From gene.huft at wachovia.com Tue Apr 1 17:54:28 2008 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Tue, 1 Apr 2008 13:54:28 -0400 Subject: Problems with build ISO via Dell DRAC Message-ID: Gurus, I have created self contained ISO builds of RHEL 5.1 and RHEL 4.6....they work flawlesslessly when mapped to a VMware Virtual CDROM, or when burned to DVD media as bootable image for physical servers. However, the same ISO(s), when mapped to a Dell server's Remote Access Card virtual CDROM device, will boot - but it will not find the kickstart file. I have cracked open the initrd.img file, put the kickstart in there, repacked it & then it doesn't choke on not finding a ks.cfg file, but it doesn't use MINE! It uses some generic one. Anyway, has anyone come across this? Is it unique to DRAC, will I see same issue on HP's iLO? If so, why does VMware handle it beautifully ¬ them? Thanks in advance Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.j.meyers at boeing.com Tue Apr 1 18:17:19 2008 From: john.j.meyers at boeing.com (Meyers, John J) Date: Tue, 1 Apr 2008 11:17:19 -0700 Subject: Human interaction in %PRE text mode with snack Dialog boxes Message-ID: <8E40364C42ABEE4089507070AEB0C81B056F01AB@XCH-NW-8V2.nw.nos.boeing.com> Hi All, The %pre mode seems to conduct I/O with tty3. With primary anaconda user interaction displays on tty1. I need to interact with the user on tty1 during %pre; Using python and snack. Due to using various weird hardware in our environment a graphical install dos not work. The Code below is just an example of what I've tried. %pre --interpreter /usr/bin/python TTY = 1 import os import sys import snack import time import fcntl import termios os.system("chvt %d" % TTY) fd = os.open("/dev/tty%d" % TTY, os.O_RDWR) os.dup2(fd, 0) os.dup2(fd, 1) os.dup2(fd, 2) sys.stdin = os.fdopen(0, "r") sys.stdout = os.fdopen(1, "w") sys.stderr = os.fdopen(2, "w") screen = snack.SnackScreen() snack.ButtonChoiceWindow(screen, "WARNING", "WARNING! Am I CRAZY!", buttons=["reboot", "continue"]): os.system("reboot") while 1: time.sleep(10) screen.finish() os.system("chvt %d" % 3) %end ANY suggestions?? J. James Meyers Boeing Network/System Analyst & Linux PDM Cell: (206) 491-0315 -------------- next part -------------- An HTML attachment was scrubbed... URL: From eric at trueblade.com Tue Apr 1 18:36:57 2008 From: eric at trueblade.com (Eric Smith) Date: Tue, 01 Apr 2008 14:36:57 -0400 Subject: Human interaction in %PRE text mode with snack Dialog boxes In-Reply-To: <8E40364C42ABEE4089507070AEB0C81B056F01AB@XCH-NW-8V2.nw.nos.boeing.com> References: <8E40364C42ABEE4089507070AEB0C81B056F01AB@XCH-NW-8V2.nw.nos.boeing.com> Message-ID: <47F280C9.106@trueblade.com> Meyers, John J wrote: > > Hi All, > The %pre mode seems to conduct I/O with tty3. With primary > anaconda user interaction displays on tty1. I need to interact with the > user on tty1 during %pre; Using python and snack. Due to using various > weird hardware in our environment a graphical install dos not work. > The Code below is just an example of what I've tried. > > %pre --interpreter /usr/bin/python > TTY = 1 > import os > import sys > import snack > import time > import fcntl > import termios > os.system("chvt %d" % TTY) > fd = os.open("/dev/tty%d" % TTY, os.O_RDWR) > os.dup2(fd, 0) > os.dup2(fd, 1) > os.dup2(fd, 2) > sys.stdin = os.fdopen(0, "r") > sys.stdout = os.fdopen(1, "w") > sys.stderr = os.fdopen(2, "w") > screen = snack.SnackScreen() > snack.ButtonChoiceWindow(screen, "WARNING", > "WARNING! Am I CRAZY!", > buttons=["reboot", "continue"]): > os.system("reboot") > while 1: time.sleep(10) > screen.finish() > os.system("chvt %d" % 3) > %end > ANY suggestions?? I use code like this: set_tty(1) # change to tty1, we're called by # kickstart with stdout as tty3 user_args = get_user_input(host_config, args.args.get('tb-host')) set_tty(3) # restore Where: def set_tty(n): f = open('/dev/tty%d' % n, 'a') os.dup2(f.fileno(), sys.stdin.fileno()) os.dup2(f.fileno(), sys.stdout.fileno()) os.dup2(f.fileno(), sys.stderr.fileno()) and get_user_input() uses snack. Hope that helps. Eric. From jake at v2gnu.com Tue Apr 1 18:50:55 2008 From: jake at v2gnu.com (Jake Vickers) Date: Tue, 01 Apr 2008 14:50:55 -0400 Subject: Human interaction in %PRE text mode with snack Dialog boxes In-Reply-To: <8E40364C42ABEE4089507070AEB0C81B056F01AB@XCH-NW-8V2.nw.nos.boeing.com> References: <8E40364C42ABEE4089507070AEB0C81B056F01AB@XCH-NW-8V2.nw.nos.boeing.com> Message-ID: <47F2840F.3090501@v2gnu.com> Meyers, John J wrote: > Hi All, > The %pre mode seems to conduct I/O with tty3. With primary > anaconda user interaction displays on tty1. I need to interact with > the user on tty1 during %pre; Using python and snack. Due to using > various weird hardware in our environment a graphical install dos not > work. > The Code below is just an example of what I've tried. > Never tried python in the %pre myself, but couldn't a chvt work for you? ex: %pre chvt 3 # changes to TTY3 echo "" echo "Enter IP address" # or whatever you need to do read IP chvt 1 # change back to TTY1 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Eric.KENNEDYvanDAM at rvponp.fgov.be Thu Apr 3 08:47:33 2008 From: Eric.KENNEDYvanDAM at rvponp.fgov.be (KENNEDY VAN DAM Eric) Date: Thu, 03 Apr 2008 10:47:33 +0200 Subject: Creating a partition without a label Message-ID: <2F4F96E90349F241809293630A893B3D045D338C@MAIL03.rvponp.fgov.be> Hello everybody. I need to install some servers with a special driver which is not working with labels. The problem is that I did not find any way to change the default for the installer. In the /etc/fstab, I don't want to have: label=/boot but /dev/sda1 Can anyone help me about this ? Thanks. -- Eric Kennedy van Dam UNIX/Storage T?l.: +32 (0)2 529 33 75 Fax.: +32 (0)2 529 39 09 Mail: eric.kennedyvandam at onp.fgov.be Office national des Pensions Tour du Midi ~ B1e ?tage 1060 Bruxelles From Pablo.Iranzo at redhat.com Thu Apr 3 09:22:53 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Thu, 3 Apr 2008 11:22:53 +0200 (CEST) Subject: Creating a partition without a label In-Reply-To: <2F4F96E90349F241809293630A893B3D045D338C@MAIL03.rvponp.fgov.be> References: <2F4F96E90349F241809293630A893B3D045D338C@MAIL03.rvponp.fgov.be> Message-ID: Hi ??Can't you change it during %post? Just some "blkid's" and some seds over your /etc/fstab and you'll have a non-labelled one. Regards Pablo -- Pablo Iranzo G??mez (http://Alufis35.uv.es/~iranzo/) (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) -- Postulado de Boling sobre la Ley de Murphy: Si se encuentra bien, no se preocupe. Se le pasar?? On Thu, 3 Apr 2008, KENNEDY VAN DAM Eric wrote: > Hello everybody. > > I need to install some servers with a special driver which is not working with labels. > > The problem is that I did not find any way to change the default for the installer. > In the /etc/fstab, I don't want to have: > > label=/boot > > but > > /dev/sda1 > > Can anyone help me about this ? > > Thanks. > > -- > Eric Kennedy van Dam > UNIX/Storage > > > T?l.: +32 (0)2 529 33 75 > Fax.: +32 (0)2 529 39 09 > Mail: eric.kennedyvandam at onp.fgov.be > > Office national des Pensions > Tour du Midi ~ B1e ?tage > 1060 Bruxelles > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Eric.KENNEDYvanDAM at rvponp.fgov.be Thu Apr 3 09:30:00 2008 From: Eric.KENNEDYvanDAM at rvponp.fgov.be (KENNEDY VAN DAM Eric) Date: Thu, 03 Apr 2008 11:30:00 +0200 Subject: Creating a partition without a label References: <2F4F96E90349F241809293630A893B3D045D338C@MAIL03.rvponp.fgov.be> Message-ID: <2F4F96E90349F241809293630A893B3D045D3482@MAIL03.rvponp.fgov.be> > -----Message d'origine----- > De : kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] De la part de > Pablo Iranzo G?mez > Envoy? : jeudi 3 avril 2008 11:23 > ? : Discussion list about Kickstart > Objet : Re: Creating a partition without a label > > > Hi > ?Can't you change it during %post? > Just some "blkid's" and some seds over your /etc/fstab > and you'll > have a non-labelled one. > Regards > Pablo Yes I think I could make this in the post. But I just want to know if there is not a kickstart directive to make this. Thanks for your response. > Pablo Iranzo G?mez > (http://Alufis35.uv.es/~iranzo/) > (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) > -- > Postulado de Boling sobre la Ley de Murphy: > > Si se encuentra bien, no se preocupe. Se le pasar? > > On Thu, 3 Apr 2008, KENNEDY VAN DAM Eric wrote: > > > Hello everybody. > > > > I need to install some servers with a special driver which > is not working with labels. > > > > The problem is that I did not find any way to change the > default for the installer. > > In the /etc/fstab, I don't want to have: > > > > label=/boot > > > > but > > > > /dev/sda1 > > > > Can anyone help me about this ? > > > > Thanks. > > > > -- > > Eric Kennedy van Dam > > UNIX/Storage > > > > > > Tl.: +32 (0)2 529 33 75 > > Fax.: +32 (0)2 529 39 09 > > Mail: eric.kennedyvandam at onp.fgov.be > > > > > Office national des Pensions > > Tour du Midi ~ B1e tage > > 1060 Bruxelles > POS_HEX1.gif> > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Pablo.Iranzo at redhat.com Thu Apr 3 09:38:11 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Thu, 03 Apr 2008 09:38:11 +0000 Subject: Creating a partition without a label In-Reply-To: <2F4F96E90349F241809293630A893B3D045D3482@MAIL03.rvponp.fgov.be> References: <2F4F96E90349F241809293630A893B3D045D338C@MAIL03.rvponp.fgov.be> <2F4F96E90349F241809293630A893B3D045D3482@MAIL03.rvponp.fgov.be> Message-ID: <1207215491.9178.3.camel@iranzo.users.redhat.com> Hi El jue, 03-04-2008 a las 11:30 +0200, KENNEDY VAN DAM Eric escribi?: > Yes I think I could make this in the post. > But I just want to know if there is not a kickstart directive to make this. FYI: http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_Guide-en-US/s1-kickstart2-options.html This is what I use to know if something is included in major anaconda-based distros, but probably Fedora will have another options. Regards Pablo -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From stan.hearn at nscorp.com Sun Apr 6 19:28:58 2008 From: stan.hearn at nscorp.com (Hearn, Stan J.) Date: Sun, 06 Apr 2008 15:28:58 -0400 Subject: Package Confusion In-Reply-To: <47F0D31C.3090603@ng23.net> Message-ID: RHEL 5.1 x64 I'm trying to determine the minimum rpm's and install rpm groups I need for my installation. I've found a few things that are confusing. I've found a great deal of information about all the RH related distros, but nothing specifically that explains this in RHEL terms. o) I thought I could perform a manual install and during the install tediously pour over the rpm and group selections to choose what I want. Then, after the build, look in anaconda.ks and see which packages/groups where installed. That didn't seem to work. o) The group listed on RHN under the package search are not necessarily the same as an install package group install. For instance: on RHN ntp shows a group of "System Environment/Daemons." There is no such install group (is there?). I'm assuming the install group and the package group are different animals. o) I found where the repo contains the install groups in comps-rhel5-server-core.xml, however none of the xml files on the DVD include ntp and the ntp package is on the DVD? o) I'm trying to not install rpm's or rpm groups on a whim AND I'm trying not to specify each package individually in kickstart. o) How do I find what install package an rpm is in? Confused as usual, Stan From Gerrard.Geldenhuis at datacash.com Mon Apr 7 09:11:31 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 7 Apr 2008 10:11:31 +0100 Subject: Package Confusion In-Reply-To: References: <47F0D31C.3090603@ng23.net> Message-ID: I must say I find this confusing as well or at least not very well documented. I would be interested in an answer too. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Hearn, Stan J. > Sent: 06 April 2008 20:29 > To: Discussion list about Kickstart > Subject: Package Confusion > > RHEL 5.1 x64 > > I'm trying to determine the minimum rpm's and install rpm groups I need > for my installation. I've found a few things that are confusing. I've > found a great deal of information about all the RH related distros, but > nothing specifically that explains this in RHEL terms. > > o) I thought I could perform a manual install and during the install > tediously pour over the rpm and group selections to choose what I want. > Then, after the build, look in anaconda.ks and see which packages/groups > where installed. That didn't seem to work. > > o) The group listed on RHN under the package search are not necessarily > the same as an install package group install. > > For instance: on RHN ntp shows a group of "System Environment/Daemons." > There is no such install group (is there?). I'm assuming the install > group and the package group are different animals. > > o) I found where the repo contains the install groups in > comps-rhel5-server-core.xml, however none of the xml files on the DVD > include ntp and the ntp package is on the DVD? > > o) I'm trying to not install rpm's or rpm groups on a whim AND I'm > trying not to specify each package individually in kickstart. > > o) How do I find what install package an rpm is in? > > Confused as usual, > Stan > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From Chip.Shabazian at bankofamerica.com Mon Apr 7 17:14:43 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 07 Apr 2008 10:14:43 -0700 Subject: Package Confusion In-Reply-To: References: <47F0D31C.3090603@ng23.net> Message-ID: I've gotten a "minimum" install by removing rpm's until I was happy, then putting those removed rpm's in the %packages section (wrote a batch script to automate it). It was simply too much work for me to modify comps.xml and go through the genhd machinations when adding the rpm to %packages with a - in front of it worked the same. Doing this, I got an install with just under 200 rpm's, although I think that was RHEL 3. As for what packages are in what group, look at comps.xml -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Gerrard Geldenhuis Sent: Monday, April 07, 2008 2:12 AM To: Discussion list about Kickstart Subject: RE: Package Confusion I must say I find this confusing as well or at least not very well documented. I would be interested in an answer too. