etherboot-users — Discussion of user issues

[Etherboot-users] AWARD BIOS and etherboot module integration - PROBLEM

[<ka...@se...>]

AWARD BIOS and  etherboot module integration - PROBLEM

Hello (and sorry please my English)

I have 4 old pentium PCs with AWARD Modulas BIOS-es 4.51PG (+-1996)
(2x chiset Triton II (430VX), 1x Triton 1 (430FX) and 1x VIA 570)

I´m NOT able to integrate *.zrom modules into motherboard BIOS-es.

First - after making place (if necessary)- i.e.:
- cbrom215 bios.bin /other 1000:0000 release (it is aha7850.16k) and
- cbrom215 bios.bin /other 0800:0000 release (it is ncr306.16k)

I get enough place -> adding module:

- cbrom215 bios.bin /isa ne.zrm D000:0000	XOR
- ...               /other ...			XOR
- ...		    /pci rtl8029.zrm		XOR
- ...		    /pci rtl8029.zrm D000:0000	XOR
- ...		    /other rtl8029.zrm D000:0000

-- "original" names are (from rom-o-matic.org):
eb-5[1].4.3-ns8390--ne--No_Delayed_INT.zrom
eb-5[1].4.3-ns8390--ne.zrom
eb-5[1].4.3-rtl8029--NO_DELAYED_INT--pci_direct.zrom
eb-5[1].4.3-rtl8029--NO_DELAYED_INT.zrom
eb-5[1].4.3-rtl8029--PCI_DIRECT--INT18.zrom
eb-5[1].4.3-rtl8029--PCI_DIRECT.zrom
eb-5[1].4.3-rtl8029.zrom
eb-5[1].4.3-rtl8139--CONFIG_PCI_DIRECT--INT18.zrom
eb-5[1].4.3-rtl8139--CONFIG_PCI_DIRECT--NO_DELAYET_INT.zrom
eb-5[1].4.3-rtl8139--CONFIG_PCI_DIRECT.zrom
eb-5[1].4.3-rtl8139.zrom

... an many other variants - different drivers,
variants (with/out NO_DELAYED_INT, CONFIG_PCI_DIRECT, ...),
lokations (C800:0000), ...


NOTHING help


 By 2 PC's - after freeing space (and adding etherboot module) ->
 - modbin says: "CHECKSUM Error" - and when I try to flash them anyway
 (with awdflash),  then the flash utility crash (after loading .BIN file
 - before asking  to flash Y/N)
 - I thing -it is the moment, when awdflash computes/checks the CHECKSUM.

 By other 2 PC's freeing space and inserting module seems to go OK.
 I flash it in AWARD BIOS, restart the computer .... and NOTHING.

 It is like if the module is not there at all- NO reaction or message
 on monitor when booting
 - just NOTHING happens - The PC's boot from FD/HDD. And if I disable 
 them - then PC stops with "No system disk or ... "

 I tryed unde MS-DOS 6.2 to run i.e. HWINFO - there are no BIOS extensions listed
 I tryed also MS-DOS utility "debug" to DUMP locations like D000:0000, ....
 - there are all "FF FF FF FF ......"

 In Norton Commander I have viewed (in hex) the module - i.e. ne.zrom - it seem to by
 OK - It begins with 55 AA 40 (32K bios module) followed also with strings like
 "...Etherboot...", ...

 I tryed also to search them in memory (again with debug utility)
 in all 64K segments C000 - F000 - there is NOTHING
 -i.e. "- s D000:0 ffff 55 aa 40" (or byte representation of string Etherboot, ..)
 NOTHING, NOTHING,NOTHING,NOTHING, ....

 => I thing, that the module is not im BIOS at all 
 (even if cbrom215 bios.bin /D says YES)
 Or the module is from BIOS not decompressed or is damaged.
 It is not possible to find in memory (debug) either the 55AA prefix, or any
 substring (i.e. from etherboot)

 there is NOTHIG - also I'm not suprized, that the main BIOS cann not 
 find an initialize this BIOS extension - > .... :-///


-- !!! *.zdsk variants works of course fine !!!

 Can someone HELP ME PLEASE? I get CRAZY !

 Where is the Problem ? What am I doing wrong ?

 Thanks for replay.

 Karel Petranek     ka...@se...

[Etherboot-users] embedding etherboot in an existing loader

[Maule M. <mar...@ya...>]

Hi:

I have an existing boot loader (custom, not based on anything standard like grub or LinuxBIOS).  Is it possible to embed an etherboot image into my loader and jump directly to it (from say a menu option in my loader)?

thanks
Mark


 
____________________________________________________________________________________
Now that's room service!  Choose from over 150,000 hotels
in 45,000 destinations on Yahoo! Travel to find your fit.
http://farechase.yahoo.com/promo-generic-14795097

Re: [Etherboot-users] Restructuring Etherboot mailing lists

[Hermann G. <eth...@mr...>]

On Thu, Mar 16, 2006 at 09:18:55AM -0500, Marty Connor wrote:
> So, I ask.  Does anyone have an opionion (for or against or neutral) =20
> about merging the Users and Developers lists?=20

i=B4m definitely pro to such a change; this should have happend before!
i hope, if you make that happen, that you move all existing users to=20
the new list; would be boring, if everbody has to resubscribe.

go on and do it ;-)

mfg hermann

[Etherboot-users] Re: [Etherboot-developers] Restructuring Etherboot mailing lists

[Roland K. <rk...@e1...>]

Hi Marty!

On 16 Mar 2006, at 15:18, Marty Connor wrote:

> Alright,
>
> Now that the LinuxWorld forms and contracts are filled out and  
> faxed in, let's get back to other stuff.
>
> I'm sending this to both lists because there may be some people on  
> the Etherboot-Users list who are interested in helping out with  
> stuff that is sort of "Developer-ish".
>
> I'm really not a fan of having two lists, unless one of the lists  
> was Etherboot-Announce, and the other one was Etherboot-Discuss.   
> The problem with Etherboot-Users and Etherboot-Developers is that  
> it segregates people by what they do, and not what they are  
> interested in.
>
Seconded. Let's have all in one neat package :-) (just my 2c)

Ciao,
                     Roland

--
TU Muenchen, Physik-Department E18, James-Franck-Str., 85748 Garching
Telefon 089/289-12575; Telefax 089/289-12570
--
CERN office: 892-1-D23 phone: +41 22 7676540 mobile: +41 76 487 4482
--
UNIX was not designed to stop you from doing stupid things, because that
would also stop you from doing clever things.
	-Doug Gwyn
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GS/CS/M/MU d-(++) s:+ a-> C+++ UL++++ P+++ L+++ E(+) W+ !N K- w--- M 
+ !V Y+
PGP++ t+(++) 5 R+ tv-- b+ DI++ e+++>++++ h---- y+++
------END GEEK CODE BLOCK------

[Etherboot-users] Restructuring Etherboot mailing lists

[Marty C. <md...@et...>]

Alright,

Now that the LinuxWorld forms and contracts are filled out and faxed  
in, let's get back to other stuff.

