Panic booting cdboot-5.2-BETA-20031201-JPSNAP.iso

2003-12-01 Thread Andreas Braukmann
Hi there,

Hardware:   Tyan 2880 (K8S), single Opteron 240
   Adaptec 2200 or 2120, 2 Arrays (20 GB RAID-0, 20 GB RAID-5)
booting the install kernel gives:
[normal boot process]
aacd0: RAID 5 on aac0
aacd0: 20480 MB (41943040 sectors)
GEOM: create disk aacd0 dp=0xc2dc1c0c
aacd1: RAID 0 (Stripe) on aac0
aacd1: 20480MB (41943168 sectors)
GEOM: create disk aacd1 dp=0xc6fa980c
panic: mutex Giant not owned at /usr/src/sys/cam/cam_xpt.c: 4822
cpuid = 0;
Debugger(panic)
Stopped at Debugger+0x55: xchgl %ebx,in_Debugger.0
db trace
Debugger(...)
panic(...)
_mtx_assert
xpt_done
aac_cam_complete
aac_complete
taskqueue_run
taskqueue_swi_run
ithread_loop
fork_exit
fork_trampoline
--- trap 0x1, eip = 0, esp = 0xe22b2d7c, ebp = 0 ---


I'm sorry for not having connected a serial console. If anybody needs
the full trace information I would gladly reproduce the crash and write
the complete information off.
Yesterdays (2003/11/30) snapshot yields the same result.

-Andreas

___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to [EMAIL PROTECTED]


Re: Panic booting cdboot-5.2-BETA-20031201-JPSNAP.iso

2003-12-01 Thread Andreas Braukmann
--On Montag, 1. Dezember 2003 17:41 Uhr +0100 Andreas Braukmann [EMAIL PROTECTED] wrote:

Hardware:   Tyan 2880 (K8S), single Opteron 240
Adaptec 2200 or 2120, 2 Arrays (20 GB RAID-0, 20 GB RAID-5)
[...]
GEOM: create disk aacd1 dp=0xc6fa980c
panic: mutex Giant not owned at /usr/src/sys/cam/cam_xpt.c: 4822
cpuid = 0;


I took another K8S / Adaptec 2200S box and (to rule out any hardware
issues) was able to reproduce the panic with all -current-snapshots back
to 20031128. I don't have former kernels available.
-Andreas

___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to [EMAIL PROTECTED]


sym0 v1.4 vs. SMP

2000-02-02 Thread Andreas Braukmann

Hi there,

 ... I have an SMP-box (dual PPro, Natoma, mptable output attached) running 
 -current (with src/sys/dev/sym/sym_hipd.c,v 1.4 2000/01/08 19:58:17)
 from around 2000/01/22. 

 The machine was running a 'ca. mid 1999' -current/SMP happily
 for the last few months. I did the upgrade (fresh install) to give 
 the mylex drivers a shot. And because of the non-booting mlx-driver
 I left my symbios logic dual channel (symbios logic '876 chip) 
 hostadaptor in the machine and included the sym-device in the
 kernel configuration. (I run the sym-driver on all our other
 symbios logic hostadaptors)

 Today I switched from GENERIC (UP) to a more specific configuration
 and to SMP (again). I attached the kernel configuration file for the
 uni-processor case, to get the 'hang' one just has to throw over
 the SMP-switch.

 The machine hangs right after this message (boot -v).

 SMP: AP CPU #1 Launched!
 SMP: CPU1 apic_initialize():
  lint0: 0x00010700 lint1: 0x00010400 TPR: 0x0010 SVR: 0x01ff


 An aequivalent uni-processor kernel (with sym0, just without SMP)
 runs flawlessly.
 I attached the dmesg-outputs of a successfully booted SMP-kernel (ncr0
 instead sym0) and and successfully booted UP-kernel (with sym0).
 If anybody needs further information, ... just ask.

 Any hints?

 Thanks,
Andreas

-- 
Me no internet, only janitor, me just wax floors.



