliament!], `Pray,
Mr. Babbage, if you put into the machine wrong figures, will the right
answers come out?' I am not able rightly to apprehend the kind of
confusion of ideas that could provoke such a question."
-- Charles Babbage
- Original Message
> From: Marco Peereboom <[EMAIL PROTECTED]>
> To: Manuel Ravasio <[EMAIL PROTECTED]>
> Cc: openbsd
> Sent: Saturday, November 29, 2008 11:05:00 PM
> Subject: Re: bioctl and RAID0
>
> Bah that is a bug though; the disk should not be knocked offline for an
> out of bounds read/write. I'll fix this.
>
> Thanks for the report.
>
> Your mistake is not to fdisk and disklabel the brand new disk that you
> created. See softraid(4) for examples.
>
> On Sat, Nov 29, 2008 at 04:34:52AM -0800, Manuel Ravasio wrote:
> > Hello list.
> > i386 PC with 3 PATA disks.
> > - a 60g Maxtor attached to motherboard's IDE controller
> > - two 160g Maxtor attached to a Promise FastTrak TX2 PCI controller
> >
> > During install all 3 disks are correctly recognized and fully assigned to
> OpenBSD.
> > Both 160g disks have a single partition (a) spanning all disk length (from
> sector 63), RAID type.
> >
> > When I run
> > # bioctl -c 0 -l /dev/wd1a,/dev/wd2a softraid0
> > the device is (AFAICT) correctly created:
> >
> > # bioctl softraid0
> > Volume Status Size Device
> > softraid0 0 Online 327843063808 sd0 RAID0
> > 0 Online 163921531904 0:0.0 noencl
> > 1 Online 163921531904 0:1.0 noencl
> > #
> >
> > On device sd0 there is a single 4.2BSD partition spanning all the disk;
> > that's
> ok with me.
> >
> > # disklabel sd0
> > # /dev/rsd0c:
> > type: SCSI
> > disk: SCSI disk
> > label: SR RAID 0
> > flags:
> > bytes/sector: 512
> > sectors/track: 63
> > tracks/cylinder: 255
> > sectors/cylinder: 16065
> > cylinders: 39857
> > total sectors: 640318485
> > rpm: 3600
> > interleave: 1
> > trackskew: 0
> > cylinderskew: 0
> > headswitch: 0 # microseconds
> > track-to-track seek: 0 # microseconds
> > drivedata: 0
> >
> > 16 partitions:
> > #size offset fstype [fsize bsize cpg]
> > a:6403184850 4.2BSD 2048 163841
> > c:6403184850 unused 0 0
> > #
> >
> > Then I try to create a new filesystem on the partition and I receive an
> > error
> message:
> >
> > # newfs /dev/rsd0a
> > newfs: wtfs: write error on block 640318484: Input/output error
> > #
> >
> > Now bioctl shows one disk as "Offline".
> >
> > # bioctl softraid0
> > Volume Status Size Device
> > softraid0 0 Offline 327843063808 sd0 RAID0
> > 0 Offline 163921531904 0:0.0 noencl
> > 1 Online 163921531904 0:1.0 noencl
> > #
> >
> > What does this mean?
> > Why is the disk offline?
> > What am I doing wrong?
> >
> > By the way, when I try to delete both sd0 and softraid0 the machine goes to
> > a
> "ddb>" prompt.
> >
> >
> > Thank you all,
> > bye,
> > Manuel
> >
> >
> > PS: During installation and during boot I received a few "interface CRC
> > error"
> messages from wd0.
> > What des this mean exactly?
> > The disk itself is quite old, 6 years at the very least. Do these messages
> mean it is going to die soon?
> >
> > Thanks again,
> > M.