I'm sending this to both lists because there may be some people on  
the Etherboot-Users list who are interested in helping out with stuff  
that is sort of "Developer-ish".

I'm really not a fan of having two lists, unless one of the lists was  
Etherboot-Announce, and the other one was Etherboot-Discuss.  The  
problem with Etherboot-Users and Etherboot-Developers is that it  
segregates people by what they do, and not what they are interested in.

What I mean is, Users vs. Developers asks people to either be a  
developer or a user. Coders vs. Non-Coders, basically.  Discuss vs.  
Announce asks whether you want to hear the discussions or just the  
result (releases, major changes).  And even the Announce list might  
be pushing it.  It would probably get used only for releases.

We just don't have the volume of traffic so that it's too big a deal  
to ignore some messages.  Further, digest-mode is available on  
SourceForge, for those who like to skim.

So, I ask.  Does anyone have an opionion (for or against or neutral)  
about merging the Users and Developers lists?  There are about 150  
people on Etherboot-Developers, and 362 on Etherboot-Users.  I'm  
guessing that most developers are on the users list, so that would  
leave about 200 people who might be on users alone.

For now, I'd like to keep the GPXE discussions (which will start  
soon) on the Etherboot list(s), until we the code base has diverged  
more from Etherboot-5.4.  Those of you who have been around long  
enough will recall that Etherboot started on the Netboot mailing  
list, and it was quite some time before it became separate.

Your thoughts and ideas --  and everything you contribute, be it  
testing, documentation, help on the mailing lists, coding, hardware,  
donations, are all appreciated very much.

Many thanks,

Marty

-- 
     Try: http://rom-o-matic.net/ to make Etherboot images instantly.

    Name: Marty Connor
US Mail: Entity Cyber, Inc.; P.O. Box 391827;
          Cambridge, MA 02139; USA
   Voice: (617) 491-6935; Fax: (617) 491-7046
   Email: md...@et...
     Web: http://www.etherboot.org/

Re: [Etherboot-developers] [Etherboot-users] Back to the Future, Part 2

[H. P. A. <hp...@zy...>]

Geert Stappers wrote:
> On Sat, Mar 11, 2006 at 07:42:18AM -0500, Marty Connor wrote:
>  <snip/> 
> 
>>For now, we'll basically rename the Etherboot-5.5 cvs module to  
>>GPXE-0.5 so as to create the minimum disruption to development.
> 
> What about a switch to successors of CVS
> like subversion[1]  or darcs[2]?
> 

How about switching to a SCM which doesn't use the fundamentally broken 
CVS model, like git or Mercurial?

	-hpa

[Etherboot-users] Etherboot Project at LinuxWorld Boston, April 3-6, 2006

[Marty C. <md...@et...>]

Well, it's getting to be that time again...

LinuxWorld Expo comes to Boston next month, and we've got a booth!

There's not a lot of time before the show, but since we've done this  
a bunch of times, we'll manage to pull it all together in time :)

The oh-so-famous-and-award-winning LTSP crew will be in town, and the  
show look quite big, with lots of vendors:

     http://www.linuxworldexpo.com/live/12/events/12BOS06A/exposition/ 
exhibitorlist

We'll be in the .ORG Pavilion, as usual, and will hopefully be  
demoing the wonders of Etherboot 5.4.2.  So, we better get it out,  
huh? ;)

Anyway,  if anyone is coming and wants to help in the booth, you're  
certainly welcome.  We should be getting 100 free passes, so if you  
want to come and hang out, we can arrange that as well.

Heck, you might even get to see rom-o-matic.net! :)

So, who's going to coming to Beantown?

Marty

-- 
     Try: http://rom-o-matic.net/ to make Etherboot images instantly.

    Name: Marty Connor
US Mail: Entity Cyber, Inc.; P.O. Box 391827;
          Cambridge, MA 02139; USA
   Voice: (617) 491-6935; Fax: (617) 491-7046
   Email: md...@et...
     Web: http://www.etherboot.org/

Re: [Etherboot-developers] [Etherboot-users] Final static ip patch

[Anselm M. H. <an...@ho...>]

Am Dienstag, den 14.03.2006, 16:29 +0100 schrieb Ricardo Carrillo Cruz:
> I suceeded to boot an image with a static ip address.
> The patches  are attached to this email.
> 
> Of course,feel free to post bugs or comments so we can improve the patch
> 
> Regards to all and special thanks to Marty Connor for his suggestions
> and David Aubin
> 
> Ricardo Carrillo Cruz

This looks like a fairly clean way to setup static addresses. Great.

Nevertheless, I miss a "subnet mask" parameter that might be necessary
in a setup that spans multiple physical networks but does not use
default subnet classing (A /8, B /16, C /24). You could construct cases
that make booting fail. We would be safe against this by having an
additional compile time parameter like "bitmask length", or a
dotted-quad subnet mask, of course.

Anselm

[Etherboot-users] Final static ip patch

[Ricardo C. C. <ema...@gm...>]

I suceeded to boot an image with a static ip address.
The patches  are attached to this email.

Of course,feel free to post bugs or comments so we can improve the patch

Regards to all and special thanks to Marty Connor for his suggestions
and David Aubin

Ricardo Carrillo Cruz

(I'm resending the mail in plain text,sorry :S)

[Etherboot-users] UDP Checksum Error

[Le T. T. <let...@iu...>]

Hi all,
    I use grub 0.95 with various nic cards (Sis900, Tigon3, eepro100, 
e1000). When workstations (all) download kernel image, etherboot print 
multiple "UDP Checksum Error".  However network switch packet error 
counter is at zero. Tftp server is atftpd. Someone know issue for this 
problem?