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Hearn, Stan J. > Sent: 06 April 2008 20:29 > To: Discussion list about Kickstart > Subject: Package Confusion > > RHEL 5.1 x64 > > I'm trying to determine the minimum rpm's and install rpm groups I need > for my installation. I've found a few things that are confusing. I've > found a great deal of information about all the RH related distros, but > nothing specifically that explains this in RHEL terms. > > o) I thought I could perform a manual install and during the install > tediously pour over the rpm and group selections to choose what I want. > Then, after the build, look in anaconda.ks and see which packages/groups > where installed. That didn't seem to work. > > o) The group listed on RHN under the package search are not necessarily > the same as an install package group install. > > For instance: on RHN ntp shows a group of "System Environment/Daemons." > There is no such install group (is there?). I'm assuming the install > group and the package group are different animals. > > o) I found where the repo contains the install groups in > comps-rhel5-server-core.xml, however none of the xml files on the DVD > include ntp and the ntp package is on the DVD? > > o) I'm trying to not install rpm's or rpm groups on a whim AND I'm > trying not to specify each package individually in kickstart. > > o) How do I find what install package an rpm is in? > > Confused as usual, > Stan > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From Gerrard.Geldenhuis at datacash.com Tue Apr 8 09:09:37 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Tue, 8 Apr 2008 10:09:37 +0100 Subject: pxe menu Message-ID: Hi I have been looking through the documentation to see if you can have sub menus in the pxe boot menu but could not find anything that talks specifically about it. I could probably adapt the function key menus but would prefer to use the normal menus. Does anyone have an example I could use or point me to the right documentation please. Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From Pablo.Iranzo at redhat.com Tue Apr 8 09:29:49 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Tue, 08 Apr 2008 11:29:49 +0200 Subject: pxe menu In-Reply-To: References: Message-ID: <1207646989.10238.6.camel@iranzo.users.redhat.com> Have a look at pxeboot (from syslinux project) and their menu.c32 ;) Regards Pablo El mar, 08-04-2008 a las 10:09 +0100, Gerrard Geldenhuis escribi?: > Hi > > I have been looking through the documentation to see if you can have > sub menus in the pxe boot menu but could not find anything that talks > specifically about it. I could probably adapt the function key menus > but would prefer to use the normal menus. Does anyone have an example > I could use or point me to the right documentation please. > > > > Regards > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From mdehaan at redhat.com Tue Apr 8 15:18:00 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 11:18:00 -0400 Subject: pxe menu In-Reply-To: <1207646989.10238.6.camel@iranzo.users.redhat.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> Message-ID: <47FB8CA8.1080800@redhat.com> Pablo Iranzo G?mez wrote: > Have a look at pxeboot (from syslinux project) and their menu.c32 ;) > > Regards > Pablo > Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate it's PXE menus. menu.c32 is of course very nice. I haven't found a good way to generate submenus in though I would be very interested in this -- for instance, if you were able to pick your cobbler distribution and then pick profiles based on that distribution, or if the lists were very long it might be more useful than having scrolling. However since menu.c32 does offer some very nice scrolling, unless your number of entries gets very large, the need for submenus may be a non-issue. Another problem you'll run into with menu.c32 is that it does not offer a working timeout -- so once you PXE to engage the menu, the user must choose from the menu... so in my case I default to "default", and menu.c32 is the "menu" target -- so to pull up the menu you have to type "menu" at the syslinux prompt before it times out to "default". Hopefully that helps, if you get a good working hierarchical setup with menu.c32 I'd be interested in seeing your configuration files on that. --Michael > El mar, 08-04-2008 a las 10:09 +0100, Gerrard Geldenhuis escribi?: > >> Hi >> >> I have been looking through the documentation to see if you can have >> sub menus in the pxe boot menu but could not find anything that talks >> specifically about it. I could probably adapt the function key menus >> but would prefer to use the normal menus. Does anyone have an example >> I could use or point me to the right documentation please. >> >> >> >> Regards >> >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list From Gerrard.Geldenhuis at datacash.com Tue Apr 8 16:06:44 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Tue, 8 Apr 2008 17:06:44 +0100 Subject: pxe menu In-Reply-To: <47FB8CA8.1080800@redhat.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> Message-ID: Just to be sure, menu.c32 is a module that extends pxe boot capabilities? Is it build into the syslinux package or do I need to download it. I could not find anythng relevant on a RH AS 4 server which is my satellite box. I will gladly pass on anything if I get it working properly. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: 08 April 2008 16:18 > To: Discussion list about Kickstart > Subject: Re: pxe menu > > Pablo Iranzo G?mez wrote: > > Have a look at pxeboot (from syslinux project) and their menu.c32 ;) > > > > Regards > > Pablo > > > > Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate it's > PXE menus. menu.c32 is of course very nice. I haven't found a good > way to generate submenus in though I would be very interested in this -- > for instance, if you were able to pick your cobbler distribution and > then pick profiles based on that distribution, or if the lists were very > long it might be more useful than having scrolling. > > However since menu.c32 does offer some very nice scrolling, unless your > number of entries gets very large, the need for submenus may be a non- > issue. > > Another problem you'll run into with menu.c32 is that it does not offer > a working timeout -- so once you PXE to engage the menu, the user must > choose from the menu... so in my case I default to "default", and > menu.c32 is the "menu" target -- so to pull up the menu you have to type > "menu" at the syslinux prompt before it times out to "default". > > Hopefully that helps, if you get a good working hierarchical setup with > menu.c32 I'd be interested in seeing your configuration files on that. > > --Michael > > El mar, 08-04-2008 a las 10:09 +0100, Gerrard Geldenhuis escribi?: > > > >> Hi > >> > >> I have been looking through the documentation to see if you can have > >> sub menus in the pxe boot menu but could not find anything that talks > >> specifically about it. I could probably adapt the function key menus > >> but would prefer to use the normal menus. Does anyone have an example > >> I could use or point me to the right documentation please. > >> > >> > >> > >> Regards > >> > >> > >> _______________________________________________ > >> Kickstart-list mailing list > >> Kickstart-list at redhat.com > >> https://www.redhat.com/mailman/listinfo/kickstart-list > >> > >> ----------------------------------------------------------------------- > - > >> > >> _______________________________________________ > >> Kickstart-list mailing list > >> Kickstart-list at redhat.com > >> https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From mdehaan at redhat.com Tue Apr 8 16:28:29 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 12:28:29 -0400 Subject: pxe menu In-Reply-To: References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> Message-ID: <47FB9D2D.8080408@redhat.com> Gerrard Geldenhuis wrote: > Just to be sure, menu.c32 is a module that extends pxe boot capabilities? > It's in the package. For Cobbler I package the menu.c32 from a Fedora version, however, because the version packaged with RHEL4 has issues (I forget details -- it's been a while). If running RHEL 5 as the PXE server this should not be a problem. > Is it build into the syslinux package or do I need to download it. I could not find anythng relevant on a RH AS 4 server which is my satellite box. > It's a seperate file. If you install Cobbler and create some profiles it will build the config files out in /tftpboot for you to look at and play with. > I will gladly pass on anything if I get it working properly. > > Regards > > >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- >> bounces at redhat.com] On Behalf Of Michael DeHaan >> Sent: 08 April 2008 16:18 >> To: Discussion list about Kickstart >> Subject: Re: pxe menu >> >> Pablo Iranzo G?mez wrote: >> >>> Have a look at pxeboot (from syslinux project) and their menu.c32 ;) >>> >>> Regards >>> Pablo >>> >>> >> Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate it's >> PXE menus. menu.c32 is of course very nice. I haven't found a good >> way to generate submenus in though I would be very interested in this -- >> for instance, if you were able to pick your cobbler distribution and >> then pick profiles based on that distribution, or if the lists were very >> long it might be more useful than having scrolling. >> >> However since menu.c32 does offer some very nice scrolling, unless your >> number of entries gets very large, the need for submenus may be a non- >> issue. >> >> Another problem you'll run into with menu.c32 is that it does not offer >> a working timeout -- so once you PXE to engage the menu, the user must >> choose from the menu... so in my case I default to "default", and >> menu.c32 is the "menu" target -- so to pull up the menu you have to type >> "menu" at the syslinux prompt before it times out to "default". >> >> Hopefully that helps, if you get a good working hierarchical setup with >> menu.c32 I'd be interested in seeing your configuration files on that. >> >> --Michael >> >>> El mar, 08-04-2008 a las 10:09 +0100, Gerrard Geldenhuis escribi?: >>> >>> >>>> Hi >>>> >>>> I have been looking through the documentation to see if you can have >>>> sub menus in the pxe boot menu but could not find anything that talks >>>> specifically about it. I could probably adapt the function key menus >>>> but would prefer to use the normal menus. Does anyone have an example >>>> I could use or point me to the right documentation please. >>>> >>>> >>>> >>>> Regards >>>> >>>> >>>> _______________________________________________ >>>> Kickstart-list mailing list >>>> Kickstart-list at redhat.com >>>> https://www.redhat.com/mailman/listinfo/kickstart-list >>>> >>>> ----------------------------------------------------------------------- >>>> >> - >> >>>> _______________________________________________ >>>> Kickstart-list mailing list >>>> Kickstart-list at redhat.com >>>> https://www.redhat.com/mailman/listinfo/kickstart-list >>>> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From dwight at supercomputer.org Tue Apr 8 16:10:45 2008 From: dwight at supercomputer.org (dwight at supercomputer.org) Date: Tue, 8 Apr 2008 08:10:45 -0800 Subject: pxe menu In-Reply-To: References: <47FB8CA8.1080800@redhat.com> Message-ID: <200804080910.45511.dwight@supercomputer.org> It's been a while since I've built it, so I don't recall the specifics. I was under the impression that it was something that you simply enabled. But as I said, it's been a while, so I could be wrong. The c32 stuff is excellent software and is very handy. Here's a sample of how to use the menuing features in your /tftpboot/pxelinux.cfg/default file: LABEL RescueImages MENU LABEL Rescue Images KERNEL menu.c32 APPEND pxelinux.cfg/RescueImages And you simply create a file called RescueImages, which contains the normal entries for booting the particular rescue image that you want. This allows you to have submenus for the rescue images. Something which is very handy when you have to support multiple architectures and multiple versions/distros, and a particular system is giving you trouble. -dwight- On Tuesday 08 April 2008 09:06:44 am Gerrard Geldenhuis wrote: > Just to