===

MPTable, version 2.0.15

---

MP Floating Pointer Structure:

  location: BIOS
  physical address: 0x000f0e70
  signature:'_MP_'
  length:   16 bytes
  version:  1.1
  checksum: 0x02
  mode: Virtual Wire

---

MP Config Table Header:

  physical address: 0x000f0e84
  signature:'PCMP'
  base table length:308
  version:  1.1
  checksum: 0x2f
  OEM ID:   'OEM0'
  Product ID:   'PROD'
  OEM table pointer:0x
  OEM table size:   0
  entry count:  30
  local APIC address:   0xfee0
  extended table length:0
  extended table checksum:  0

---

MP Config Base Table Entries:

--
Processors: APIC ID Version State   Family  Model   StepFlags
 0   0x11BSP, usable 6   1   9   0xfbff
 1   0x11AP, usable  6   1   9   0xfbff
--
Bus:Bus ID  Type
 0   PCI   
 1   ISA   
--
I/O APICs:  APIC ID Version State   Address
 2   0x11usable  0xfec0
--
I/O Ints:   TypePolarityTrigger Bus ID   IRQAPIC ID PIN#
ExtINT   conformsconforms1 0  20
INT  conformsconforms1 1  21
INT  conformsconforms1 0  22
INT  conformsconforms1 3  23
INT  conformsconforms1 4  24
INT  conformsconforms1 5  25
INT  conformsconforms1 6  26
INT  conformsconforms1 7  27
INT  conformsconforms1 8  28
INT  conformsconforms1 9  29
INT  conformsconforms110  2   10
INT  conformsconforms111  2   11
INT  conformsconforms112  2   12
INT  conformsconforms113  2   13
INT  conformsconforms114  2   14
INT  conformsconforms115  2   15
INT active-lo   level0   8:A  2   16
INT active-lo   level0   9:A  2   17
INT active-lo   level0  10:A  2   18
INT active-lo   level0  11:A  2   19
INT active-lo   level0   7:A  2   19
INT active-lo   level0  12:A  2   16
SMI  conformsconforms1 0  2   23
--
Local Ints: TypePolarityTrigger Bus ID   IRQAPIC ID PIN#

Re: very silent, but heavy filesystem-crash

2000-01-29 Thread Andreas Braukmann

Hi,

 ... thanks for your assistance.
 I

On Fri, Jan 28, 2000 at 08:22:32AM -0800, Rodney W. Grimes wrote:
 Next step is run the ``save'' fsck:
 fsck -n /scratch /someplacewithspace 21

  there is some illuminated filesystem-hacker who would like to
  take a peek. (I would allow ssh-access to the box ...)
 
 $1.00/MB for me to recover it 
... I was used to this kind of service for hosed FAT-filesystems 
and hosed partition tables ;) ...
(I did it by stuffing the 'hosed disks' into one of my unix-boxes
and working on the raw devices ...) 

 if you need it.  
... fortunately not. Just have a look at the mountpoint's name 

 I don't do it for fun.
... understandable :), I made the offer just for the case 
anybody wants to track down a potential bug.

But as it comes out I had (in fact :( ) a typo in the disklabel.

I checked the disklabel two times before my post to -current, but
I had obviously a little problem with my 'eye-brain-interlink'.

I really shouldn't admit this in this forum, should I? ... But
in relabeling the disk I left the disklabel as follows:

#size   offsetfstype   [fsize bsize bps/cpg]
  b:  10485760  swap# (Cyl.0 - 65*)
  c: 179124120unused0 0 # (Cyl.0 - 1114*)
  g: 1791241204.2BSD 1024  819216   # (Cyl.0 - 1114*)


Sorry for bothering all of you, feeling soo stultified 
and the big and heavy hat (with a certain pattern on it) weighs
on my head ...


Thanks,
Andreas

-- 
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Mylex AcceleRAID 250

2000-01-22 Thread Andreas Braukmann

Hi Mike,

 ... today I tried to build our first Mylex-RAID based FreeBSD-system.