> >
> > ==
> > dmesg right after install
> > ==
> >
> > OpenBSD 4.4 (GENERIC) #1021: Tue Aug 12 17:16:55 MDT 2008
> >[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
> > cpu0: AMD-K7(tm) Processor ("AuthenticAMD" 686-class, 512KB L2 cache) 550
> > MHz
> > cpu0: FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,MMX
> > real mem = 267939840 (255MB)
> > avail mem = 250646528 (239MB)
> > mainbus0 at root
> > bios0 at mainbus0: AT/286+ BIOS, date 08/02/00, BIOS32 rev. 0 @ 0xfd9c0,
> SMBIOS rev. 2.3 @ 0xf04f0 (30 entries)
> > bios0: vendor American Megatrends Inc. version "627.10" date 02/29/2000
> > bios0: ASUSteK Computer INC. K7M
> > apm0 at bios0: Power Management spec V1.2
> > apm0: AC on, battery charge unknown
> > acpi at bios0 function 0x0 not configured
> > pcibios0 at bios0: rev 2.1 @ 0xf/0x1
> > pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf8120/144 (7 entries)
> > pcibios0: PCI Interrupt Router at 000:04:0 ("VIA VT82C586 ISA" rev 0x00)
> > pcibios0: PCI bus #1 is the last bus
> > bios0: ROM list: 0xc/0x8000 0xc8000/0xc000
> > cpu0 at mainbus0
Re: bioctl and RAID0 -- RAID1+0 and 0+1?
I read that softraid now supports RAID0 and RAID1 only. I'm thinking of adding two more disks to the i386 pc I wrote about in this thread. Would a RAID 1+0 or 0+1 supported in this case? I can think of a procedure like this: - fdisk and disklabel all 4 disks with a single RAID partition - create a RAID0 sd0/softraid0 device on, say, wd1 and wd2 - create a RAID0 sd1/softraid1 device on wd3 and wd4 - disklabel sd0 creating a single RAID partition - idem with sd1 - create a RAID1 sd2/softraid2 device on sd0 and sd1 - disklabel sd2 with a single 4.2BSD partition - create filesystem, mount, etc etc Would something like this work? Unfortunately I have no suitable hardware available right now. I'll try tampering with a OpenBSD 4.4 virtual machine as soon as I have a couple of spare hours. Manuel -- "On two occasions I have been asked [by members of Parliament!], `Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question." -- Charles Babbage
Re: dhclient regression? 4.3 -> 4.4
On Sun, 2008-11-30 at 15:23 -0500, System Administrator wrote: > The two configuration files are shown below. The only significant > difference is in how the alternate location uses a non-zero key index. > > working hostname.rum0 (in primary location): > dhcp nwid "HOME" nwkey "HomeWEPString" > > no-longer working hostname.rum0: > dhcp nwid "WIFI" nwkey "2:,SomeKeyString,," > I had the same behavior. For n:key1,key2,key3,key4 if n=1 it always worked for me. Feature ;) http://marc.info/?l=openbsd-misc&m=121811679523684&w=2
Re: OpenBSD 4.4-release installation hangs on large disk (x86)
> 2) in an earlier message you indicated that there was some kind of > RAID on this system, I think it is safe to say that it is a BIOS-assisted > software RAID, which COULD be causing you problems if it is still > configured in the BIOS. And even if it isn't causing this problem, > it WILL bite you in the future when the BIOS decides top copy something > from one drive to the other... Sorry about my previous post but there is no RAID. > So, start by disabling the BIOS "RAID" do-hickey thing and unplug > the second drive. I suspect you will have no problems then. I have tried pulling the plug on the second disk (wd1) with no luck. > Once you convince yourself the size of the drive is not directly > an issue, I'd suggest starting over with a more sane partitioning > plan. Just because you have a cheap 500G disk doesn't mean you > need to allocate all or most of it. For one, the bigger the disk, > the longer it takes to fsck after you trip over the power cord. I actually have two "cheap" 500GB; so that makes it 1 terabyte. I don't mind fsck doing its thing when I trip over the cable. But I really do need a large /data1 partition to rsync stuff over from other places. Here's the last disk partition I tried (and it didn't work): / 10g, swap 3g, /tmp 20g, /home 100g, /var/ 50g, /usr/ 50g. It was sitting there for 30 minutes after the last file set was installed (xserv44.tgz) and then I turned it off. Thanks for any further help.