be sure, menu.c32 is a module that extends pxe boot > capabilities? > > Is it build into the syslinux package or do I need to download it. > I could not find anythng relevant on a RH AS 4 server which is my > satellite box. > > I will gladly pass on anything if I get it working properly. > > Regards > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > > bounces at redhat.com] On Behalf Of Michael DeHaan > > Sent: 08 April 2008 16:18 > > To: Discussion list about Kickstart > > Subject: Re: pxe menu > > > > Pablo Iranzo G?mez wrote: > > > Have a look at pxeboot (from syslinux project) and their > > > menu.c32 ;) > > > > > > Regards > > > Pablo > > > > Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate > > it's PXE menus. menu.c32 is of course very nice. I haven't > > found a good way to generate submenus in though I would be very > > interested in this -- for instance, if you were able to pick > > your cobbler distribution and then pick profiles based on that > > distribution, or if the lists were very long it might be more > > useful than having scrolling. > > > > However since menu.c32 does offer some very nice scrolling, > > unless your number of entries gets very large, the need for > > submenus may be a non- issue. > > > > Another problem you'll run into with menu.c32 is that it does > > not offer a working timeout -- so once you PXE to engage the > > menu, the user must choose from the menu... so in my case I > > default to "default", and menu.c32 is the "menu" target -- so to > > pull up the menu you have to type "menu" at the syslinux prompt > > before it times out to "default". > > > > Hopefully that helps, if you get a good working hierarchical > > setup with menu.c32 I'd be interested in seeing your > > configuration files on that. > > > > --Michael > > > > > El mar, 08-04-2008 a las 10:09 +0100, Gerrard Geldenhuis escribi?: > > >> Hi > > >> > > >> I have been looking through the documentation to see if you > > >> can have sub menus in the pxe boot menu but could not find > > >> anything that talks specifically about it. I could probably > > >> adapt the function key menus but would prefer to use the > > >> normal menus. Does anyone have an example I could use or > > >> point me to the right documentation please. > > >> > > >> > > >> > > >> Regards > > >> > > >> > > >> _______________________________________________ > > >> Kickstart-list mailing list > > >> Kickstart-list at redhat.com > > >> https://www.redhat.com/mailman/listinfo/kickstart-list > > >> > > >> ------------------------------------------------------------- > > >>---------- > > > > - > > > > >> _______________________________________________ > > >> Kickstart-list mailing list > > >> Kickstart-list at redhat.com > > >> https://www.redhat.com/mailman/listinfo/kickstart-list > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From elliot at rpath.com Tue Apr 8 18:32:00 2008 From: elliot at rpath.com (Elliot Peele) Date: Tue, 8 Apr 2008 14:32:00 -0400 Subject: pxe menu In-Reply-To: <47FB8CA8.1080800@redhat.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> Message-ID: <20080408183200.GB18954@rpath.com> On Tue, Apr 08, 2008 at 11:18:00AM -0400, Michael DeHaan wrote: > Pablo Iranzo G?mez wrote: > > Have a look at pxeboot (from syslinux project) and their menu.c32 ;) Also might want to take a look at vesamenu.c32. > Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate it's > PXE menus. menu.c32 is of course very nice. I haven't found a good > way to generate submenus in though I would be very interested in this -- > for instance, if you were able to pick your cobbler distribution and > then pick profiles based on that distribution, or if the lists were very > long it might be more useful than having scrolling. As of syslinux 3.54 sub menu support was added. The following example will create a top level menu option of "rPath Linux" with two sub levels of "rPath Linux 1.0.7" and "rPath Linux 2 Beta 2". Example: menu begin rpl menu label rPath Linux label - menu label rPath Linux 1.0.7 menu disabled label 1.0.7 menu indent 1 kernel 1.0.7/vmlinuz append initrd=1.0.7/initrd.img label 1.0.7-x86_64 menu indent 1 kernel 1.0.7-x86_64/vmlinuz append initrd=1.0.7-x86_64/initrd.img menu separator label - menu label rPath Linux 2 Beta 2 menu disabled label rpl2-beta2 menu indent 1 kernel rpl2-beta2/vmlinuz append initrd=rpl2-beta2/initrd.img menu separator menu end > Another problem you'll run into with menu.c32 is that it does not offer > a working timeout -- so once you PXE to engage the menu, the user must > choose from the menu... so in my case I default to "default", and > menu.c32 is the "menu" target -- so to pull up the menu you have to type > "menu" at the syslinux prompt before it times out to "default". Not sure what version the timeout issue was fixed in, but it works in 3.62. Elliot -- Elliot Peele rPath, Inc. elliot at rpath.com From elliot at bentlogic.net Tue Apr 8 17:28:36 2008 From: elliot at bentlogic.net (Elliot Peele) Date: Tue, 8 Apr 2008 13:28:36 -0400 Subject: pxe menu In-Reply-To: <47FB8CA8.1080800@redhat.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> Message-ID: <20080408172836.GB8141@bentlogic.net> On Tue, Apr 08, 2008 at 11:18:00AM -0400, Michael DeHaan wrote: > Pablo Iranzo G?mez wrote: > > Have a look at pxeboot (from syslinux project) and their menu.c32 ;) Also might want to take a look at vesamenu.c32. > Cobbler (http://cobbler.et.redhat.com) uses menu.c32 to generate it's > PXE menus. menu.c32 is of course very nice. I haven't found a good > way to generate submenus in though I would be very interested in this -- > for instance, if you were able to pick your cobbler distribution and > then pick profiles based on that distribution, or if the lists were very > long it might be more useful than having scrolling. As of syslinux 3.54 sub menu support was added. The following example will create a top level menu option of "rPath Linux" with two sub levels of "rPath Linux 1.0.7" and "rPath Linux 2 Beta 2". Example: menu begin rpl menu label rPath Linux label - menu label rPath Linux 1.0.7 menu disabled label 1.0.7 menu indent 1 kernel 1.0.7/vmlinuz append initrd=1.0.7/initrd.img label 1.0.7-x86_64 menu indent 1 kernel 1.0.7-x86_64/vmlinuz append initrd=1.0.7-x86_64/initrd.img menu separator label - menu label rPath Linux 2 Beta 2 menu disabled label rpl2-beta2 menu indent 1 kernel rpl2-beta2/vmlinuz append initrd=rpl2-beta2/initrd.img menu separator menu end > Another problem you'll run into with menu.c32 is that it does not offer > a working timeout -- so once you PXE to engage the menu, the user must > choose from the menu... so in my case I default to "default", and > menu.c32 is the "menu" target -- so to pull up the menu you have to type > "menu" at the syslinux prompt before it times out to "default". Not sure what version the timeout issue was fixed in, but it works in 3.62. Elliot -- Elliot Peele elliot at bentlogic.net From terry.mcintyre at gmail.com Tue Apr 8 18:37:32 2008 From: terry.mcintyre at gmail.com (Terry McIntyre) Date: Tue, 8 Apr 2008 11:37:32 -0700 Subject: can't CVS pykickstart Message-ID: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> Per instructions at http://fedoraproject.org/wiki/pykickstart, I did the following: export CVSROOT=:pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS cvs co pykickstart cvs server: cannot find module `pykickstart' - ignored cvs [checkout aborted]: cannot expand modules echo $CVSROOT :pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS Am I missing something? Thanks! -- Terry McIntyre UNIX for hire Software Development, Systems Administration, Security terry.mcintyre at gmail.com From mdehaan at redhat.com Tue Apr 8 18:44:43 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 14:44:43 -0400 Subject: pxe menu In-Reply-To: <20080408183200.GB18954@rpath.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> <20080408183200.GB18954@rpath.com> Message-ID: <47FBBD1B.3050200@redhat.com> > Not sure what version the timeout issue was fixed in, but it works in > 3.62. > > Elliot > > Good to know. I'll see if I can take this up with the maintainer to get newer versions in. Thanks! --Michael From mdehaan at redhat.com Tue Apr 8 18:49:04 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 14:49:04 -0400 Subject: pxe menu In-Reply-To: <47FBBD1B.3050200@redhat.com> References: <1207646989.10238.6.camel@iranzo.users.redhat.com> <47FB8CA8.1080800@redhat.com> <20080408183200.GB18954@rpath.com> <47FBBD1B.3050200@redhat.com> Message-ID: <47FBBE20.1010305@redhat.com> Michael DeHaan wrote: > >> Not sure what version the timeout issue was fixed in, but it works in >> 3.62. >> >> Elliot >> > > Good to know. I'll see if I can take this up with the maintainer to > get newer versions in. Thanks! > > --Michael > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list I'm going to test subpackaging a newer menu.c32 with a cobbler today, and if that works with an older syslinux, then we can have things popping up to the regular menus by default -- (currently I subpackage menu.c32 to ensure older OS versions can still get nice menus). --Michael From elliot at rpath.com Tue Apr 8 18:58:09 2008 From: elliot at rpath.com (Elliot Peele) Date: Tue, 8 Apr 2008 14:58:09 -0400 Subject: can't CVS pykickstart In-Reply-To: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> References: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> Message-ID: <20080408185809.GC18954@rpath.com> On Tue, Apr 08, 2008 at 11:37:32AM -0700, Terry McIntyre wrote: > Per instructions at http://fedoraproject.org/wiki/pykickstart, I did > the following: > > export CVSROOT=:pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS > > cvs co pykickstart > > cvs server: cannot find module `pykickstart' - ignored > cvs [checkout aborted]: cannot expand modules > > echo $CVSROOT > :pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS > > Am I missing something? pykickstart moved to git, the wiki page is just out of date. git://git.fedorahosted.org/pykickstart.git Elliot -- Elliot Peele rPath, Inc. elliot at rpath.com From terry.mcintyre at gmail.com Tue Apr 8 19:05:40 2008 From: terry.mcintyre at gmail.com (Terry McIntyre) Date: Tue, 8 Apr 2008 12:05:40 -0700 Subject: can't CVS pykickstart In-Reply-To: <20080408185809.GC18954@rpath.com> References: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> <20080408185809.GC18954@rpath.com> Message-ID: <67f096ea0804081205s20dcd4a4k58abe478881a3ba9@mail.gmail.com> On Tue, Apr 8, 2008 at 11:58 AM, Elliot Peele wrote: > On Tue, Apr 08, 2008 at 11:37:32AM -0700, Terry McIntyre wrote: > > Per instructions at http://fedoraproject.org/wiki/pykickstart, I did > > the following: > > > > export CVSROOT=:pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS > > > > cvs co pykickstart > > > > cvs server: cannot find module `pykickstart' - ignored > > cvs [checkout aborted]: cannot expand modules > > > > echo $CVSROOT > > :pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS > > > > Am I missing something? > > pykickstart moved to git, the wiki page is just out of date. > > git://git.fedorahosted.org/pykickstart.git > Thanks, Elliot I can spell "git", but am not yet familiar with it. What's the quick-and-dirty way to use git to grab pykickstart? Thanks! From mdehaan at redhat.com Tue Apr 8 19:04:58 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 15:04:58 -0400 Subject: can't CVS pykickstart In-Reply-To: <67f096ea0804081205s20dcd4a4k58abe478881a3ba9@mail.gmail.com> References: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> <20080408185809.GC18954@rpath.com> <67f096ea0804081205s20dcd4a4k58abe478881a3ba9@mail.gmail.com> Message-ID: <47FBC1DA.1040008@redhat.com> Terry McIntyre wrote: > On Tue, Apr 8, 2008 at 11:58 AM, Elliot Peele wrote: > >> On Tue, Apr 08, 2008 at 11:37:32AM -0700, Terry McIntyre wrote: >> > Per instructions at http://fedoraproject.org/wiki/pykickstart, I did >> > the following: >> > >> > export CVSROOT=:pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS >> > >> > cvs co pykickstart >> > >> > cvs server: cannot find module `pykickstart' - ignored >> > cvs [checkout aborted]: cannot expand modules >> > >> > echo $CVSROOT >> > :pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS >> > >> > Am I missing something? >> >> pykickstart moved to git, the wiki page is just out of date. >> >> git://git.fedorahosted.org/pykickstart.git >> >> > Thanks, Elliot > > I can spell "git", but am not yet familiar with it. > > What's the quick-and-dirty way to use git to grab pykickstart? > > Thanks! > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > yum install git-core (or possibly just "git" if you want everything) cd ~/ git clone git://git.fedorahosted.org/pykickstart.git cd ~/pykickstart From terry.mcintyre at gmail.com Tue Apr 8 21:36:51 2008 From: terry.mcintyre at gmail.com (Terry McIntyre) Date: Tue, 8 Apr 2008 14:36:51 -0700 Subject: ks_validator error? Message-ID: <67f096ea0804081436v1a127df3w17b6d5748fcc2d83@mail.gmail.com> When I validate my kickstart file using ks_validator, I get this generic message: [root at admin cobbler]# ksvalidator /var/www/cobbler/kickstarts_sys/00:15:17:3a:c2:78/ks.cfg General error in input file: 'str' object has no attribute 'partition' Is this something I should worry about? Any notion what I should