 Hardware (if you need any more details, just feel free to ask):

Gigabyte Dual PPro Board, Natoma, 2*PPro/200
dual channel SymbiosLogic SCSI-Hostadapter (BIOS deactivated),
- no disks
- just one CD-ROM on the first channel

Mylex AcceleRAID 250/8MB ECC-Cache
- EzAssist 1.00-16
- Controller 0: DAC960PTL1, FW V4.07-29
- 8 GB BIOS translation

- 3 * IBM DNES DNES-309170W
- RAID-5, 2 logical disks

FreeBSD 4.0-current snapshot 2103

 The installation (over nfs) went very well.
 I sliced and labeled just the mlxd0.
 mlxd0s1 was set bootable and I've chosen to install a standard mbr.

 On Reboot I just get an 'missing operating system'.

 If I boot from the install floppies again, the RAID-volume ist detected
 fine.

 Do you have any hints on what to try next?

 
 As I am talking to you yet in any case, I would like to ask you
 what cache-size to choose. Is it really advisable to stuff the
 full 64 MByte onto the controller? Where lies the 'sweet-spot'
 on the cache-size/performance curve?

Kind regards and thanks in advance,
Andreas


P.S. Thanks for this very valuable driver. We're now able to support
 RAID-configurations even in a 'small-server' setup.
 The AccelRaid 250 is about $500 in Germany, very affordable
 compared to a full-blown SCSI2SCSI CMD5440 or the like.



-- 
: TSE GmbH Neue Medien  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Filesystem crash / no softupdates / jail running ...

2000-01-19 Thread Andreas Braukmann

Hi there,

 ... I had a major filesystem crash just a few minutes ago.
 The crash was so heavy, that I first had to think about a
 hardware-failure. But I didn't found anything in the logfiles.

 The story:
Yesterday I began to play around with an experimentail jail-based
virtual server setup.
The 'jail world' was located under /scratch/jails/zappa/.
'inetd', 'telnetd', a 'sh' were running jailed.
It sees to run rather well.

Today I changed my working directory (running a regular, non-jailed
shell) to /scratch and found the directory empty.

'What the hell is going on here?' I thought ran a 'ps' just to 
find the jailed processes running happily.
I did a telnet to the virtual server, was able to login successfully
and found a apparently intact filesystem.
... and then.. ... and then ... the screen froze and a few seconds
later the machines rebooted silently.
No tracks in the logs.

 Sorry, I was running X and currently have no serial console available.
 Therefor I just can't tell, if it was really a silent reboot or 
 a 'not-seen' panic.

 The filesystem was totally hosed. 
 'fsck' spat out masses of 'unknown file type', 'partially allocated
 inode', 'incorrect block count' (with rather large discrepancies) 
 and 'excessive bad blks' messages.
 Partially transcripts of 'fsck -y '-logs are available on request.

 

 System:K7/700, 256MB, sym0, fxp0
 Disk:  IBM DNES-309170W SA30

 The disk contained a ca. 9GB filesystem (default newfs) mounted
 without any special options (on /scratch).

 FreeBSD xxx.xxx..de  4.0-CURRENT FreeBSD 4.0-CURRENT #0: Wed Jan
 5 14:30:28 CET 2000

 I have soft-updates compiled in but deactivated on all filesystems.
 

Regards,
Andreas

... sorry, ... my written English is rather ugly ...

-- 
Andreas Braukmann - [EMAIL PROTECTED] - 


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: XFree86 3.9.17

2000-01-18 Thread Andreas Braukmann

(trimmed the long cc:-list)

On Tue, Jan 18, 2000 at 07:51:22AM +0200, Vallo Kallaste wrote:

 The problem with 3.3.5 was something to do with refreshing, example
 after doing some ls -la in xterm the xterm screen got blank completely
 or showed some lines of text in the bottom but otherwise remained black.
I noticed no such problems since running 3.3.5 in 1600x1200/bpp16.
Am I just lucky? My board is a G400DH32.