Thank you.
Tristan

 -- 
--------
Tristan Le Toullec
Service Informatique CAMPUS III - IUT De Caen
Rue Anton Tcheckov - 14123 IFS
Bureau 239 - Tel : 02.31.52.55.26

Re: [Etherboot-developers] [Etherboot-users] Back to the Future, Part 2

[<sta...@st...>]

On Sat, Mar 11, 2006 at 07:42:18AM -0500, Marty Connor wrote:
 <snip/>=20
> For now, we'll basically rename the Etherboot-5.5 cvs module to =20
> GPXE-0.5 so as to create the minimum disruption to development.

What about a switch to successors of CVS
like subversion[1]  or darcs[2]?


Cheers
Geert Stappers

[1] http://subversion.tigris.org
[2] http://abridgegame.org/darcs/

[Etherboot-users] Back to the Future, Part 2

[Marty C. <md...@et...>]

Hello Everyone,

About a year ago, I sent the following message:

     http://sourceforge.net/mailarchive/message.php?msg_id=11633136

It was a grand vision to begin migration of our primary focus from  
Etherboot to GPXE.

Well, I think the time is right to revive the idea, and take steps to  
make it happen.

So, Etherboot 5.4 will be the last branch of Etherboot, and Etherboot  
5.5 will shortly be renamed to GPXE 0.5.5.

All the reasons to do it remain the same from a year ago, and now  
there are even more compelling reasons:

First, we have a critical mass of talented hackers nearby who are  
ready to help make it happen.

Second, we have to get over the legacy associations with Etherboot.   
Just the other day I found a website that claimed that Etherboot did  
not support PXE and could only load Linux kernels:

     http://www.argontechnology.com/products.aspx?id=65

There's just so much documentation and history associated with the  
name "Etherboot" that it's hard to overcome it.  There are just too  
many misconceptions to correct.

GPXE gives us a chance for a new start.  It will more fully support  
the PXE spec, and will allow us to explore the new BIOS directions of  
the http://www.uefi.org/ folks, and understand what the implications  
of 64-bit processors network booting.

I've already secured the domains GPXE.ORG (and .NET and .COM) for  
that matter, so people will be able to find us easily.

For now, we'll basically rename the Etherboot-5.5 cvs module to  
GPXE-0.5 so as to create the minimum disruption to development.

The Etherboot-5.4 branch will remain, and fixes and improvements will  
continue to be applied for some time.  I'm sure the current Etherboot  
code base will benefit from what we discover.

Soon, I'd like to start populating GPXE.ORG with documentation and a  
new wiki focussing on how to use GPXE as your PXE stack.

Michael has done a lot of work getting Etherboot-5.5 PXE-ified.  The  
code base is much more modular, and has a new build system, an  
integration of the Adam Dunkel's uIP stack, and lots of other  
improvements.

There is still a whole lot of work to do, and the opportunity to  
really learn about the low-level business of loading operating  
systems and working at bare metal.

There is also the need for a first-class wiki, documentation re- 
writing, testing, testing, and more testing, and dozens of other  
activities.

The GPXE-Discuss mailing list exists, and sometime soon I'd like to  
copy addresses from Etherboot-Users and Etherboot-Developers to GPXE- 
Discuss so we can have a single address for GPXE discussion.

I think this will be exciting.  If you've ever wanted to learn more  
about how computers really work, and want to work on something new  
and useful, now is the time.  GPXE is the place.

Let's get it on.

Cordially,

Marty

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Jim M. <ja...@Mc...>]

On Thu, March 9, 2006 2:48 pm, Marty Connor wrote:
> On Mar 9, 2006, at 2:30 PM, Jim McQuillan wrote:
>> On Thu, March 9, 2006 2:14 pm, Marty Connor wrote:
>>> Isn't there some way to use DEFAULT_BOOTFILE to get what you want
>>> using Etherboot?
>>>
>>>      setting DEFAULT_BOOTFILE to:
>>>
>>>          tftp://192.168.1.xxx/tftpboot/lts/kernel.nb
>>>
>>> (I believe) would get an IP address from a DHCP server, and then
>>> (assuming the DHCP server didn't give a filename) would use the
>>> DEFAULT_BOOTFILE.
>     ...
>> He'd still be missing the root-path option that is required by LTSP
>> (assuming he's using LTSP).
>
> =22There's always somethin'=22... :)
>
>> Fortunately, it's possible to fix that in LTSP (Don't ya just love
>> open
>> source?)
>
> OK, I'm dying of suspense.  How do you fix this in LTSP?
> Inquiring minds, WanT tO KnOw...

You'd have to edit the linuxrc script in the initrd to accept a new optio=
n
for forcing root-path.

In fact, you'd have to figure out if DHCP is completely out of the
picture, or of it's still OK for the workstation to query dhcp for it's I=
P
address and netmask/broadcast/gateway info.

If you want DHCP completely out of the picture, you'd have to pass all of=

the information via the kernel cmdline, and then the linuxrc script would=

have to pick that up and NOT call dhclient.

So, I suppose we could add a couple of kernel cmdline options, like:

   ROOTPATH=3D<ipaddr and directory>
   NODHCP=3D=22ipaddr:netmask:broadcast:gateway:rootpath=22

So, you could either set ROOTPATH to the root-path value and use dhcp to
get the rest, OR, you could set NODHCP to contain everything.

of course, with Etherboot, you'd have to set option-128/129 to pass that
kernel cmdline.

Also, I think there's a limit to how long the kernel cmdline can be, but
i'm not sure if the limit is in the range that we should be worried about=
.

Also, the comments above about 'initrd' will all change with LTSP-4.2,
because we are switching over to initramfs.  But, the concepts are still
pretty much the same.

btw, we're on track to deliver LTSP-4.2 for LinuxWorld in Boston (Apr 3-6=
).

Jim.




>
> Marty
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the live
> webcast
> and join the prime developer group breaking into this new coding
> territory=21
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=
=3D121642
> _______________________________________________
> Etherboot-users mailing list
> Etherboot-users=40lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Marty C. <md...@et...>]

On Mar 9, 2006, at 2:30 PM, Jim McQuillan wrote:
> On Thu, March 9, 2006 2:14 pm, Marty Connor wrote:
>> Isn't there some way to use DEFAULT_BOOTFILE to get what you want
>> using Etherboot?
>>
>>      setting DEFAULT_BOOTFILE to:
>>
>>          tftp://192.168.1.xxx/tftpboot/lts/kernel.nb
>>
>> (I believe) would get an IP address from a DHCP server, and then
>> (assuming the DHCP server didn't give a filename) would use the
>> DEFAULT_BOOTFILE.
    ...
> He'd still be missing the root-path option that is required by LTSP
> (assuming he's using LTSP).