look for? Thanks for any info. -- Terry McIntyre UNIX for hire Software Development, Systems Administration, Security terry.mcintyre at gmail.com From mdehaan at redhat.com Tue Apr 8 21:54:26 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 17:54:26 -0400 Subject: ks_validator error? In-Reply-To: <67f096ea0804081436v1a127df3w17b6d5748fcc2d83@mail.gmail.com> References: <67f096ea0804081436v1a127df3w17b6d5748fcc2d83@mail.gmail.com> Message-ID: <47FBE992.5070905@redhat.com> Terry McIntyre wrote: > When I validate my kickstart file using ks_validator, I get this > generic message: > > > [root at admin cobbler]# ksvalidator > /var/www/cobbler/kickstarts_sys/00:15:17:3a:c2:78/ks.cfg > General error in input file: 'str' object has no attribute 'partition' > > Is this something I should worry about? > > Any notion what I should look for? > > Thanks for any info. > > The partition method is apparently new to Python 2.5, so that is probably the problem. You probably want an older version of pykickstart that matches your distro? --Michael From mdehaan at redhat.com Tue Apr 8 21:57:26 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 08 Apr 2008 17:57:26 -0400 Subject: can't CVS pykickstart In-Reply-To: <47FBC1DA.1040008@redhat.com> References: <67f096ea0804081137q232ef6bn37ba080eb35da5e9@mail.gmail.com> <20080408185809.GC18954@rpath.com> <67f096ea0804081205s20dcd4a4k58abe478881a3ba9@mail.gmail.com> <47FBC1DA.1040008@redhat.com> Message-ID: <47FBEA46.5080301@redhat.com> Michael DeHaan wrote: > Terry McIntyre wrote: >> On Tue, Apr 8, 2008 at 11:58 AM, Elliot Peele wrote: >> >>> On Tue, Apr 08, 2008 at 11:37:32AM -0700, Terry McIntyre wrote: >>> > Per instructions at http://fedoraproject.org/wiki/pykickstart, I did >>> > the following: >>> > >>> > export CVSROOT=:pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS >>> > >>> > cvs co pykickstart >>> > >>> > cvs server: cannot find module `pykickstart' - ignored >>> > cvs [checkout aborted]: cannot expand modules >>> > >>> > echo $CVSROOT >>> > :pserver:anonymous at rhlinux.redhat.com:/usr/local/CVS >>> > >>> > Am I missing something? >>> >>> pykickstart moved to git, the wiki page is just out of date. >>> >>> git://git.fedorahosted.org/pykickstart.git >>> >>> >> Thanks, Elliot >> >> I can spell "git", but am not yet familiar with it. >> >> What's the quick-and-dirty way to use git to grab pykickstart? >> >> Thanks! >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > > yum install git-core (or possibly just "git" if you want everything) > > cd ~/ > > git clone git://git.fedorahosted.org/pykickstart.git > > cd ~/pykickstart > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list Confirmed, just replacing menu.c32 with a version from the latest syslinux snapshot is all that is required. (No need to update syslinux itself). Excellent -- I'll include this in the Cobbler 0.9.x branch :) --Michael From debian at herakles.homelinux.org Tue Apr 8 23:44:20 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Wed, 09 Apr 2008 07:44:20 +0800 Subject: pxe menu In-Reply-To: References: Message-ID: <47FC0354.9050602@herakles.homelinux.org> Gerrard Geldenhuis wrote: > Hi > > I have been looking through the documentation to see if you can have sub > menus in the pxe boot menu but could not find anything that talks > specifically about it. I could probably adapt the function key menus but > would prefer to use the normal menus. Does anyone have an example I > could use or point me to the right documentation please. The etherboot project now has a PXE "ROM" that can be chained from pxelinux and have a different menu. It might go part way. It also has other features worth investigating. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From robin-lists at robinbowes.com Wed Apr 9 18:32:11 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Wed, 09 Apr 2008 19:32:11 +0100 Subject: What does kssendmac set? Message-ID: Hi, I have a cgi script that I use to generate kickstart files for various hosts. I use it something like this: ... ks=http://example.com/cgi/ks?host=foobar.example.com I'd like to use the machine mac address to determine the host configuration to be used and have just discovered the "kssendmac" directive. Can someone tell me what additional information that directive sets and how I should look for it in my CGI? Thanks, R. From elliot at rpath.com Wed Apr 9 19:30:38 2008 From: elliot at rpath.com (Elliot Peele) Date: Wed, 9 Apr 2008 15:30:38 -0400 Subject: What does kssendmac set? In-Reply-To: References: Message-ID: <20080409193038.GA13727@rpath.com> On Wed, Apr 09, 2008 at 07:32:11PM +0100, Robin Bowes wrote: > I'd like to use the machine mac address to determine the host > configuration to be used and have just discovered the "kssendmac" > directive. > > Can someone tell me what additional information that directive sets and > how I should look for it in my CGI? Looks like it sends a http header in the form: X-RHN-Provisioning-MAC-$interfaceIndex: $deviceName $macAddress Elliot -- Elliot Peele rPath, Inc. elliot at rpath.com From mdehaan at redhat.com Wed Apr 9 20:33:35 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 09 Apr 2008 16:33:35 -0400 Subject: What does kssendmac set? In-Reply-To: <20080409193038.GA13727@rpath.com> References: <20080409193038.GA13727@rpath.com> Message-ID: <47FD281F.2070906@redhat.com> Elliot Peele wrote: > On Wed, Apr 09, 2008 at 07:32:11PM +0100, Robin Bowes wrote: > >> I'd like to use the machine mac address to determine the host >> configuration to be used and have just discovered the "kssendmac" >> directive. >> >> Can someone tell me what additional information that directive sets and >> how I should look for it in my CGI? >> > > Looks like it sends a http header in the form: > > X-RHN-Provisioning-MAC-$interfaceIndex: $deviceName $macAddress > > > Elliot > > Robin, not sure if you just asked this in #kickstart, but here's some example code of reading that from Python: http://git.fedorahosted.org/git/?p=cobbler;a=blob_plain;f=scripts/findks.cgi;hb=HEAD For those that aren't there, #kickstart is on irc.freenode.net and is the IRC companion to this channel. Join up, as it could use some more people in it. --Michael From robin-lists at robinbowes.com Thu Apr 10 00:26:40 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Thu, 10 Apr 2008 01:26:40 +0100 Subject: What does kssendmac set? In-Reply-To: <47FD281F.2070906@redhat.com> References: <20080409193038.GA13727@rpath.com> <47FD281F.2070906@redhat.com> Message-ID: Michael DeHaan wrote: > Robin, not sure if you just asked this in #kickstart, I did indeed - nice talking to you. I'll be back... :) R. From mad at vorsicht-bissig.de Thu Apr 10 06:46:13 2008 From: mad at vorsicht-bissig.de (mad at vorsicht-bissig.de) Date: Thu, 10 Apr 2008 08:46:13 +0200 Subject: Create custom boot image for cobbler pxe boot Message-ID: <20080410064613.143000@gmx.net> Hi! I have a desktop PC I want to install with cobbler (Fedora Core 5) via network. The problem is, that the e1000 network card is to new for the pxeboot kernel from the FC5 CD. The cobbler boot runs without problems until it wants to download the kickstart file from the cobbler server. The error message is, that it can't find a network device. I already compiled a new e1000 module on an installed FC5 PC. The module works fine. Now how can I use the new module? I tried to update the initrd.img from my /tftpboot directory and that worked like described in http://www.sharktooth.de/doku.php/linux:fedora:fc5_e1000_hack But this is a very dirty hack (in my opinion). Is there a 'right way' to create a new vmlinuz and initrd.img on the already installed PC that I can use for tftpboot? Or is there another possibilty to use the new e1000 module? Thanks in advance mad From rbrown at BallistaSec.com Thu Apr 10 13:05:10 2008 From: rbrown at BallistaSec.com (Rodrick Brown) Date: Thu, 10 Apr 2008 09:05:10 -0400 Subject: Create custom boot image for cobbler pxe boot In-Reply-To: <20080410064613.143000@gmx.net> References: <20080410064613.143000@gmx.net> Message-ID: Make your own initrd with the newer e1000 driver the modules are located in modules.tgz Question why do you insist on using such an old version of fc? -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of mad at vorsicht-bissig.de Sent: Thursday, April 10, 2008 2:46 AM To: kickstart-list at redhat.com Subject: Create custom boot image for cobbler pxe boot Hi! I have a desktop PC I want to install with cobbler (Fedora Core 5) via network. The problem is, that the e1000 network card is to new for the pxeboot kernel from the FC5 CD. The cobbler boot runs without problems until it wants to download the kickstart file from the cobbler server. The error message is, that it can't find a network device. I already compiled a new e1000 module on an installed FC5 PC. The module works fine. Now how can I use the new module? I tried to update the initrd.img from my /tftpboot directory and that worked like described in http://www.sharktooth.de/doku.php/linux:fedora:fc5_e1000_hack But this is a very dirty hack (in my opinion). Is there a 'right way' to create a new vmlinuz and initrd.img on the already installed PC that I can use for tftpboot? Or is there another possibilty to use the new e1000 module? Thanks in advance mad _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From pgroven at 2wire.com Thu Apr 10 23:45:14 2008 From: pgroven at 2wire.com (Phillip Groven) Date: Thu, 10 Apr 2008 16:45:14 -0700 Subject: oracle Red Hat kickstart Message-ID: <299E8BB0D5475E4FABB162E7DCA1BBE902387759@phxexch01.corp.2wire.com> Does anyone have a sample ks.cfg to automate an oracle install using Red Hat kickstart? We are using oracle 10G R1 I have a working kickstart but I am not sure how to automate it for oracle Does anyone know of a web resource link? -------------- next part -------------- An HTML attachment was scrubbed... URL: From hhoffman at ip-solutions.net Fri Apr 11 00:12:43 2008 From: hhoffman at ip-solutions.net (Harry Hoffman) Date: Thu, 10 Apr 2008 20:12:43 -0400 Subject: oracle Red Hat kickstart In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE902387759@phxexch01.corp.2wire.com> References: <299E8BB0D5475E4FABB162E7DCA1BBE902387759@phxexch01.corp.2wire.com> Message-ID: <47FEACFB.6050107@ip-solutions.net> Here you go... this uses cobbler for provisioning and it's for CentOS4.x (let me know if you need 5.x). Let me know if you have any questions. If the lists filters attachments it'll be here shortly http://www.ip-solutions.net/oracle/cobbler-oracle.tar.gz Cheers, Harry Phillip Groven wrote: > * * > > * * > > *Does anyone have a sample ks.cfg to automate an oracle install using > Red Hat kickstart?* > > *We are using oracle 10G R1* > > *I have a working kickstart but I am not sure how to automate it for oracle* > > *Does anyone know of a web resource link?* > > * * > > * * > > * * > > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -------------- next part -------------- A non-text attachment was scrubbed... Name: cobbler-oracle.tar.gz Type: application/x-gzip Size: 7631 bytes Desc: not available URL: From CallahanT at tessco.com Fri Apr 11 15:20:20 2008 From: CallahanT at tessco.com (Callahan, Tom) Date: Fri, 11 Apr 2008 11:20:20 -0400 Subject: oracle Red Hat kickstart In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE902387759@phxexch01.corp.2wire.com> Message-ID: Define automate? I have kickstarts that setup the systems will all the pre-reqs required to be installed for Oracle 10g RAC, but we have our DBA do the installation of the Oracle software itself. On 4/10/08 7:45 PM, "Phillip Groven" wrote: > > > Does anyone have a sample ks.cfg to automate an oracle install using Red Hat > kickstart? > We are using oracle 10G R1 > I have a working kickstart but I am not sure how to automate it for oracle > Does anyone know of a web resource link? > > > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -------------- next part -------------- An HTML attachment was scrubbed... URL: From Aleksandar.Lazic at netavis.net Thu Apr 17 08:45:29 2008 From: Aleksandar.Lazic at netavis.net (Lazic Aleksandar) Date: Thu, 17 Apr 2008 10:45:29 +0200 Subject: Error with kickstart and raid Message-ID: <30B7EA1EBAB8274BA6A2BDA3C16FF9D115D6E2@vieneu.universe.intra> Hi all, we use a custom layout in ks preinstall script --- clearpart --all --initlabel part raid.01 --ondisk hdb --size=... --asprimary part raid.02 --ondisk hdb --size=... --asprimary part raid.03 --ondisk hdb --size=... part raid.04 --ondisk hdb --size=... part swap --ondisk hdb --size part /data/data1 --ondisk hdb --asprimary --size 1 --grow --fstype=ext3 part raid.11 --ondisk hdc --size=... --asprimary part raid.12 --ondisk hdc --size=... --asprimary part raid.13 --ondisk hdc --size=... part raid.14 --ondisk hdc --size=... part swap --ondisk hdc --size part /data/data2 --ondisk hdc --asprimary --size 1 --grow --fstype=ext3 raid /boot --level=1 --device=md0 --fstype=ext3 raid.01 raid.11 raid / --level=1 --device=md1 --fstype=ext3 raid.02 raid.12 raid /netavis --level=1 --device=md2 --fstype=ext3 raid.03 raid.13 raid /var/lib/mysql --level=1 --device=md3 --fstype=ext3 raid.04 raid.14 --- After a test installation we want to make the final installation but we get always the following error: ValueError: md1 is already in mdList. This is on a