dhclient regression? 4.3 -> 4.4
I have an i386 box that used to be running 4.3-stable and was recently upgraded to 4.4 using a CD and following the instructions. Everything seemed to be working fine including rum wireless in its primary location. However, a previously working configuration in an alternate location now results in the following log entries: DHCPDISCOVER on rum0 to 255.255.255.255 port 67 interval 1 DHCPDISCOVER on rum0 to 255.255.255.255 port 67 interval 2 ... No DHCPOFFERS received. No working leases in persistent database - sleeping. The two configuration files are shown below. The only significant difference is in how the alternate location uses a non-zero key index. working hostname.rum0 (in primary location): dhcp nwid "HOME" nwkey "HomeWEPString" no-longer working hostname.rum0: dhcp nwid "WIFI" nwkey "2:,SomeKeyString,," The box is a P-III class running GENERIC kernel. I did not include a dmesg because currently it lacks connectivity (the very reason for the posting) and the above information was hand-typed. However, if it does become really important, I will find a way to transfer the dmesg and/or output of any other command as requested. All input greatly appreciated, -Jacob.
Re: Missing security announcements
On Sun, Nov 30, 2008 at 10:23:56AM -0800, new_guy wrote: > Martin SchrC6der wrote: > > > > Why do you maintain stable by issuing security patches for it if you > > don't care if anybody installs them (by not telling them about the > > patches through one of the designated channels)? Don't you want > > people installing them? > > > > Is it so hard to write a mail to the list once every few months? The > > content is already there... > > > > I just check the errata web page every now and then. When/if anything huge > is discovered (very seldom) then it's slashdotted or something. So in the > end, I always seem to find out somehow. If someone is following stable, and really cares about keeping their system(s) up to date, I can't imagine why they wouldn't take the few seconds per day required to glance at the errata page. I mean, if you're reading Slashdot, The Guardian, Al-Jazeera, The Onion, or what-have-you, on a regular basis, why not just toss the errata page into the mix? For Christ's sake, the errata are listed in reverse chronological order so you don't even have to hit the space bar to see what's new. Do they have to toss in a soother as well? Not to mention that checking the errata page daily only underlines the extent to which these people---who give away for free a complete operating system---are really on top of the game. cheers, -wb
Re: Missing security announcements
Martin SchrC6der wrote: > > Why do you maintain stable by issuing security patches for it if you > don't care if anybody installs them (by not telling them about the > patches through one of the designated channels)? Don't you want > people installing them? > > Is it so hard to write a mail to the list once every few months? The > content is already there... > I just check the errata web page every now and then. When/if anything huge is discovered (very seldom) then it's slashdotted or something. So in the end, I always seem to find out somehow. -- View this message in context: http://www.nabble.com/Missing-security-announcements-tp20465932p20760480.html Sent from the openbsd user - misc mailing list archive at Nabble.com.
Re: bash for root?
farhan ahmed wrote: > Question is how can you make shell statically linked? I thought when you > install package it should be linked rather than manual compiling and > installing I think that is best left as an exercise for the asker. Here's what it boils down to: There is nothing wrong with a properly implemented 'bash' or any other shell for root. Hint: when the system comes up single user mode, it will ASK you what shell to use. The statically compiled part isn't even critical in OpenBSD, unless you are intent on running bash in single-user mode before all partitions are mounted. The problem is when you break things, you break 'em BIG. Original thread is a case in point. You win awards for courage, not wisdom, for still being intent on using bash as the root shell while you are still walking with a limp from your last experience. There's a lot of stuff that can go wrong when changing a user's default shell over the lifecycles of the system (think upgrades!), virtually all operator error, all avoidable, but errors that can happen tend to happen. When you break JoeAverage's account, no big deal, as long as you can get back as root and fix it. When you break root, you have a problem. Yes, the goal is to do everything right, but another goal is to make it more difficult to do things wrong. If you don't know how to do it right, test it right, and recover it right, don't change the root shell. I realize how it is such finger breaking work to type the five keystrokes "b a s h [enter]" at a command prompt after logging in...so horrible, I know, but until you know what you are doing, just manually invoke bash. You will know you know what you are doing when you realize you don't need or want to use bash on OpenBSD. The only good reason I've found to use bash on OpenBSD is to make it feel like some other OS, and that's really not a good thing when you are administering the system (i.e., logging in as root!). ksh rocks on OpenBSD. :) Nick.