"There's always somethin'"... :)

> Fortunately, it's possible to fix that in LTSP (Don't ya just love  
> open
> source?)

OK, I'm dying of suspense.  How do you fix this in LTSP?
Inquiring minds, WanT tO KnOw...

Marty

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Jim M. <ja...@Mc...>]

On Thu, March 9, 2006 2:14 pm, Marty Connor wrote:
> On Mar 9, 2006, at 4:52 PM, lonnie=40outstep.com wrote:
>> When a boot client does a broadcast for a DHCP request, then there
>> could be more than 1 DHCP server responding and the client would
>> receive requests from all responding DHCP servers looking for the
>> one that accepted that particular client and gave it a lease, right?
>
> It's more elaborate than that.  PXE says the client should wait for
> up to 3 DHCP offers, and pick the one it likes the best (correct me
> if I'm wrong, PXE mavens).
>
>> Would this also mean that, in general, for all of the DHCP servers,
>> then if only one of the had the client MAC address in their list
>> then that would be the request that would have the valid
>> information (IP, gateway, etc....) for that client?
>> I guess what I am getting at is that it is probably ok for there to
>> be multiple DHCP servers on the net and subnets as long a only one
>> of them had the client address as an acceptable MAC address.
>
> It depends on if they're (the DHCP server) are =22authoritative=22 for
> the subnet or =22not authoritative=22 in ISC dhcpd.conf parlance.
> If they have the =22not authoritative=22 directive turned on, they don'=
t
> say anything. If they're authoritative, I believe they say =22no lease=22=

> if you don't have a pool of addresses defined.
>
>> The problem with changing ports for the DHCP server and Etherboot
>> is that there might already be other DHCP servers on those ports
>> without us bing aware of it by some chance and then we would still
>> have multiple requests beign fielded by multiple DHCP servers and
>> in which the problem is really not solved.
>
> We're looking at allowing people to put STATIC information in their
> Etherboot configuration so that no DHCP server is required to load a
> file.
> This means that someone could mess up and walk on an IP address on
> the LAN, but this is already possible on just about any device I know
> about.
> The consequences would be between you and your LAN administrator if
> you blow it.
>
> Isn't there some way to use DEFAULT_BOOTFILE to get what you want
> using Etherboot?
>
>      setting DEFAULT_BOOTFILE to:
>
>          tftp://192.168.1.xxx/tftpboot/lts/kernel.nb
>
> (I believe) would get an IP address from a DHCP server, and then
> (assuming the DHCP server didn't give a filename) would use the
> DEFAULT_BOOTFILE.
>
> I even found an old patch that would FORCE_BOOTFILE (although I would
> rename it FORCE_DEFAULT_BOOTFILE :) that ignores all the DHCP
> filenames and uses the one you supplied.
>
> This strikes me as a likely scenario.  You can get an IP, Router,
> DNS, etc, but you can't modify the DHCP server to do the filename
> thing, even though you have your own little tftp server (they're so
> much less objectionable than DHCP servers to LAN admins, after all).
>
> So you get everything from normal DHCP except DEFAULT_BOOTFILE, and
> you get your kernel from where you need to.
>
> Does this help?

He'd still be missing the root-path option that is required by LTSP
(assuming he's using LTSP).

Fortunately, it's possible to fix that in LTSP (Don't ya just love open
source?)

Jim.


>
> Marty
>
> --- etherboot-5.3.6/src/core/nic.c      2003-12-13 07:27:21.000000000
> +0100
> +++ etherboot-5.3.6.new/src/core/nic.c  2003-12-21 15:22:27.000000000
> +0100
> =40=40 -267,7 +267,11 =40=40
>          rpc_init();
>   =23endif
>   =23ifdef DEFAULT_BOOTFILE
> +=23ifdef FORCE_BOOTFILE
> +       kernel =3D DEFAULT_BOOTFILE;
> +=23else
>          kernel =3D KERNEL_BUF=5B0=5D =21=3D '=5C0' ? KERNEL_BUF : DEFA=
ULT_BOOTFILE;
> +=23endif /*FORCE_BOOTFILE */
>   =23else
>          kernel =3D KERNEL_BUF;
>   =23endif
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the live
> webcast
> and join the prime developer group breaking into this new coding
> territory=21
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=
=3D121642
> _______________________________________________
> Etherboot-users mailing list
> Etherboot-users=40lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Marty C. <md...@et...>]

On Mar 9, 2006, at 4:52 PM, lo...@ou... wrote:
> When a boot client does a broadcast for a DHCP request, then there  
> could be more than 1 DHCP server responding and the client would  
> receive requests from all responding DHCP servers looking for the  
> one that accepted that particular client and gave it a lease, right?

It's more elaborate than that.  PXE says the client should wait for  
up to 3 DHCP offers, and pick the one it likes the best (correct me  
if I'm wrong, PXE mavens).

> Would this also mean that, in general, for all of the DHCP servers,  
> then if only one of the had the client MAC address in their list  
> then that would be the request that would have the valid  
> information (IP, gateway, etc....) for that client?
> I guess what I am getting at is that it is probably ok for there to  
> be multiple DHCP servers on the net and subnets as long a only one  
> of them had the client address as an acceptable MAC address.

It depends on if they're (the DHCP server) are "authoritative" for  
the subnet or "not authoritative" in ISC dhcpd.conf parlance.
If they have the "not authoritative" directive turned on, they don't  
say anything. If they're authoritative, I believe they say "no lease"  
if you don't have a pool of addresses defined.

> The problem with changing ports for the DHCP server and Etherboot  
> is that there might already be other DHCP servers on those ports  
> without us bing aware of it by some chance and then we would still  
> have multiple requests beign fielded by multiple DHCP servers and  
> in which the problem is really not solved.

We're looking at allowing people to put STATIC information in their  
Etherboot configuration so that no DHCP server is required to load a  
file.
This means that someone could mess up and walk on an IP address on  
the LAN, but this is already possible on just about any device I know  
about.
The consequences would be between you and your LAN administrator if  
you blow it.