CentOS 5.1 installation. I have try to add a: os.system("mdadm -S /dev/md*") But this wasn't th solution ;-(. I have also try with --useexisting but in this stage there aren't any md's?! When I make a 'mdadm --detail -s' then I see only md1 not md0 on the tty2. Please can anybody help me :-) The screenshots of the output are available under: http://solutions.netavis.net/Part_error01.jpg http://solutions.netavis.net/Part_error02.jpg http://solutions.netavis.net/Part_error03.jpg BR Aleks From jgranado at redhat.com Thu Apr 17 09:02:50 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Thu, 17 Apr 2008 11:02:50 +0200 Subject: Error with kickstart and raid In-Reply-To: <30B7EA1EBAB8274BA6A2BDA3C16FF9D115D6E2@vieneu.universe.intra> References: <30B7EA1EBAB8274BA6A2BDA3C16FF9D115D6E2@vieneu.universe.intra> Message-ID: <4807123A.2050702@redhat.com> Lazic Aleksandar wrote: > Hi all, > > we use a custom layout in ks preinstall script > > --- > clearpart --all --initlabel > part raid.01 --ondisk hdb --size=... --asprimary part raid.02 --ondisk > hdb --size=... --asprimary part raid.03 --ondisk hdb --size=... > part raid.04 --ondisk hdb --size=... > part swap --ondisk hdb --size > part /data/data1 --ondisk hdb --asprimary --size 1 --grow --fstype=ext3 > part raid.11 --ondisk hdc --size=... --asprimary part raid.12 --ondisk > hdc --size=... --asprimary part raid.13 --ondisk hdc --size=... > part raid.14 --ondisk hdc --size=... > part swap --ondisk hdc --size > part /data/data2 --ondisk hdc --asprimary --size 1 --grow --fstype=ext3 > raid /boot --level=1 --device=md0 --fstype=ext3 raid.01 raid.11 > raid / --level=1 --device=md1 --fstype=ext3 raid.02 raid.12 > raid /netavis --level=1 --device=md2 --fstype=ext3 raid.03 raid.13 > raid /var/lib/mysql --level=1 --device=md3 --fstype=ext3 raid.04 raid.14 > --- > > After a test installation we want to make the final installation but we > get always the following error: > > ValueError: md1 is already in mdList. > > This is on a CentOS 5.1 installation. > I have try to add a: > > os.system("mdadm -S /dev/md*") > > But this wasn't th solution ;-(. > > I have also try with --useexisting but in this stage there aren't any > md's?! > > When I make a 'mdadm --detail -s' then I see only md1 not md0 on the > tty2. > > Please can anybody help me :-) > > The screenshots of the output are available under: > > http://solutions.netavis.net/Part_error01.jpg > http://solutions.netavis.net/Part_error02.jpg > http://solutions.netavis.net/Part_error03.jpg > > BR > > Aleks > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list This is a known bug, it happens when you are trying to install over a system that had raid in the past. This issue is already solved in RHEL5.2. My advice to you is to format the HD before installation. With the previously installed system, or put a %pre section in the kickstart that does that for you. Regards -- Joel Andres Granados Red Hat / Brno, Czech Republic From Pablo.Iranzo at redhat.com Thu Apr 17 09:05:53 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Thu, 17 Apr 2008 11:05:53 +0200 Subject: Error with kickstart and raid In-Reply-To: <4807123A.2050702@redhat.com> References: <30B7EA1EBAB8274BA6A2BDA3C16FF9D115D6E2@vieneu.universe.intra> <4807123A.2050702@redhat.com> Message-ID: <1208423153.17672.4.camel@iranzo.users.redhat.com> I've experienced something similar with existing lvm layouts... dd is your friend ;) Regards Pablo El jue, 17-04-2008 a las 11:02 +0200, Joel Andres Granados escribi?: > Lazic Aleksandar wrote: > > Hi all, > > > > we use a custom layout in ks preinstall script > > > > --- > > clearpart --all --initlabel > > part raid.01 --ondisk hdb --size=... --asprimary part raid.02 --ondisk > > hdb --size=... --asprimary part raid.03 --ondisk hdb --size=... > > part raid.04 --ondisk hdb --size=... > > part swap --ondisk hdb --size > > part /data/data1 --ondisk hdb --asprimary --size 1 --grow --fstype=ext3 > > part raid.11 --ondisk hdc --size=... --asprimary part raid.12 --ondisk > > hdc --size=... --asprimary part raid.13 --ondisk hdc --size=... > > part raid.14 --ondisk hdc --size=... > > part swap --ondisk hdc --size > > part /data/data2 --ondisk hdc --asprimary --size 1 --grow --fstype=ext3 > > raid /boot --level=1 --device=md0 --fstype=ext3 raid.01 raid.11 > > raid / --level=1 --device=md1 --fstype=ext3 raid.02 raid.12 > > raid /netavis --level=1 --device=md2 --fstype=ext3 raid.03 raid.13 > > raid /var/lib/mysql --level=1 --device=md3 --fstype=ext3 raid.04 raid.14 > > --- > > > > After a test installation we want to make the final installation but we > > get always the following error: > > > > ValueError: md1 is already in mdList. > > > > This is on a CentOS 5.1 installation. > > I have try to add a: > > > > os.system("mdadm -S /dev/md*") > > > > But this wasn't th solution ;-(. > > > > I have also try with --useexisting but in this stage there aren't any > > md's?! > > > > When I make a 'mdadm --detail -s' then I see only md1 not md0 on the > > tty2. > > > > Please can anybody help me :-) > > > > The screenshots of the output are available under: > > > > http://solutions.netavis.net/Part_error01.jpg > > http://solutions.netavis.net/Part_error02.jpg > > http://solutions.netavis.net/Part_error03.jpg > > > > BR > > > > Aleks > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > This is a known bug, it happens when you are trying to install over a system that had raid in the past. This issue is already solved in RHEL5.2. My advice to you is to format the HD before installation. With the previously installed system, or put a %pre section in the kickstart that does that for you. > > Regards > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Aleksandar.Lazic at netavis.net Thu Apr 17 09:47:00 2008 From: Aleksandar.Lazic at netavis.net (Lazic Aleksandar) Date: Thu, 17 Apr 2008 11:47:00 +0200 Subject: Error with kickstart and raid Message-ID: <30B7EA1EBAB8274BA6A2BDA3C16FF9D115D6E7@vieneu.universe.intra> Joel Andres Granados wrotes on Thu., April 17, 2008 11:03 AM >Lazic Aleksandar wrote: >> Hi all, >> >> we use a custom layout in ks preinstall script [snipp] >> After a test installation we want to make the final installation but >> we get always the following error: >> >> ValueError: md1 is already in mdList. This is a known bug, it happens when you are trying to install over a system that had raid in the past. This issue is already solved in RHEL5.2. My advice to you is to format the HD before installation. With the previously installed system, or put a %pre section in the kickstart that does that for you. Thank you ;-) Cheers Aleks From alst74 at gmail.com Fri Apr 18 12:05:28 2008 From: alst74 at gmail.com (=?ISO-8859-1?Q?Alexander_St=E5hlberg?=) Date: Fri, 18 Apr 2008 14:05:28 +0200 Subject: creating boot image Message-ID: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> Couldn't find a nice guide through google, so I try here. I'd like to create a boot image, where I already on the image itself point out my nfs server and kickstart file so I don't have to write the boot options everytime (linux ks=nfs:myserver:/somedir/ks.cfg), always using the same nfs server for installation. I then mount the image in my ilo (hp machine) as a usb or cdrom image. In short, trying to perform a diskless/usbless installation with preconfigured boot options. kind regards alex From jgranado at redhat.com Fri Apr 18 13:50:53 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Fri, 18 Apr 2008 15:50:53 +0200 Subject: creating boot image In-Reply-To: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> References: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> Message-ID: <4808A73D.6090107@redhat.com> Alexander St?hlberg wrote: > Couldn't find a nice guide through google, so I try here. > > I'd like to create a boot image, where I already on the image itself > point out my nfs server and kickstart file so I don't have to write > the boot options everytime (linux ks=nfs:myserver:/somedir/ks.cfg), > always using the same nfs server for installation. > I then mount the image in my ilo (hp machine) as a usb or cdrom image. > > In short, trying to perform a diskless/usbless installation with > preconfigured boot options. > > kind regards > alex > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list I'm assuming you have this problem because you have a large amount of machines that you need to install with certain regularity. In this case I would really suggest PXE, it is extremely configurable and can accomplish what you want. Cobbler is also a good alternative (I've been told, haven't actually used it). I think the boot image approach, though possible, is too much work. Regards -- Joel Andres Granados Red Hat / Brno, Czech Republic From rainer at ultra-secure.de Fri Apr 18 14:38:59 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Fri, 18 Apr 2008 16:38:59 +0200 Subject: creating boot image In-Reply-To: <4808A73D.6090107@redhat.com> References: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> <4808A73D.6090107@redhat.com> Message-ID: <4808B283.6070209@ultra-secure.de> Joel Andres Granados schrieb: > > I'm assuming you have this problem because you have a large amount of > machines that you need to install with certain regularity. In this > case I would really suggest PXE, it is extremely configurable and can > accomplish what you want. Cobbler is also a good alternative (I've > been told, haven't actually used it). > Yup. Cobbler is the way to go. > I think the boot image approach, though possible, is too much work. > Yes, probably. I thought of setting up a PXE+TFTP-server and the rest of the infrastructure by hand, but after I found cobbler, I never looked back. In fact, it's the tool I have been looking for for years (OK, I'm actually looking for something that can provision Linux, Windows, FreeBSD and Solaris - but cobbler is a good start ;-))) ) Rainer From debian at herakles.homelinux.org Fri Apr 18 21:37:37 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Sat, 19 Apr 2008 05:37:37 +0800 Subject: creating boot image In-Reply-To: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> References: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> Message-ID: <480914A1.1030902@herakles.homelinux.org> ? wrote: > Couldn't find a nice guide through google, so I try here. > > I'd like to create a boot image, where I already on the image itself > point out my nfs server and kickstart file so I don't have to write > the boot options everytime (linux ks=nfs:myserver:/somedir/ks.cfg), > always using the same nfs server for installation. > I then mount the image in my ilo (hp machine) as a usb or cdrom image. > > In short, trying to perform a diskless/usbless installation with > preconfigured boot options. I've seen the others' replies. I have done it, with CentOS5. In my case, the network connection to my install server wasn't as good as my Internet connexion, but I wanted to ks. I don't recall now wither I used the Internet, or installed from DVD. Probably the former, the machine I used doesn't have a DVD drive, but I did set up transparent proxy (and cache). I don't remember all the commands, but it would have been something like this: read all the relevant docs including how to specify where the ks file is. In my case, I decided to put the ks file on one of my servers so I didn't have to burn another CD or DVD to change it. Unpack the ISO: mount -o loop,ro bf.iso /mnt/iso cp -pr /mnt/iso /tmp/iso find /mnt/iso -name TRANS.TBL -exec rm -fv {} \; ensure it's writable chmod -r +w /tmp/iso edit the isolinux configuration to specify the ks location vim /tmp/iso/... put the ks file in place whatever create a new iso. mkisofs Burn to rewritable media. Important this, one chap I know of used a whole box of CDs before he thought of this. One of cdrecord growisofs Note to Anaconda hackers, if any are listening. Is there a good reason Anaconda shouldn't just have a peek at local floppy, USB and maybe optical media to see whether there's a ks file to be found, and use it? -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From markmc at redhat.com Mon Apr 21 07:57:59 2008 From: markmc at redhat.com (Mark McLoughlin) Date: Mon, 21 Apr 2008 08:57:59 +0100 Subject: [PATCH] Initialize cmd.handler earlier; fixes repo.methodToRepo() Message-ID: <1208764679-4816-1-git-send-email-markmc@redhat.com> FC8_Repo.methodToRepo() references self.handler.method which, since a commands "handler" attribute is only initialized by BaseHandler.dispatcher(), means that a "repo" command must have been parsed before methodToRepo() is called. Clearly this is an invalid assumption - a kickstart may not contain any "repo" commands if e.g. it only has a "url" command. Fix this by making a commands "handler" attribute be initialized as soon as the command is registered during the BaseHandler constructor. Signed-off-by: Mark McLoughlin --- pykickstart/base.py | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/pykickstart/base.py b/pykickstart/base.py index 54ce86d..26e8d58 100644 --- a/pykickstart/base.py +++ b/pykickstart/base.py @@ -302,6 +302,7 @@ class BaseHandler: # Finally, add the mapping to the commands dict. self.commands[cmdName] = cmdObj + self.commands[cmdName].handler = self # We also need to create attributes for the various data objects. # No checks here because dMap