-Andreas

-- 
 The computer revolution is over. The computers won.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: XFree86 3.9.17

2000-01-17 Thread Andreas Braukmann

Hi,

On Mon, Jan 17, 2000 at 06:05:54PM -0500, Donn Miller wrote:
  Have you tried to install it? It does builds but barfs at some point for
  install process. Not that I care so much about it, 3.3.6 works for my
  G400 now, thanks god.
... what was the problem with the G400 in earlier revisions?
3.3.5 (built from the port) lives quite happily on my G400?

I'm very eager to get my hands on a XFree86 with dual-head support 
and truetype-rendering, therefor my interest in the XFree86 prereleases.

 During make install, I switched compilers by setting
 PATH=/usr/local/bin:$PATH. 
ahhh.

 The compiler I used was the development version of gcc,
 gcc-2.96-current. 
Is the gcc-2.96-current in the ports collection? 
Browsing through /usr/ports/lang I've detected only the older egcs-port?


-Andreas

-- 
Andreas Braukmann - [EMAIL PROTECTED]
 The computer revolution is over. The computers won.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: No disks found msg in /stand/sysinstall, how come?

1999-12-19 Thread Andreas Braukmann

Hi Mike,

On Sat, Dec 18, 1999 at 06:01:21PM -0800, Mike Smith wrote:
  On Sat, Dec 18, 1999 at 04:21:19PM +0200, Vallo Kallaste wrote:
Is is not possible to configure and add a second disk via the
/stand/sysinstall Label and Fdisk menu's?
  ... I got this, too, ... but
  
   Hmm, yes I got it too. Me thinks it's related to change from block
   devices to char and to the MAKEDEV. Anybody out there please
   correct me if I'm wrong. I had to do standard disklabel -rw ad1
   auto and after that edit the disklabel by hand. It's certainly not
 
 Sysinstall is not build with 'make world', and should not be used on a 
 system that's been updated that way.
... but my /stand/sysinstall came from a freshly installed snapshot
(19991208). 

In principle I'm (or was) rather used to using disklabel
for labeling new disks.
But recently (I would say for the last 12 months) I had various
difficulties in disklabeling new drives. (on -stable and on -current)
I found a solution in each of the cases eventually, but it made me not
really happy. Once I just 'dd-ed' an existing label from another disk.
This time I tried this strategy, too, but to no avail, since writing
back the afterwards edited disklabel seemed not to be possible.
(the "no space left on device"-message appeared ... )

  ... just after having the standard disklabel -rw ad0 auto (my boot disk is
  da0) complained about 'no space left on device'. Yes, ... I did the
  disklabel -W ad0 ...
 
 You forgot to overwrite the beginning of the disk.
... and yes I did (before doing the 'fdisk') a nice little
dd if=/dev/zero of=/dev/rad0 count=1 

Do you have any further hints?

Kind regards,
Andreas


Sidemark: Yes, ... I'm running -current on a few machines (and yes I
know, that I get what I pay for ;). I'm used to live with
-current since I switched my cvsup-Tag from stable to current on a
'brand new' dual PPro/200.

standard disclaimer: I really don't want to sound insulting or arrogating 
 or whatsoever.
 If anybody really might read such an undertone from my
 message(s), please ignore it and blame it on my rather 
 ugly written English.


-- 
Andreas Braukmann - TSE GmbH - Neue Medien - http://www.tse-online.de



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Flash (was: Re: Sound card support)

1999-12-10 Thread Andreas Braukmann

Hi,

On Thu, Dec 09, 1999 at 11:16:58PM +0300, Dmitrij Tejblum wrote:
  Good luck using it under current.
  
  First site you hit quits netscape without reasons...
... yes. I built the port for the first time just yesterday and made
the same experience.

  ...until you drop out of X and see a __sh_getcontext  IIRC warning on
  your console.
 