Isn't there some way to use DEFAULT_BOOTFILE to get what you want  
using Etherboot?

     setting DEFAULT_BOOTFILE to:

         tftp://192.168.1.xxx/tftpboot/lts/kernel.nb

(I believe) would get an IP address from a DHCP server, and then  
(assuming the DHCP server didn't give a filename) would use the  
DEFAULT_BOOTFILE.

I even found an old patch that would FORCE_BOOTFILE (although I would  
rename it FORCE_DEFAULT_BOOTFILE :) that ignores all the DHCP  
filenames and uses the one you supplied.

This strikes me as a likely scenario.  You can get an IP, Router,  
DNS, etc, but you can't modify the DHCP server to do the filename  
thing, even though you have your own little tftp server (they're so  
much less objectionable than DHCP servers to LAN admins, after all).

So you get everything from normal DHCP except DEFAULT_BOOTFILE, and  
you get your kernel from where you need to.

Does this help?

Marty

--- etherboot-5.3.6/src/core/nic.c      2003-12-13 07:27:21.000000000  
+0100
+++ etherboot-5.3.6.new/src/core/nic.c  2003-12-21 15:22:27.000000000  
+0100
@@ -267,7 +267,11 @@
         rpc_init();
  #endif
  #ifdef DEFAULT_BOOTFILE
+#ifdef FORCE_BOOTFILE
+       kernel = DEFAULT_BOOTFILE;
+#else
         kernel = KERNEL_BUF[0] != '\0' ? KERNEL_BUF : DEFAULT_BOOTFILE;
+#endif /*FORCE_BOOTFILE */
  #else
         kernel = KERNEL_BUF;
  #endif

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Peter <pl...@ac...>]

When a DHCP server has no data for a client it does not respond. 'No 
data' means the client is not in its SOA (or explicitly excluded by 
MAC).

Peter

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Jim M. <ja...@Mc...>]

lo...@ou... wrote:
> Hi Jim,
>
> I see what you are saying and this all looks like a very limiting 
> factor of DHCP.
>
> Let me ask you this, ok.
>
> When a boot client does a broadcast for a DHCP request, then there 
> could be more than 1 DHCP server responding and the client would 
> receive requests from all responding DHCP servers looking for the one 
> that accepted that particular client and gave it a lease, right?
>
> Would this also mean that, in general, for all of the DHCP servers, 
> then if only one of the had the client MAC address in their list then 
> that would be the request that would have the valid information (IP, 
> gateway, etc....) for that client?

Well, sort of.

Many people configure their dhcp servers to automatically hand out an IP 
address from a pool of addresses.

So, you could have a 2nd dhcp server that responds to the dhcp request 
BEFORE your ltsp server gets a chance to.

Then, the Etherboot or PXE bootrom would receive the offer, and possibly 
accept it.  I think recent versions of Etherboot will ignore any 
responses that don't have a 'filename' parameter, so as far as etherboot 
is concerned, you'd only accept a reply from a dhcp server properly 
configured for diskless booted.

BUT, after the kernel gets loaded in memory, and the 2nd DHCP request is 
sent out by the dhclient program inside the initrd, it doesn't make the 
distinction about the 'filename' parameter, and will accept an address 
from the first dhcp server that replies.  This could cause a problem, 
because the reply that it gets might not have a 'root-path' parameter, 
and without that, the workstation will just die with an error.

So, my advice is anytime you are thinking of putting a dhcp server on an 
existing network, you should make sure you know what other dhcp servers 
already exist, and if there are any, make sure they are configured to 
ignore requests from unknown clients, or they are configured to hand out 
the information that you need.

I realize it's not always easy to know that information, but if I was 
one of the guys responsible for one of the existing DHCP servers, i'd be 
very angry if somebody dropped another dhcp server on my network without 
informing me.

Many people just setup a separate lan segment for their LTSP thin 
clients that is completely separate from the existing network.  Also, a 
smart switch with VLAN capabilities can help to separate the network for 
you.

Jim McQuillan
ja...@Lt...



>
> I guess what I am getting at is that it is probably ok for there to be 
> multiple DHCP servers on the net and subnets as long a only one of 
> them had the client address as an acceptable MAC address.
>
> The problem with changing ports for the DHCP server and Etherboot is 
> that there might already be other DHCP servers on those ports without 
> us bing aware of it by some chance and then we would still have 
> multiple requests beign fielded by multiple DHCP servers and in which 
> the problem is really not solved.
>
> Thanks and have a good day,
>
> Lonnie T. Cumberland
> OutStep Technologies Incorporated
>
> Email: Lo...@ou...
>       Lon...@ya...
>
> Recommended sites:
>
>        http://www.peoplesquest.com
>
>
>
> Quoting Jim McQuillan <ja...@Mc...>:
>
>> Lonnie,
>>
>> While it's true that you can configure Etherboot to set the 
>> 'REQUIRE_VCI_ETHERBOOT' option, it won't do everything you think it 
>> will.
>>
>> The problem is, booting an LTSP workstation involves 2 DHCP requests. 
>>  The first one is from the bootrom (Etherboot OR PXE).  The 2nd DHCP 
>> request comes from dhclient inside the initrd that is downloaded 
>> along with the kernel.  That 2nd DHCP request won't do the 
>> 'REQUIRE_VCI_ETHERBOOT' option, it will simply make a DHCP request 
>> and accept the first response, no matter which dhcp server offers 
>> that response.  Then, it'll fail, if the wrong server offers it, 
>> because the wrong server probably isn't sending a 'root-path' value.
>>
>> You might instead try to set your LTSP dhcp to use ports 1067/1068, 
>> instead of the standard 67/68.  That way, the workstation will make a 
>> request, and the only DHCP server that will ever see the request is 
>> the LTSP dhcp server.
>>
>> To use ports 1067/1068, there are 3 places you'll have to make a change:
>>
>>  1)  The etherboot bootrom configuration
>>  2)  In the dhcpd.conf file, you'll have to set 'DPORT=1067' in an 
>> 'option-129' field.
>>  3)  You'll have to tell your dhcpd to ONLY listen on port 1067
>>
>> Take a look at:  
>> http://wiki.ltsp.org/twiki/bin/view/Ltsp/DHCP#Multiple_DHCP_servers_on_the_sam 
>>
>> for more information.
>>
>> ALSO, keep in mind, that using Ports 1067/1068 is ONLY an option for 
>> Etherboot.  You can't do this with PXE.
>>
>> Hope that helps,
>>
>> Jim McQuillan
>> ja...@Lt...
>>
>>
>>
>> Lonnie Cumberland wrote:
>>> Greetings All,
>>>
>>> I am trying to figure out how to set up a floppy boot image so that 
>>> I can boot a LTSP client but take advantage of possibly 
>>> "*REQUIRE_VCI_ETHERBOOT*" and not loose the MAC address as well.
>>>
>>> The problem is that we have multiple DHCP servers and need to the 
>>> client to ignore all requests except from a specific one that is 
>>> located on another subnet.
>>>
>>> I have been told that there is supposed to be a way for me to add in 
>>> an additional string of information into the floppy boot rom and 
>>> also on my DHCP server so that when the client nic receives a 
>>> response from the correct server then it will check for this string 
>>> as well.
>>>
>>> If this is possible, then I would like to be able to specify my own 
>>> string and not some default one like "Etherboot" which what I have 
>>> read seems to be in the documentation.
>>>
>>> Can anyone please tell me more about this?
>>>
>>> Thanks and have a good day,
>>>
>>> Lonnie T. Cumberland
>>> OutStep Technologies Incorporated
>>>
>>> Email: Lo...@ou...
>>>       Lon...@ya...
>>> Recommended sites:
>>>     http://www.peoplesquest.com
>>>
>>>
>>>
>>>
>>> -------------------------------------------------------
>>> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
>>> language
>>> that extends applications into web and mobile media. Attend the live 
>>> webcast
>>> and join the prime developer group breaking into this new coding 
>>> territory!
>>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 
>>>
>>> _______________________________________________
>>> Etherboot-users mailing list
>>> Eth...@li...
>>> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>>
>>
>> -------------------------------------------------------
>> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
>> language
>> that extends applications into web and mobile media. Attend the live 
>> webcast
>> and join the prime developer group breaking into this new coding 
>> territory!
>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
>> _______________________________________________
>> Etherboot-users mailing list
>> Eth...@li...
>> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
> language
> that extends applications into web and mobile media. Attend the live 
> webcast
> and join the prime developer group breaking into this new coding 
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> Etherboot-users mailing list
> Eth...@li...
> https://lists.sourceforge.net/lists/listinfo/etherboot-users