is a bijection. At least, that's what @@ -323,7 +324,6 @@ class BaseHandler: elif self.commands[cmd] != None: self.commands[cmd].currentCmd = cmd self.commands[cmd].currentLine = self.currentLine - self.commands[cmd].handler = self self.commands[cmd].lineno = lineno self.commands[cmd].parse(args[1:]) -- 1.5.4.5 From jgranado at redhat.com Mon Apr 21 08:22:12 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Mon, 21 Apr 2008 10:22:12 +0200 Subject: creating boot image In-Reply-To: <480914A1.1030902@herakles.homelinux.org> References: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> <480914A1.1030902@herakles.homelinux.org> Message-ID: <480C4EB4.5040607@redhat.com> John Summerfield wrote: > ? wrote: >> Couldn't find a nice guide through google, so I try here. >> >> I'd like to create a boot image, where I already on the image itself >> point out my nfs server and kickstart file so I don't have to write >> the boot options everytime (linux ks=nfs:myserver:/somedir/ks.cfg), >> always using the same nfs server for installation. >> I then mount the image in my ilo (hp machine) as a usb or cdrom image. >> >> In short, trying to perform a diskless/usbless installation with >> preconfigured boot options. > > I've seen the others' replies. > I have done it, with CentOS5. In my case, the network connection to my > install server wasn't as good as my Internet connexion, but I wanted to > ks. I don't recall now wither I used the Internet, or installed from > DVD. Probably the former, the machine I used doesn't have a DVD drive, > but I did set up transparent proxy (and cache). > > I don't remember all the commands, but it would have been something like > this: > read all the relevant docs including how to specify where the ks file > is. In my case, I decided to put the ks file on one of my servers so I > didn't have to burn another CD or DVD to change it. > > Unpack the ISO: > mount -o loop,ro bf.iso /mnt/iso > cp -pr /mnt/iso /tmp/iso > find /mnt/iso -name TRANS.TBL -exec rm -fv {} \; > ensure it's writable > chmod -r +w /tmp/iso > edit the isolinux configuration to specify the ks location > vim /tmp/iso/... > put the ks file in place > whatever > create a new iso. > mkisofs too, so read anaconda if you can't find it with google, I think Chris > Kloiber had a script....> > Burn to rewritable media. Important this, one chap I know of used a > whole box of CDs before he thought of this. > One of > cdrecord > growisofs > > Note to Anaconda hackers, if any are listening. Is there a good reason > Anaconda shouldn't just have a peek at local floppy, USB and maybe > optical media to see whether there's a ks file to be found, and use it? > > > You can tell anaconda where to look in the cmdline. from http://fedoraproject.org/wiki/Anaconda/Options " ks * Kickstart over NFS. ks=cdrom: * Kickstart from CDROM ks=file: * Kickstart from a file (path = 'fd0/ks.cfg') ks=ftp:// * Kickstart from FTP. ks=hd:: * Kickstart via harddrive (dev = 'hda1', for example) ks=http:// * Kickstart from HTTP. ks=nfs(:options):: * Kickstart from NFS. NFS mount options are optional. " -- Joel Andres Granados Red Hat / Brno, Czech Republic From debian at herakles.homelinux.org Mon Apr 21 12:22:55 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Mon, 21 Apr 2008 20:22:55 +0800 Subject: creating boot image In-Reply-To: <480C4EB4.5040607@redhat.com> References: <89abee5a0804180505i7572d6a9la1bd5b7510be5eb7@mail.gmail.com> <480914A1.1030902@herakles.homelinux.org> <480C4EB4.5040607@redhat.com> Message-ID: <480C871F.708@herakles.homelinux.org> Joel Andres Granados wrote: > John Summerfield wrote: >> ? wrote: >>> Couldn't find a nice guide through google, so I try here. >>> >>> I'd like to create a boot image, where I already on the image itself >>> point out my nfs server and kickstart file so I don't have to write >>> the boot options everytime (linux ks=nfs:myserver:/somedir/ks.cfg), >>> always using the same nfs server for installation. >>> I then mount the image in my ilo (hp machine) as a usb or cdrom image. >>> >>> In short, trying to perform a diskless/usbless installation with >>> preconfigured boot options. >> >> I've seen the others' replies. >> I have done it, with CentOS5. In my case, the network connection to my >> install server wasn't as good as my Internet connexion, but I wanted >> to ks. I don't recall now wither I used the Internet, or installed >> from DVD. Probably the former, the machine I used doesn't have a DVD >> drive, but I did set up transparent proxy (and cache). >> >> I don't remember all the commands, but it would have been something >> like this: >> read all the relevant docs including how to specify where the ks file >> is. In my case, I decided to put the ks file on one of my servers so I >> didn't have to burn another CD or DVD to change it. >> >> Unpack the ISO: >> mount -o loop,ro bf.iso /mnt/iso >> cp -pr /mnt/iso /tmp/iso >> find /mnt/iso -name TRANS.TBL -exec rm -fv {} \; >> ensure it's writable >> chmod -r +w /tmp/iso >> edit the isolinux configuration to specify the ks location >> vim /tmp/iso/... >> put the ks file in place >> whatever >> create a new iso. >> mkisofs > anaconda too, so read anaconda if you can't find it with google, I >> think Chris Kloiber had a script....> >> Burn to rewritable media. Important this, one chap I know of used a >> whole box of CDs before he thought of this. >> One of >> cdrecord >> growisofs >> >> Note to Anaconda hackers, if any are listening. Is there a good reason >> Anaconda shouldn't just have a peek at local floppy, USB and maybe >> optical media to see whether there's a ks file to be found, and use it? >> >> >> > > You can tell anaconda where to look in the cmdline. from > http://fedoraproject.org/wiki/Anaconda/Options I know, but setting that as a default requires fiddling with the ISO as I just described. It would be handy if, as I said, it just has a look around. In comparison, it's what Windows XP does (at least, it looks for a floppy). -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From Gerrard.Geldenhuis at datacash.com Mon Apr 21 15:48:00 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 21 Apr 2008 16:48:00 +0100 Subject: Debugging kickstart general advice sought Message-ID: Hi I am debugging a problem with my kickstart but am founding it immensely cumbersome to rebuild a machine everytime just to test the last part of the install a stubborn pre-post script that does not want to work. I' will probably ask a question about it here later on. Can anyone suggest a more productive way of debugging kickstart files that takes up less time than actually building a machine each time? Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From Gerrard.Geldenhuis at datacash.com Mon Apr 21 15:55:47 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 21 Apr 2008 16:55:47 +0100 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: I use sleep a lot and () > /mnt/sysimage/root/nochroot-post.log 2>&1 but apart from that there must be an easier way. Regards ________________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Gerrard Geldenhuis Sent: 21 April 2008 16:48 To: Discussion list about Kickstart Subject: Debugging kickstart general advice sought Hi I am debugging a problem with my kickstart but am founding it immensely cumbersome to rebuild a machine everytime just to test the last part of the install a stubborn pre-post script that does not want to work. I' will probably ask a question about it here later on. Can anyone suggest a more productive way of debugging kickstart files that takes up less time than actually building a machine each time? Regards From jake at v2gnu.com Mon Apr 21 16:06:37 2008 From: jake at v2gnu.com (Jake Vickers) Date: Mon, 21 Apr 2008 12:06:37 -0400 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: <480CBB8D.5070704@v2gnu.com> Gerrard Geldenhuis wrote: > I use sleep a lot and () > /mnt/sysimage/root/nochroot-post.log 2>&1 but > apart from that there must be an easier way. > > I use sleep between commands so I can see whats going on, and do my testing in VMWare. Still cumbersome though. From Pablo.Iranzo at redhat.com Mon Apr 21 16:07:41 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Mon, 21 Apr 2008 18:07:41 +0200 (CEST) Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: What I do is to use the following trick: %pre ( commands )|tee -a /tmp/pre.log %post --nochroot ( commands cp /tmp/pre.log /mnt/sysimage/root/pre.log ) |tee -a /mnt/sysimage/root/post.log And of course, lot's of "echo", and "read" if you want to do some pause ;) Regards Pablo -- Pablo Iranzo G?mez (http://Alufis35.uv.es/~iranzo/) (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) -- Postulado de Boling sobre la Ley de Murphy: Si se encuentra bien, no se preocupe. Se le pasar? On Mon, 21 Apr 2008, Gerrard Geldenhuis wrote: > Hi > > I am debugging a problem with my kickstart but am founding it immensely > cumbersome to rebuild a machine everytime just to test the last part of > the install a stubborn pre-post script that does not want to work. I' > will probably ask a question about it here later on. > > > > Can anyone suggest a more productive way of debugging kickstart files > that takes up less time than actually building a machine each time? > > > > Regards > > From jgranado at redhat.com Mon Apr 21 16:09:54 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Mon, 21 Apr 2008 18:09:54 +0200 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: <480CBC52.4020905@redhat.com> Gerrard Geldenhuis wrote: > Hi > > I am debugging a problem with my kickstart but am founding it immensely > cumbersome to rebuild a machine everytime just to test the last part of > the install a stubborn pre-post script that does not want to work. I' > will probably ask a question about it here later on. > > > > Can anyone suggest a more productive way of debugging kickstart files > that takes up less time than actually building a machine each time? > If your just debugging, you might want to just install the @core systems and additionally install the packages that are used for the post script. In this way you wont have to install 1000 packages but just a couple of hundred. -- Joel Andres Granados Red Hat / Brno, Czech Republic From Gerrard.Geldenhuis at datacash.com Mon Apr 21 16:15:20 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 21 Apr 2008 17:15:20 +0100 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: Yikes, I imagined that there could(should?) be a way to do a "state" change that would allow you to return to a previous state/position in the kickstart and try something again. Does anybody use the debug mode or now how to use it? I am not sure how to start it at will. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > Sent: 21 April 2008 17:08 > To: Discussion list about Kickstart > Subject: Re: Debugging kickstart general advice sought > > What I do is to use the following trick: > > %pre > ( > commands > )|tee -a /tmp/pre.log > > %post --nochroot > ( > commands > cp /tmp/pre.log /mnt/sysimage/root/pre.log > ) |tee -a /mnt/sysimage/root/post.log > > > And of course, lot's of "echo", and "read" if you want to do some > pause ;) > > Regards > Pablo > > > > > -- > Pablo Iranzo G?mez > (http://Alufis35.uv.es/~iranzo/) > (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) > -- > Postulado de Boling sobre la Ley de Murphy: > > Si se encuentra bien, no se preocupe. Se le pasar? > > On Mon, 21 Apr 2008, Gerrard Geldenhuis wrote: > > > Hi > > > > I am debugging a problem with my kickstart but am founding it immensely > > cumbersome to rebuild a machine everytime just to test the last part of > > the install a stubborn pre-post script that does not want to work. I' > > will probably ask a question about it here later on. > > > > > > > > Can anyone suggest a more productive way of debugging kickstart files > > that takes up less time than actually building a machine each time? > > > > > > > > Regards > > > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From Gerrard.Geldenhuis at datacash.com Mon Apr 21 16:16:40 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 21 Apr 2008 17:16:40 +0100 Subject: Debugging kickstart general advice sought In-Reply-To: <480CBC52.4020905@redhat.com> References: <480CBC52.4020905@redhat.com> Message-ID: > > If your just debugging, you might want to just install the @core systems > and additionally install the packages that are used for the post script. > In this way you wont have to install 1000 packages but just a couple of > hundred. Aggreed, I do that already and my install is very small, relatively anyway. It takes about 1min30sec, but that is still a long time to wait. Regards From Gerrard.Geldenhuis at datacash.com Mon Apr 21 17:39:01 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 21 Apr 2008 18:39:01 +0100 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > Sent: 21 April 2008 17:08 > To: Discussion list about Kickstart > Subject: Re: Debugging kickstart general advice sought > > What I do is to use the following trick: > > %pre > ( > commands > )|tee -a /tmp/pre.log > > %post --nochroot > ( > commands > cp /tmp/pre.log /mnt/sysimage/root/pre.log > ) |tee -a /mnt/sysimage/root/post.log > > > And of course, lot's of "echo", and "read" if you want to do some > pause ;) > > Regards > Pablo Does the tee trick