 If you can hack on the flash plugin's Makefile, try add -fno-exceptions 
 there.
... many thanks for the tip.
It's up and running.

Regards,
Andreas


P.S.: ... the port's maintainer is cc-ed ...

-- 
: TSE GmbH Neue Medien  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: The Matrix screensaver, v.0.2

1999-08-21 Thread Andreas Braukmann

On Fri, Aug 20, 1999 at 08:10:51PM -0400, Keith Stevenson wrote:
 On Fri, Aug 20, 1999 at 04:11:42PM -0600, Oscar Bonilla wrote:
  then xlock already includes a module for this... I'm using it on my
  laptop..
... ahh. ... Meanwhile I've updated my xlock to xlockmore-4.14 including
'the matrix'.  I'm rather disappointed.

a) it crawls like a dog without legs :( (especially on my 1600x1200 screen)
 I like Andrzej's module a _lot_ better than the xlock module.  It looks (IMO)
 more like the screen from the movie and doesn't eat all of my CPU cycles.
b) I second this.

-ab

-- 
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: The Matrix screensaver, v.0.2

1999-08-20 Thread Andreas Braukmann

Hi,

On Fri, Aug 20, 1999 at 07:34:31PM +0200, Andrzej Bialecki wrote:
 Due to unexpected demand (I did it mostly as a distraction from real
 work), I backported this screensaver to 3.2-RELEASE (don't know about
... I saw the announcement for the first release, realised that
it's only for -current; went home, tested it on my personal -current
machine and instantly wondered how many messages about a port to
-stable you might had got already.
I decided to just let the others do the begging ;)))

What a pity, that our servers are running on those smallish
gray-scale 9" screens.
And what a pity for 'Chuck' being banned from our consoles ...;)

What about a xlock-more module?

Thank's alot for the amusement ...

-Andreas



-- 
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



just pulled the egcs-switch

1999-05-07 Thread Andreas Braukmann
Hi there,

 ... yesterday I decided, that it might be the right time to
 actually pull the 'egcs'-switch on my 4.0-current SMP-box.
 I cvsupped, built and installed the world, tweaked the 
 kernel config (newbus) ...
 The transition was smooth, so far.
 
 The question remained: ... how to easily figure out which ports
 are in  the egcs-caused need for a rebuild?

 ok. ... all the c++-based ones, but how to find out which ports
 are c++ ones?

 Is anybody out there, with hints or experiences to share?

TIA,
Andreas

-- 
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message



Re: just pulled the egcs-switch

1999-05-07 Thread Andreas Braukmann
Hi,

On Fri, May 07, 1999 at 01:10:31AM -0700, Amancio Hasty wrote:
 Is not a good idea to use the default egcs as it is configured to compile
 c++ programs which use threads and exceptions for instance Netscape's
 Eletrical Fire (A Java VM/Compiler) does not work.
do you suggest to use the egcs from the ports collection as
the 'more-or-less' default compiler for the ports?

I've used the beast for a few c++ applications (from ports
and various other sources), which needed the more advanced c++-compiler
than gcc 2.7.x.

But then I have '2 different compilers' and '2 different set of libs',
again, havn't I?

-Andreas


-- 
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message



Re: default route not set up??

1999-05-02 Thread Andreas Braukmann
Hi,

On Sat, May 01, 1999 at 03:18:23PM +0200, Dag-Erling Smorgrav wrote:
 Daniel C. Sobral d...@newsguy.com writes:
  Dag-Erling Smorgrav wrote:
   # Configure routing
interesting code deleted ... [..]
  Looking there, it *does* seem that there is a problem. defaultrouter
  is only used to set route_default, which is *not* used. At least on
  the code you quoted.
 This is getting fun :) Look closer at the first line I commented, then
 at the loop, and especially the eval.

since I've stumbled over the same 'what the the hell is going on
here?'-piece of shell script once upon a time, I've foreseen the
further development of this after having read the first two 
messages.
I had to set a few static routes, saw the empty 'static_routes'
configuration variable and just didn't grasped what to do know
with that variable.
I considered, that there has to be some underlying 'magic' similar
to the method used for the interface configuration.
Despite of this cognition I had to jump head first into the
rc.network script to investigate on 'how to use' the core variable
(static_routes) and the supplementary variables the right way.