[Etherboot-users] Re: [Ltsp-discuss] Does something like wireless pxe exist?

[Jim M. <ja...@Mc...>]

Marty Connor wrote:

snipped ....
>
> Now, there may be another way to go.  There is a company that makes a 
> really cool adapter the plugs into an Ethernet port on a workstation, 
> and then does wireless.  Here's a link to that:
>
>     http://macsense.com/product/broadband/wua800.html
>
> With this device, I bet I could set up a thin client with Etherboot or 
> PXE, let it use its regular Ethernet port to boot from, and have this 
> box handle all the wireless stuff for me.  You can configure it with 
> SSIDs and various other parameters, and the client machine has no idea 
> what's going on.
>
> I'm so intrigued, I just ordered one to test.  With one of these, I 
> might be able to turn any thin client into a wireless thin client.
>

This has actually been done successfully by several people.  I've got a 
d-link travel AP that has the capability of being a wireless-bridge.  
Cost about $60 at Frys, and works quite well.  there's also something 
from Linksys called a 'gaming adapter' that also works well.  And, you 
can just take any old access point and use that in bridge mode, and it 
should work just fine.

Take a look at:

    
http://wiki.ltsp.org/twiki/bin/view/Ltsp/WirelessLTSPClientsUsingAnEthernetBridge


Still, i'm interested in hearing how the Macsense product works.


Thanks Marty and see you at Linux World,

Jim McQuillan
ja...@Lt...


> Anyway, I just wanted to put a few ideas out there.  I hope some of 
> this is interesting and useful and will stimulate some conversation.
>
> Marty
>
> --    Try: http://rom-o-matic.net/ to make Etherboot images instantly.
>
>    Name: Marty Connor
> US Mail: Entity Cyber, Inc.; P.O. Box 391827;
>          Cambridge, MA 02139; USA
>   Voice: (617) 491-6935; Fax: (617) 491-7046
>   Email: md...@et...
>     Web: http://www.etherboot.org/
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
> language
> that extends applications into web and mobile media. Attend the live 
> webcast
> and join the prime developer group breaking into this new coding 
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _____________________________________________________________________
> Ltsp-discuss mailing list.   To un-subscribe, or change prefs, goto:
>      https://lists.sourceforge.net/lists/listinfo/ltsp-discuss
> For additional LTSP help,   try #ltsp channel on irc.freenode.net

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[<lo...@ou...>]

Hi Jim,

I see what you are saying and this all looks like a very limiting 
factor of DHCP.

Let me ask you this, ok.

When a boot client does a broadcast for a DHCP request, then there 
could be more than 1 DHCP server responding and the client would 
receive requests from all responding DHCP servers looking for the one 
that accepted that particular client and gave it a lease, right?

Would this also mean that, in general, for all of the DHCP servers, 
then if only one of the had the client MAC address in their list then 
that would be the request that would have the valid information (IP, 
gateway, etc....) for that client?

I guess what I am getting at is that it is probably ok for there to be 
multiple DHCP servers on the net and subnets as long a only one of them 
had the client address as an acceptable MAC address.

The problem with changing ports for the DHCP server and Etherboot is 
that there might already be other DHCP servers on those ports without 
us bing aware of it by some chance and then we would still have 
multiple requests beign fielded by multiple DHCP servers and in which 
the problem is really not solved.

Thanks and have a good day,

Lonnie T. Cumberland
OutStep Technologies Incorporated

Email: Lo...@ou...
       Lon...@ya...