catch all stdout and stderr messages? Regards From Pablo.Iranzo at redhat.com Mon Apr 21 20:30:34 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Mon, 21 Apr 2008 22:30:34 +0200 (CEST) Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: Hi You can put before the "|" one 2>&1 to get every single message on files Regards Pablo -- Pablo Iranzo G??mez (http://Alufis35.uv.es/~iranzo/) (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) -- Postulado de Boling sobre la Ley de Murphy: Si se encuentra bien, no se preocupe. Se le pasar?? On Mon, 21 Apr 2008, Gerrard Geldenhuis wrote: > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > > Sent: 21 April 2008 17:08 > > To: Discussion list about Kickstart > > Subject: Re: Debugging kickstart general advice sought > > > > What I do is to use the following trick: > > > > %pre > > ( > > commands > > )|tee -a /tmp/pre.log > > > > %post --nochroot > > ( > > commands > > cp /tmp/pre.log /mnt/sysimage/root/pre.log > > ) |tee -a /mnt/sysimage/root/post.log > > > > > > And of course, lot's of "echo", and "read" if you want to do some > > pause ;) > > > > Regards > > Pablo > > Does the tee trick catch all stdout and stderr messages? > > Regards > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From jake at v2gnu.com Tue Apr 22 17:03:43 2008 From: jake at v2gnu.com (Jake Vickers) Date: Tue, 22 Apr 2008 13:03:43 -0400 Subject: Not really a kickstart question..... Message-ID: <480E1A6F.3070201@v2gnu.com> So we can't use netconfig in RHEL5 kickstarts anymore to set the IP during the installation phase. I can't seem to get the setup-tool to run on first boot (where you can config the net card, turn off SELinux, etc.). Am I missing something obvious, or does everyone else who has a need for this just script it in the kickstart file? Thanks for any suggestions. From email at jasonkohles.com Tue Apr 22 18:33:13 2008 From: email at jasonkohles.com (Jason Kohles) Date: Tue, 22 Apr 2008 14:33:13 -0400 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: <8D41813D-3A03-400D-9E55-7E35409DA782@jasonkohles.com> On Apr 21, 2008, at 12:15 PM, Gerrard Geldenhuis wrote: > Yikes, > > I imagined that there could(should?) be a way to do a "state" change > that would allow you to return to a previous state/position in the > kickstart and try something again. > This is why I use vmware for testing, you can let the install get up to the point where you want to test, create a snapshot, let it continue, and then roll back to that snapshot at any point. -- Jason Kohles, RHCA RHCDS RHCE email at jasonkohles.com - http://www.jasonkohles.com/ "A witty saying proves nothing." -- Voltaire From john.j.meyers at boeing.com Wed Apr 23 18:31:52 2008 From: john.j.meyers at boeing.com (Meyers, John J) Date: Wed, 23 Apr 2008 11:31:52 -0700 Subject: Not really a kickstart question..... In-Reply-To: <480E1A6F.3070201@v2gnu.com> References: <480E1A6F.3070201@v2gnu.com> Message-ID: <8E40364C42ABEE4089507070AEB0C81B056F01CC@XCH-NW-8V2.nw.nos.boeing.com> Hi Jake, In the %pre section of the ks file, I run a python script that is VERY derived from the anaconda network_text.py code. This allows the installer to input the required network info. After which the rest of the install is automated. I can prefill the subnet, ip to last octet, DNS server... If you don't have fancy requirements, you should be able to import network_text *, and call the various modules from anaconda to provide netconfig functionality. Yeah Kickstart! And long live yum. -----Original Message----- From: Jake Vickers [mailto:jake at v2gnu.com] Sent: Tuesday, April 22, 2008 10:04 AM To: Discussion list about Kickstart Subject: Not really a kickstart question..... So we can't use netconfig in RHEL5 kickstarts anymore to set the IP during the installation phase. I can't seem to get the setup-tool to run on first boot (where you can config the net card, turn off SELinux, etc.). Am I missing something obvious, or does everyone else who has a need for this just script it in the kickstart file? Thanks for any suggestions. From mdehaan at redhat.com Wed Apr 23 18:37:49 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 23 Apr 2008 14:37:49 -0400 Subject: Not really a kickstart question..... In-Reply-To: <8E40364C42ABEE4089507070AEB0C81B056F01CC@XCH-NW-8V2.nw.nos.boeing.com> References: <480E1A6F.3070201@v2gnu.com> <8E40364C42ABEE4089507070AEB0C81B056F01CC@XCH-NW-8V2.nw.nos.boeing.com> Message-ID: <480F81FD.2010303@redhat.com> Meyers, John J wrote: > > Hi Jake, > > In the %pre section of the ks file, I run a python script that > is VERY derived from the anaconda network_text.py code. This allows the > installer to input the required network info. After which the rest of > the install is automated. I can prefill the subnet, ip to last octet, > DNS server... If you don't have fancy requirements, you should be able > to import network_text *, and call the various modules from anaconda to > provide netconfig functionality. Yeah Kickstart! And long live yum. > Cobbler has a way of keeping track of all of these variables and filling in these sections automatically. See this customized section that Robin Bowes contributed today: https://fedorahosted.org/cobbler/wiki/KickstartSnippets#Networkconfiguration This allows you to do things like: cobbler system add --name=foo --hostname=foo.example.com --ip=X ... and have everything substituted into all the kickstarts correctly based on the system that you are installing. (http://cobbler.et.redhat.com) --Michael > -----Original Message----- > From: Jake Vickers [mailto:jake at v2gnu.com] > Sent: Tuesday, April 22, 2008 10:04 AM > To: Discussion list about Kickstart > Subject: Not really a kickstart question..... > > So we can't use netconfig in RHEL5 kickstarts anymore to set the IP > during the installation phase. I can't seem to get the setup-tool to run > on first boot (where you can config the net card, turn off SELinux, > etc.). > Am I missing something obvious, or does everyone else who has a need for > this just script it in the kickstart file? > Thanks for any suggestions. > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From ebrown at lanl.gov Wed Apr 23 21:35:46 2008 From: ebrown at lanl.gov (Ed Brown) Date: Wed, 23 Apr 2008 15:35:46 -0600 Subject: Debugging kickstart general advice sought In-Reply-To: References: Message-ID: <480FABB2.8060009@lanl.gov> Gerrard Geldenhuis wrote: > Does anybody use the debug mode or now how to use it? I am not sure how to start it at will. Is there a debug mode? There really should be if there isn't. After all, this is a nearly universal problem, especially when first setting up kickstart, when making changes, even with ordinarily working configurations that suddenly develop problems with any of the variables involved (network connectivity, web/ftp/nfs services, repos, etc). I argued for a better built-in debug capability or framework in this thread from August and September 06: My workaround for the lack of a native debug-ability is to execute the pre and post scripts on tty3, use error-checking like you would in any other scripting that you do, and open a shell on problems. (You can also switch to tty2 if you need another commandline.) You can do discovery/debugging in the actual pre/post environment, very possibly learn how to fix your script, or work around some temporary external issue, and when you exit from the shell, continue with the rest of the pre/post script. %post chvt 3 exec < /dev/tty3 > /dev/tty3 2>/dev/tty3 ...some commands... ... some-command that deserves an error check... if [ $? -ne 0 ]; then echo 'some-command failed, check it out...' sh fi ...more commands -Ed From jturro at redhat.com Thu Apr 24 07:24:29 2008 From: jturro at redhat.com (Pep Turro Mauri) Date: Thu, 24 Apr 2008 08:24:29 +0100 Subject: Debugging kickstart general advice sought In-Reply-To: <480FABB2.8060009@lanl.gov> References: <480FABB2.8060009@lanl.gov> Message-ID: <20080424072429.GI18336@pep.fab.redhat.com> On Wed, Apr 23, 2008 at 03:35:46PM -0600, Ed Brown wrote: > > Is there a debug mode? There really should be if there isn't. This can be useful: http://fedoraproject.org/wiki/Anaconda and in particular: http://fedoraproject.org/wiki/Anaconda/Stage2DevelopmentGuide While this is about debugging anaconda itself and not the the %pre or %post scripts, some tips might help... -- Pep Turr? Mauri Technical Support Engineer Red Hat Global Support Services - EMEA From pgroven at 2wire.com Thu Apr 24 20:02:14 2008 From: pgroven at 2wire.com (Phillip Groven) Date: Thu, 24 Apr 2008 13:02:14 -0700 Subject: cobbler questions Message-ID: <299E8BB0D5475E4FABB162E7DCA1BBE902E2837B@phxexch01.corp.2wire.com> If I have a working kickstart will cobbler overwrite my existing kickstart? What OS versions does cobbler run on? What OS versions can cobbler kickstart? What versions of vmware will it support? How is cobbler different than Red Hat Satellite server? Is cobbler supported by Red Hat Enterprise? Is there a cobbler-list instead of kickstart-list? -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Thu Apr 24 20:22:12 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 24 Apr 2008 16:22:12 -0400 Subject: cobbler questions In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE902E2837B@phxexch01.corp.2wire.com> References: <299E8BB0D5475E4FABB162E7DCA1BBE902E2837B@phxexch01.corp.2wire.com> Message-ID: <4810EBF4.4010309@redhat.com> Phillip Groven wrote: > * * > > * * > > *If I have a working kickstart will cobbler overwrite my existing > kickstart?* > No, thankfully not :) Having a migration path is important. Cobbler can use existing kickstarts by referencing them as http:// URLs, but also has a more powerful kickstart templating system that helps you construct kickstarts with subtle differences between your various profiles and systems. In this way, you can share a common kickstart template, but maintain your package lists or other machine specific options outside of your kickstart template. This makes it very workable for static IP configurations, or when you have a lot of customization to do that would be hard with having a kickstart for each individual system. Cobbler uses the widely-used template language Cheetah to power it's templating technology. > > ** > > *What OS versions does cobbler run on?* > All of this information is on http://cobbler.et.redhat.com -- The server needs EL 4, EL 5, Fedora 7 and later. koan can also run on RHEL3. Many non-Fedora/Red-Hat based distributions can also be provisioned. > ** > > *What OS versions can cobbler kickstart?* > Any RHEL or derivative distro, Any Fedora, and also some support for deploying Debian/Ubuntu or SuSE (though this is less evolved -- but growing over time). > ** > > *What versions of vmware will it support?* > At this point, just VMware server (the free, but not open source, one). Someone has added code to do VMware workstation but I need to incorporate it back into "koan", which is Cobbler's helper tool. Cobbler works much better with tools like Xen and qemu/KVM, both of which are supported by libvirt. VMware is not and the management tools are less useful. > > ** > > * * > > *How is cobbler different than Red Hat Satellite server?* > Cobbler offers more customization options in lots of ways, but also does not include monitoring or config management built in, like Satellite can do. Cobbler does not require a database -- there is a web interface, though most of the power of cobbler can be accessed from the command line. Package management in Cobbler is also mainly keyed around using yum. Cobbler also supports deploying distributions that Satellite does not yet support in some cases -- such as also deploying Fedora. koan also supports some additional virtualization features that Satellite does not support -- namely qemu/KVM, VMware (experimental), installing in LVM partitions, multiple network interfaces, and so forth. If you are not large enough to need a Satellite -- you can still use Cobbler, though if you are that large, with something like rhn_regks, you can still use Cobbler to deploy systems that register against Satellite. It's easy to do. > > ** > > * * > > *Is cobbler supported by Red Hat Enterprise?* > Cobbler is not generally available in an "officially supported" context yet. It is however available in EPEL ( http://fedoraproject.org/wiki/EPEL ) and many people do use Cobbler to deploy their Red Hat Enterprise Linux. Cobbler does have community support via the mailing list and IRC, and maintains seperate stable/development releases. Regardless, the Red Hat Enterprise Linux you deploy /with/ Cobbler will still be supported :) > ** > > *Is there a cobbler-list instead of kickstart-list?