dhclient
I am running a fairly simple dhclient on my OBSD 4.4 box and it runs as a firewall. bge0 = lan bge1 = wan dhcp to ISP What I have discovered is that all works well UNTIL the ISP modem is rebooted. At that point, dhclient seems to sleep and then VANISH. For example...I am running fine and then reboot my ISP modem: Here is what happens: 06:45:48 fw dhclient[16611]: DHCPREQUEST on bge1 to 255.255.255.255 port 67 06:45:55 fw dhclient[16611]: DHCPREQUEST on bge1 to 255.255.255.255 port 67 06:46:02 fw dhclient[16611]: DHCPDISCOVER on bge1 to 255.255.255.255 port 67 interval 4 06:46:06 fw dhclient[16611]: DHCPDISCOVER on bge1 to 255.255.255.255 port 67 interval 6 06:46:12 fw dhclient[16611]: DHCPDISCOVER on bge1 to 255.255.255.255 port 67 interval 11 06:46:15 fw dhclient[16611]: DHCPREQUEST on bge1 to 255.255.255.255 port 67 06:46:19 fw dhclient[16611]: DHCPREQUEST on bge1 to 255.255.255.255 port 67 06:46:19 fw dhclient[16611]: DHCPACK from 75.9.96.1 06:46:19 fw dhclient[16611]: bound to 75.9.X.X -- renewal in 300 seconds. 06:46:23 fw dhclient[16611]: DHCPDISCOVER on bge1 to 255.255.255.255 port 67 interval 13 06:46:36 fw dhclient[16611]: No DHCPOFFERS received. 06:46:36 fw dhclient[16611]: Trying recorded lease 75.9.X.X 06:46:39 fw dhclient[16611]: No working leases in persistent database - sleeping. ..and then nothing more. Sleep becomes death, until a reboot or a manual run of 'dhclient bge1' is invoked. Any thoughts? -JD
Re: bash for root?
Dieter wrote: 2. don't use bash as shell for root. Or at least understand what you are doing. What is wrong with bash as shell for root? (Assuming bash is in /bin and statically linked.) There's nothing wrong with that if you make it statically linked and put it in /bin. You know what, and why, you are doing it ;) My only advice to the OP was to be careful which shell to choose for the root account, especially bash which is dynamically linked and installed by default in /usr/local/bin/. I hope I didn't offend too many others with my suggestion... All the best, Stijn
Re: Samba printing, OpenBSD client to Windows server
Ed Ahlsen-Girard wrote: Alexander Hall wrote: Ed Ahlsen-Girard wrote: OK, I've installed Samba, and gotten printcap set such that I printed a straight text fire, but nothing else works now that I tried to print other formats through gv and open-office. Perhaps Samba is not the way to go? Printcap below. #$OpenBSD: printcap,v 1.4 2003/03/28 21:32:30 jmc Exp $ #lp|local line printer:\ #:lp=/dev/lp:sd=/var/spool/output:lf=/var/log/lpd-errs: #rp|remote line printer:\ #:lp=:rm=printhost:rp=lp:sd=/var/spool/output:lf=/var/log/lpd-errs: lp|hpoffice:rp=hpoffice:rm=192.168.1.100:sd=/var/spool/lpd/hpoffice:af=/var/spool/lpd/hpoffice/acct:if=/usr/local/bin/smbprint:mx=0:lp=/dev/null: For local printing, samba does nothing. Unless your printer supports postcsript natively (most cheap printers don't) you need some kind of converting filter. For my canon i550, i'm using apsfilter combined with ghostscript, both available as packages/ports. Dont know if /usr/local/bin/smbprint in your printcap is some filter like that or where it comes from. Can't find it in any port. /Alexander It's not local printing. It's an HP OfficeJet hung on a Windows XP machine. Ok, so your machine is a samba client. That explains the if=... part. :) Anyway, if I'm not terribly mistaken, samba or that XP host will not help you convert from postscript (or whatever the program printing from is delivering) to the printers native language, so if your printer does not grok postscript you still need some kind of filter. However, most printers, if not all, can print normal ascii text. I have no idea what capabilities the HP OfficeJet has, but to me it still seems like ghostscript or something else is needed to convert your output to a suitable format for the printer. /Alexander