Recommended sites:

        http://www.peoplesquest.com



Quoting Jim McQuillan <ja...@Mc...>:

> Lonnie,
>
> While it's true that you can configure Etherboot to set the 
> 'REQUIRE_VCI_ETHERBOOT' option, it won't do everything you think it 
> will.
>
> The problem is, booting an LTSP workstation involves 2 DHCP requests. 
>  The first one is from the bootrom (Etherboot OR PXE).  The 2nd DHCP 
> request comes from dhclient inside the initrd that is downloaded 
> along with the kernel.  That 2nd DHCP request won't do the 
> 'REQUIRE_VCI_ETHERBOOT' option, it will simply make a DHCP request 
> and accept the first response, no matter which dhcp server offers 
> that response.  Then, it'll fail, if the wrong server offers it, 
> because the wrong server probably isn't sending a 'root-path' value.
>
> You might instead try to set your LTSP dhcp to use ports 1067/1068, 
> instead of the standard 67/68.  That way, the workstation will make a 
> request, and the only DHCP server that will ever see the request is 
> the LTSP dhcp server.
>
> To use ports 1067/1068, there are 3 places you'll have to make a change:
>
>  1)  The etherboot bootrom configuration
>  2)  In the dhcpd.conf file, you'll have to set 'DPORT=1067' in an 
> 'option-129' field.
>  3)  You'll have to tell your dhcpd to ONLY listen on port 1067
>
> Take a look at:  
> http://wiki.ltsp.org/twiki/bin/view/Ltsp/DHCP#Multiple_DHCP_servers_on_the_sam
> for more information.
>
> ALSO, keep in mind, that using Ports 1067/1068 is ONLY an option for 
> Etherboot.  You can't do this with PXE.
>
> Hope that helps,
>
> Jim McQuillan
> ja...@Lt...
>
>
>
> Lonnie Cumberland wrote:
>> Greetings All,
>>
>> I am trying to figure out how to set up a floppy boot image so that 
>> I can boot a LTSP client but take advantage of possibly 
>> "*REQUIRE_VCI_ETHERBOOT*" and not loose the MAC address as well.
>>
>> The problem is that we have multiple DHCP servers and need to the 
>> client to ignore all requests except from a specific one that is 
>> located on another subnet.
>>
>> I have been told that there is supposed to be a way for me to add in 
>> an additional string of information into the floppy boot rom and 
>> also on my DHCP server so that when the client nic receives a 
>> response from the correct server then it will check for this string 
>> as well.
>>
>> If this is possible, then I would like to be able to specify my own 
>> string and not some default one like "Etherboot" which what I have 
>> read seems to be in the documentation.
>>
>> Can anyone please tell me more about this?
>>
>> Thanks and have a good day,
>>
>> Lonnie T. Cumberland
>> OutStep Technologies Incorporated
>>
>> Email: Lo...@ou...
>>       Lon...@ya...
>> Recommended sites:
>>     http://www.peoplesquest.com
>>
>>
>>
>>
>> -------------------------------------------------------
>> This SF.Net email is sponsored by xPML, a groundbreaking scripting language
>> that extends applications into web and mobile media. Attend the live webcast
>> and join the prime developer group breaking into this new coding territory!
>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
>> _______________________________________________
>> Etherboot-users mailing list
>> Eth...@li...
>> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting language
> that extends applications into web and mobile media. Attend the live webcast
> and join the prime developer group breaking into this new coding territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> Etherboot-users mailing list
> Eth...@li...
> https://lists.sourceforge.net/lists/listinfo/etherboot-users
>

[Etherboot-users] Re: [Ltsp-discuss] Does something like wireless pxe exist?

[Marty C. <md...@et...>]

On Mar 8, 2006, at 4:18 PM, Richard Bos wrote:
> Op woensdag 8 maart 2006 21:32, schreef Evan Hisey:
>> The only real problem with a pxe wireless is that none of the card
>> makers have considered this a desired option.

Actually, it's more subtle than that.  For example, there is support  
in Etherboot (which has PXE support) for prism-based PCI cards to  
boot wirelessly.  I have demonstrated it at LinuxWorld in the past,  
and it's slow but it works.  We also don't have easy ways to specify  
SSIDs and various other parameters that would be desirable, but it is  
possible, using specific PCI cards to boot wirelessly.

>> Nothing real fancy is
>> need other than a way to set the acces point for initial connection.

Well, it's a bit messier than that.  See below

> thanks for you confirmation, it is what I thought already.  Should  
> we perhaps
> organize a coordinated effort to ask a network card manufacturer to  
> add
> wireless pxe support to 1 of their cards. ... ...

The problem is that most wireless cards are really PCMCIA cards, and  
whether you have a laptop or desktop, there is an extra interface  
called a PCI to PCMCIA bridge that must be initialized before the  
computer can talk to the wireless card.

The PXE driver in general has no idea which PCI to PCMCIA bridge is  
in the computer it happens to be in.  Have you ever noticed how late  
in booting Linux that PCMCIA support is enabled?  In order for  
wireless booting, the BIOS must get involved and cooperate with the  
PCMCIA card. It must initialize the PCMCIA to PCI bridge, and the  
driver has to know where to probe.  The PXE loader could do this, but  
it's not trivial to do, since you'd have to put PCMCIA support in the  
driver and probe.

Now, there is a company that sells PXE on a disk for various adapters:

     http://www.argontechnology.com/product.aspx/cid1/22

They also sell adapters:

     http://www.argontechnology.com/product.aspx/cid1/103

including a PCMCIA adapter that does PXE, so they must have limited  
PCMCIA support to find the PCMCIA card and initialize it.   
Unfortunately, it's not wireless.

Heck, they even sell Etherboot ROMS (clearly an old version without  
PXE support):

     http://www.argontechnology.com/product.aspx/cid1/102/cid2/65

The problem is that they don't seem to mention Etherboot is GPL'ed,  
and there is no obvious way to download the source for the code in  
the ROMs, so I think they may not be in compliance with the GPL, and  
since Etherboot supports PXE now, their $10 PXE on a disk things seem  
a lot more pricey...

Now, there may be another way to go.  There is a company that makes a  
really cool adapter the plugs into an Ethernet port on a workstation,  
and then does wireless.  Here's a link to that:

     http://macsense.com/product/broadband/wua800.html

With this device, I bet I could set up a thin client with Etherboot  
or PXE, let it use its regular Ethernet port to boot from, and have  
this box handle all the wireless stuff for me.  You can configure it  
with SSIDs and various other parameters, and the client machine has  
no idea what's going on.

I'm so intrigued, I just ordered one to test.  With one of these, I  
might be able to turn any thin client into a wireless thin client.

Anyway, I just wanted to put a few ideas out there.  I hope some of  
this is interesting and useful and will stimulate some conversation.

Marty

-- 
     Try: http://rom-o-matic.net/ to make Etherboot images instantly.

    Name: Marty Connor
US Mail: Entity Cyber, Inc.; P.O. Box 391827;
          Cambridge, MA 02139; USA
   Voice: (617) 491-6935; Fax: (617) 491-7046
   Email: md...@et...
     Web: http://www.etherboot.org/

Re: [Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Jim M. <ja...@Mc...>]

Lonnie,

While it's true that you can configure Etherboot to set the 
'REQUIRE_VCI_ETHERBOOT' option, it won't do everything you think it will.