It would be very nice, if 'somebody' (TM) would document the 
various 'dark magic' parts of the rc scripts.

But, ... a person priviledged to setup the routing for 
a system (aka system administrator aka 'root') should really be
able to parse 'sh'-code, shouldn't he?

-Andreas


-- 
: TSE TeleService GmbH  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message



Subnotebook capable of running -current

1999-03-30 Thread Andreas Braukmann
Hi,
   I want to purchase some kind of 'subnotebook', that has to
   be as -current-compatible as possible.

   I anticipate the crowd whispering Libretto, ... Libretto, but
   I have to say, that I don't like the Libretto very much for its
   crude screen resolution.

   Currently I'm tending towards the Toshiba Portege 3010
   or the Fujitsu LifeBook B110 Biblo.

   Any experiences with these decices?
   
   I'm wondering about the CD-drives connected thru a PCMCIA-Card 
   and FreeBSD-compatible PCMCIA-Network-Cards.

Thanks,
Andreas

-- 
: TSE TeleService GmbH  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: Anti-Spam Petition: http://www.politik-digital.de/spam/:
: PGP-Key:http://www.tse-online.de/~ab/public-key:
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :


To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message



Re: adding DHCP client to src/contrib/

1999-02-08 Thread Andreas Braukmann
Hi,

On Mon, Feb 08, 1999 at 02:28:20PM -0800, David O'Brien wrote:
 I am planning on adding the Wide-DHCP client to src/contrib/ and
 src/sbin/ in a few days.
speaking about DHCP-things (clients and servers) I would like to 
hear from some 'export' on choosing a the 'right' dhcp-software
for FreeBSD.
I'm mostly interested in choosing the server part. In the past I
deployed only the wide-dhcp server from the ports collection; but
only because I somehow decided to test the wide-dhcp before the isc one.

A few days ago I had a look (triggered by a piece of samba documentation)
at isc-dhcp and found the configuration syntax somewhat more user friendly.

Are there any really hard facts on how to decide?


 The plan is to make a boot floppy / boot CDROM with a DHCP client on it.
fine.

Kind regards,
Andreas

-- 
: TSE TeleService GmbH  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: PGP-Key:  http://www.tse-online.de/~ab/public-key  :
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :

To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message


Re: adding DHCP client to src/contrib/

1999-02-08 Thread Andreas Braukmann
Hi,
 ...sorry, I just didn't get it ...

On Mon, Feb 08, 1999 at 04:32:39PM -0800, Steve Kargl wrote:
 I am planning on adding the Wide-DHCP client to src/contrib/ and
 ^^
^^^
! client !
 src/sbin/ in a few days.
   [...]

  If we want FreeBSD to have any credibility as a workstation OS, we
  need DHCP. 


  It should be possible for a user or admin to smack in the
  boot floppy, have it autoconfigure the selected network interface, and
  perform an FTP installation.
 
 So, we'll import a pop server, apache, g77, ad nauseam
 to increase the credibility of FreeBSD as a workstation OS.
Read his statement again; the (IMHO very valuable) itention is to
be able to build a sysinstall boot floppy for network-installations
without having the installing person go through the hassle of setting
up a _static_ ip configuration. (Especially if the workstation 
should be a dhcp client later on.)

Furthermore I know more than one network where 'dhcp' is the only
mean to get a valid ip-configuration.