The problem is, booting an LTSP workstation involves 2 DHCP requests.  
The first one is from the bootrom (Etherboot OR PXE).  The 2nd DHCP 
request comes from dhclient inside the initrd that is downloaded along 
with the kernel.  That 2nd DHCP request won't do the 
'REQUIRE_VCI_ETHERBOOT' option, it will simply make a DHCP request and 
accept the first response, no matter which dhcp server offers that 
response.  Then, it'll fail, if the wrong server offers it, because the 
wrong server probably isn't sending a 'root-path' value.

You might instead try to set your LTSP dhcp to use ports 1067/1068, 
instead of the standard 67/68.  That way, the workstation will make a 
request, and the only DHCP server that will ever see the request is the 
LTSP dhcp server.

To use ports 1067/1068, there are 3 places you'll have to make a change:

  1)  The etherboot bootrom configuration
  2)  In the dhcpd.conf file, you'll have to set 'DPORT=1067' in an 
'option-129' field.
  3)  You'll have to tell your dhcpd to ONLY listen on port 1067

Take a look at:  
http://wiki.ltsp.org/twiki/bin/view/Ltsp/DHCP#Multiple_DHCP_servers_on_the_sam
for more information.

ALSO, keep in mind, that using Ports 1067/1068 is ONLY an option for 
Etherboot.  You can't do this with PXE.

Hope that helps,

Jim McQuillan
ja...@Lt...



Lonnie Cumberland wrote:
> Greetings All,
>
> I am trying to figure out how to set up a floppy boot image so that I 
> can boot a LTSP client but take advantage of possibly 
> "*REQUIRE_VCI_ETHERBOOT*" and not loose the MAC address as well.
>
> The problem is that we have multiple DHCP servers and need to the 
> client to ignore all requests except from a specific one that is 
> located on another subnet.
>
> I have been told that there is supposed to be a way for me to add in 
> an additional string of information into the floppy boot rom and also 
> on my DHCP server so that when the client nic receives a response from 
> the correct server then it will check for this string as well.
>
> If this is possible, then I would like to be able to specify my own 
> string and not some default one like "Etherboot" which what I have 
> read seems to be in the documentation.
>
> Can anyone please tell me more about this?
>
> Thanks and have a good day,
>
> Lonnie T. Cumberland
> OutStep Technologies Incorporated
>
> Email: Lo...@ou...
>       Lon...@ya...
> Recommended sites:
>     http://www.peoplesquest.com
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
> language
> that extends applications into web and mobile media. Attend the live 
> webcast
> and join the prime developer group breaking into this new coding 
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> Etherboot-users mailing list
> Eth...@li...
> https://lists.sourceforge.net/lists/listinfo/etherboot-users

[Etherboot-users] REQUIRE_VCI_ETHERBOOT Question

[Lonnie C. <lo...@ou...>]

Greetings All,

I am trying to figure out how to set up a floppy boot image so that I 
can boot a LTSP client but take advantage of possibly 
"*REQUIRE_VCI_ETHERBOOT*" and not loose the MAC address as well.

The problem is that we have multiple DHCP servers and need to the client 
to ignore all requests except from a specific one that is located on 
another subnet.

I have been told that there is supposed to be a way for me to add in an 
additional string of information into the floppy boot rom and also on my 
DHCP server so that when the client nic receives a response from the 
correct server then it will check for this string as well.

If this is possible, then I would like to be able to specify my own 
string and not some default one like "Etherboot" which what I have read 
seems to be in the documentation.

Can anyone please tell me more about this?

Thanks and have a good day,

Lonnie T. Cumberland
OutStep Technologies Incorporated

Email: Lo...@ou...
       Lon...@ya... 

Recommended sites: 

	http://www.peoplesquest.com

Re: [Etherboot-users] oil immersed beowulf cluster.

[Marty C. <md...@et...>]

On Mar 8, 2006, at 2:39 AM, Cameron Bergh wrote:

> Pardon me if this has been asked before, but i have looked  
> everywhere and i
> have no idea how to write this .zhd file that the rom-o-matic has  
> given me.

It's nice of you to write.  The conversation is pleasant.  Here's how  
I found an answer:

Google

     etherboot .zhd

Check out the 2nd link:

     http://sourceforge.net/mailarchive/forum.php? 
forum_id=6402&max_rows=25&style=nested&viewmonth=200508

See that that relevant message there (the one with .zhd in it) quotes  
a message from the previous month:

     http://sourceforge.net/mailarchive/message.php?msg_id=12512726

This method erases a partition, and would thus not be suitable for  
all situations.  If you have a hard drive, and you have (or are  
willing to put) a filesystem on it, you may want to consider .zlilo  
files, and a kernel bootloader, such as grub or lilo.  For example:

Note that there are other ways to load Etherboot from a hard disk.   
One such way is to use a .zlilo file, and have grub or lilo load it.

Such as:

     http://www.wizzy.org.za/article/articlestatic/14/1/2/

or

     ftp://ftp.scyld.com/private/jlehan/pxe-on-a-disk.html

Let us know how things go.

Marty

> I am building an oil immersed beowulf cluster out of random old  
> computers
> with 256MB flash hard drives, I was hoping to have etherboot load  
> from the
> hard drives because some of the mobo's dont support PXE booting and  
> i dont
> want to go buy a bunch of EEPROMS to flash.
>
> what do i use to write a ZHD file? do these things even work?
>
> Thanks!
> -Cameron Bergh


-- 
     Try: http://rom-o-matic.net/ to make Etherboot images instantly.

    Name: Marty Connor
US Mail: Entity Cyber, Inc.; P.O. Box 391827;
          Cambridge, MA 02139; USA
   Voice: (617) 491-6935; Fax: (617) 491-7046
   Email: md...@et...
     Web: http://www.etherboot.org/

[Etherboot-users] oil immersed beowulf cluster.

[Cameron B. <can...@gm...>]

Pardon me if this has been asked before, but i have looked everywhere and i
have no idea how to write this .zhd file that the rom-o-matic has given me.

I am building an oil immersed beowulf cluster out of random old computers
with 256MB flash hard drives, I was hoping to have etherboot load from the
hard drives because some of the mobo's dont support PXE booting and i dont
want to go buy a bunch of EEPROMS to flash.

what do i use to write a ZHD file? do these things even work?

Thanks!
-Cameron Bergh