Scenario: User-A wants to setup his brand new FreeBSD-workstation:

[without dhcp-client on boot-disk]

- boot the bootdisk ;)
- choose distribution media 'ftp' or 'nfs'
- get to the network-configuration screen
- don't find a knob named 'use dhcp'
- yell after the net-admin Hey, ... I need a static IP, which one
  should I use?
  [ - The ever-friendly net-admin is not available
- wait several days ... ;) ]
- install / configure the machine for the static ip
- cd /usr/ports/net/dhcp*; make  make install;
- change the machine's configuration (/etc/rc.conf etc.) to make
  use of the dhcp-client
- tell the net-admin that the static ip could now be 'recycled'

[with dhcp-client on boot-disk]

- boot the bootdisk ;)
- choose distribution media 'ftp' or 'nfs'
- get to the network-configuration screen
- select 'use dhcp'
- install
- be happy

Including a dhcp-client for the boot-disk is something very, very
different from including various 'convenience' software in the
base system. 


Regards,
Andreas

-- 
: TSE TeleService GmbH  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: PGP-Key:  http://www.tse-online.de/~ab/public-key  :
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :

To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message


Re: some woes about rc.conf.site

1999-02-07 Thread Andreas Braukmann
Hi,

On Sun, Feb 07, 1999 at 06:05:15PM +0100, Andreas Klemm wrote:
  Sorry to say this, but you really don't understand it. :)
sorry Andreas, ... I have to second this ;)

   When we had one central rc.conf file it was fun to browse through
   it and having all supported knobs visible at a glance.
  
  And you still have this now.  In fact, with the unadulterated rc.conf, you
  have the original default values for youre reference.
hmmm. I hated the old behaviour of sysinstall to make changes directly to
/etc/rc.conf.
Why? Because I'm used to use /etc/rc.conf just as a 'reference manual'
for all the 'knobs'. If mergemaster tells me that rc.conf has changed,
I have the 'diff' as a rough guideline if I have to change my 
rc.conf.locale, too. 

Typically I use 'sysinstall' exactly once in one machine's lifetime.
My old method of dealing with 'rc.conf' and 'rc.conf.local' was:
= sysinstall generates a modified rc.conf
= mv rc.conf rc.conf.local
= cp /usr/src/etc/rc.conf rc.conf
= vi rc.conf.local
   delete all the lines not suitable for rc.conf.local

after making a new world:
= mergemaster
   if there are diffs between the old and the new rc.conf
   = let mergemaster install the new rc.conf
   = have a close look at the 'diffs' and check if any of the
  changes conflict with my current rc.conf.local.

Now, with 'rc.conf.site' I just don't have to bother with rc.conf
after a fresh installation. I would just move rc.conf.site to rc.conf.local
and then procede as earlier mentioned.

   Then rc.conf.site has a totally different sort order which is
   not very helpful/comfortable, when comparing rc.conf and rc.conf.site.
I have to admit, that I havn't met a real rc.conf.site, yet. 
If the sort order differs significantly it should really be corrected.

   Then rc.conf.site doesn't contain every knob which rc.conf has.
  Erm, it's not supposed to.  It's supposed to contain only those knobs
  you want to change.  
Just as I have only the minimum set of knobs in rc.conf.local.

 or am I really the only person, who hacks rc.conf.site with vi 
no :=)

 both files at the same time and is a bit annoyed by having to compare
 every single line and then to add the knob in rc.conf.site ?!
hmm. 
diff rc.conf.old rc.conf should point you directly to the changed
options or changed default behaviors.

 But, if I'm the only person who complains, then forget about it.
 It's not so important then ;-)
;) ..

Regards,
Andreas

-- 
: TSE TeleService GmbH  :  Gsf: Arne Reuter: :
: Hovestrasse 14:   Andreas Braukmann  : We do it with   :
: D-48351 Everswinkel   :  HRB: 1430, AG WAF   :  FreeBSD/SMP:
::
: PGP-Key:  http://www.tse-online.de/~ab/public-key  :
: Key fingerprint:  12 13 EF BC 22 DD F4 B6  3C 25 C9 06 DC D3 45 9B :

To Unsubscribe: send mail to majord...@freebsd.org
with unsubscribe freebsd-current in the body of the message