* > > * * > Yes! et-mgmt-tools at redhat.com -- which you can find the signup page from https://fedorahosted.org/cobbler (which hosts the Wiki and Trac instance for Cobbler). > > * * > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From john.j.meyers at boeing.com Thu Apr 24 20:21:05 2008 From: john.j.meyers at boeing.com (Meyers, John J) Date: Thu, 24 Apr 2008 13:21:05 -0700 Subject: Not really a kickstart question..... In-Reply-To: <4810812C.6070102@v2gnu.com> References: <480E1A6F.3070201@v2gnu.com> <8E40364C42ABEE4089507070AEB0C81B056F01CC@XCH-NW-8V2.nw.nos.boeing.com> <4810812C.6070102@v2gnu.com> Message-ID: <8E40364C42ABEE4089507070AEB0C81B056F01CD@XCH-NW-8V2.nw.nos.boeing.com> Hi Jake, The attached script is called in the following manner: As you can see I do network based installs; You can modify to your favorite method(s). ################## begin ks.cfg ##################### # Kickstart file automatically generated by anaconda. install nfs --server= --dir=/vol/rhel5repo/repodata/yum/RedHat/5Client/x86_64 key #System language lang en_US keyboard us #Reboot after installation reboot #Use text mode install text #Network information %include /tmp/net.cfg #XWindows configuration information skipx rootpw --iscrypted firewall --disabled authconfig --enableshadow --enablemd5 selinux --disabled timezone --utc America/Los_Angeles bootloader --location=mbr --md5pass= # The following is the partition information you requested # Note that any partitions you deleted are not expressed # here so unless you clear all partitions first, this is # not guaranteed to work clearpart --all part /boot --fstype ext3 --size=100 --asprimary part / --fstype ext3 --size=32768 --asprimary part swap --size=4096 --asprimary %packages @office @engineering-and-scientific @admin-tools @editors @system-tools @gnome-software-development @text-internet @x-software-development @gnome-desktop @core @authoring-and-publishing @base @java-development @games @legacy-software-development @java @legacy-software-support @base-x @graphics @printing @kde-software-development @kde-desktop @sound-and-video @development-libs @development-tools @graphical-internet @workstation xinetd system-config-kickstart kdegames k3b vim-X11 emacs bind bind-chroot system-config-bind dhcp -nmap -bluez-hcidump -isdn4k-utils -lrzsz -sysreport kernel-devel %pre --interpreter=/bin/sh # cd /tmp wget ftp:///yum/ks/bin/ksnetcfg.py chmod 777 ksnetcfg.py /usr/bin/python ksnetcfg.py # ###################### end ks.cfg #################### Cheers, James Meyers -----Original Message----- From: Jake Vickers [mailto:jake at v2gnu.com] Sent: Thursday, April 24, 2008 5:47 AM To: Meyers, John J Subject: Re: Not really a kickstart question..... Meyers, John J wrote: > > Hi Jake, > > In the %pre section of the ks file, I run a python script that is > VERY derived from the anaconda network_text.py code. This allows the > installer to input the required network info. After which the rest of > the install is automated. I can prefill the subnet, ip to last octet, > DNS server... If you don't have fancy requirements, you should be able > to import network_text *, and call the various modules from anaconda > to provide netconfig functionality. Yeah Kickstart! And long live yum. > Sorry to be a pain, but would it be possible for me to get a copy of that python script? Right now I am using the firstboot --enable option but this is not ideal for me. Especially since it does not allow you to configure name servers. Thanks! -------------- next part -------------- A non-text attachment was scrubbed... Name: ksnetcfg.py Type: application/octet-stream Size: 8140 bytes Desc: ksnetcfg.py URL: From pgroven at 2wire.com Fri Apr 25 01:23:37 2008 From: pgroven at 2wire.com (Phillip Groven) Date: Thu, 24 Apr 2008 18:23:37 -0700 Subject: sha rootpw encryption in ks.cfg Message-ID: <299E8BB0D5475E4FABB162E7DCA1BBE902E28380@phxexch01.corp.2wire.com> In my ks.cfg files I encrypt the root password using grub-md5-crypt According to the below article I an use sha512 How can I encrypt the password using sha? What version of Red Hat is Rawhide? Tips and tricks: Choosing the password hashing algorithm for /etc/shadow during installation by Alexander Todorov This is a new feature that has been requested in the Red Hat Enterprise Linux product line. The supported algorithms are DES , MD5 , and SHA-256/512 . They will be available after the upcoming updates releases as well as in Rawhide . The new algorithms will be configurable via Kickstart only. Here is the correct code. DES authconfig --enableshadow --passalgo=descrypt rootpw fedora MD5 authconfig --enableshadow --enablemd5 rootpw fedora or authconfig --enableshadow --passalgo=md5 rootpw fedora SHA-256 authconfig --enableshadow --passalgo=sha256 rootpw fedora SHA-512 authconfig --enableshadow --passalgo=sha512 rootpw fedora -------------- next part -------------- An HTML attachment was scrubbed... URL: From gene.huft at wachovia.com Fri Apr 25 01:26:37 2008 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Thu, 24 Apr 2008 21:26:37 -0400 Subject: Huft, Gene is out of the office. Message-ID: I will be out of the office starting 04/24/2008 and will not return until 04/29/2008. I am out of office until Tues, Apr 29. I will be checking email infrequently & will reply as soon as possible. Best regards, Gene L Huft -------------- next part -------------- An HTML attachment was scrubbed... URL: From mytechacc at gmail.com Fri Apr 25 03:13:50 2008 From: mytechacc at gmail.com (MyTechAccount) Date: Thu, 24 Apr 2008 23:13:50 -0400 Subject: Not really a kickstart question..... In-Reply-To: <20080425012404.5A4FA618F77@hormel.redhat.com> References: <20080425012404.5A4FA618F77@hormel.redhat.com> Message-ID: I wrote a kickstart cgi that incorporates our machine database so I can dump the correct netinfo based on incoming mac address. so all I have to do it enter the mac/ip into a table and boot the machine. the kickme cgi then returns the correct info and autopopulates the network config data based on settings we have on our internal subnets. you don't need anything that fancy though, just complete the network config in the kickstart.cfg output skipx network --device eth0 --bootproto static --ip 192.168.20.189 -- netmask 255.255.2 55.0 --gateway 192.168.20.254 --nameserver 192.168.20.96 --hostname something.something.something rootpw --iscrypted yadayada and it's completely automated On Apr 24, 2008, at 9:24 PM, kickstart-list-request at redhat.com wrote: >> >> Hi Jake, >> >> In the %pre section of the ks file, I run a python script that > is >> VERY derived from the anaconda network_text.py code. This allows the >> installer to input the required network info. After which the rest of >> the install is automated. I can prefill the subnet, ip to last octet, >> DNS server... If you don't have fancy requirements, you should be >> able > >> to import network_text *, and call the various modules from anaconda >> to provide netconfig functionality. Yeah Kickstart! And long live >> yum. -------------- next part -------------- An HTML attachment was scrubbed... URL: From email at jasonkohles.com Fri Apr 25 04:10:14 2008 From: email at jasonkohles.com (Jason Kohles) Date: Fri, 25 Apr 2008 00:10:14 -0400 Subject: sha rootpw encryption in ks.cfg In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE902E28380@phxexch01.corp.2wire.com> References: <299E8BB0D5475E4FABB162E7DCA1BBE902E28380@phxexch01.corp.2wire.com> Message-ID: <33A85F2A-46E1-4514-AFE8-CFAA7C1C170A@jasonkohles.com> On Apr 24, 2008, at 9:23 PM, Phillip Groven wrote: > > In my ks.cfg files I encrypt the root password using grub-md5-crypt > > According to the below article I an use sha512 > How can I encrypt the password using sha? The article you quoted tells you how to do that.... authconfig --enableshadow --passalgo=sha512 > What version of Red Hat is Rawhide? Rawhide isn't a specific version, it's an automatically built distribution of whatever the latest versions of all the packages are in the source repository. It's intended for use by developers doing testing and it is frequently broken or unstable. What the comment means is that if you want it now you have to get it from Rawhide and hope you can make it work, otherwise you have to wait until the next release, which will include these changes. > Tips and tricks: Choosing the password hashing algorithm for /etc/ > shadow during installation > by Alexander Todorov > This is a new feature that has been requested in the Red Hat > Enterprise Linux product line. The supported algorithms are DES,MD5, > and SHA-256/512. They will be available after the upcoming updates > releases as well as in Rawhide. The new algorithms will be > configurable via Kickstart only. Here is the correct code. > DES > authconfig --enableshadow --passalgo=descrypt > rootpw fedora > MD5 > authconfig --enableshadow --enablemd5 > rootpw fedora > or > authconfig --enableshadow --passalgo=md5 > rootpw fedora > SHA-256 > authconfig --enableshadow --passalgo=sha256 > rootpw fedora > SHA-512 > authconfig --enableshadow --passalgo=sha512 > rootpw fedora > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Jason Kohles, RHCA RHCDS RHCE email at jasonkohles.com - http://www.jasonkohles.com/ "A witty saying proves nothing." -- Voltaire -------------- next part -------------- An HTML attachment was scrubbed... URL: From wfelipe at gmail.com Fri Apr 25 15:28:07 2008 From: wfelipe at gmail.com (Wilson Felipe) Date: Fri, 25 Apr 2008 12:28:07 -0300 Subject: problem with the module bnx2 Message-ID: hello, We are starting to install redhat 5.1 and centos 5.1 in our environment, and it all went well until we tried with Dell machines (PE1950 and 860), which have network cards using bnx2 (using MSI). The problem is at the bnx2 module, we found out 2 way: - using dhcp at pxe and at the boot: the kernel and the initrd loads from pxe boot, and stops at downloading the ks file, though it gets the ip from the dhcp, but the network link is never put up, but in messages it prints "the link becomes ready" - using dhcp at pxe, and static ip at the boot the kernel and the initrd loads from pxe boot, and also stops at download the ks file, but after a few seconds, the links becomes ready and we can get the ks file and the installation goes all well has anyone seen this behavior using bnx2? From Gerrard.Geldenhuis at datacash.com Mon Apr 28 08:43:59 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 28 Apr 2008 09:43:59 +0100 Subject: problem with the module bnx2 In-Reply-To: References: Message-ID: Hi > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Wilson Felipe > Sent: 25 April 2008 16:28 > To: kickstart-list at redhat.com > Subject: problem with the module bnx2 > > hello, > > We are starting to install redhat 5.1 and centos 5.1 in our > environment, and it all went well until we tried with Dell machines > (PE1950 and 860), which have network cards using bnx2 (using MSI). The > problem is at the bnx2 module, we found out 2 way: > > - using dhcp at pxe and at the boot: > the kernel and the initrd loads from pxe boot, and stops at > downloading the ks file, though it gets the ip from the dhcp, but the > network link is never put up, but in messages it prints "the link > becomes ready" > > - using dhcp at pxe, and static ip at the boot > the kernel and the initrd loads from pxe boot, and also stops at > download the ks file, but after a few seconds, the links becomes ready > and we can get the ks file and the installation goes all well > > has anyone seen this behavior using bnx2? > We use HP machines with bnx2(broadcom) and e1000(intel) nics and we have not seen the same behaviour. Could it be switch related? Regards From acapotorto at lastminute.com Mon Apr 28 09:41:28 2008 From: acapotorto at lastminute.com (Anibal Capotorto) Date: Mon, 28 Apr 2008 10:41:28 +0100 Subject: problem with the module bnx2 In-Reply-To: Message-ID: I?ve run into a lot of stress when using the bnx2 cards with rhel4u4 /u5 on the HPs. Basically 3 things that are a problem with this nics. 1st the problem you?re talking about I think it?s definitely because of the topology of the network given that the bnx2 family sometimes takes some time in bringing the link up which will lead into The ks process timing out and asking you to enter the ip manually. With the HPs I?ve had 2 problems, 1st one the nics weren?t detected @ boot time, the pxe was doing fine, but the OS wasn?t able to bring up the nics This one was fixed disabled the apci=off to on (or vice versa, I can?t really remember well that one, but it?s definitely something to do with that. THEN when the boxes were installed ok, the problem was that if I wanted to do a /etc/init.d/network restart the nic died and won?t come back until the next reboot The error in the restart was something related to ? chipset failed to reset ? This was due to the complicated nature of the board itself being HP and supporting iSCSI, the only solution was to replace the nics This were PCI-E nics, the onboard ones worked fine with the apci=off thing.. Hope this helps On 28/04/2008 09:43, "Gerrard Geldenhuis" wrote: > Hi > >> > -----Original Message----- >> > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- >> > bounces at redhat.com] On Behalf Of Wilson Felipe >> > Sent: 25 April 2008 16:28 >> > To: kickstart-list at redhat.com >> > Subject: problem with the module bnx2 >> > >> > hello, >> > >> > We are starting to install redhat 5.1 and centos 5.1 in our >> > environment, and it all went well until we tried with Dell machines >> > (PE1950 and 860), which have network cards using bnx2 (using MSI). The >> > problem is at the bnx2 module, we found out 2 way: >> > >> > - using dhcp at pxe and at the boot: >> > the kernel and the initrd loads from pxe boot, and stops at >> > downloading the ks file, though it gets the ip from the dhcp, but the >> > network link is never put up, but in messages it prints "the link >> > becomes ready" >> > >> > - using dhcp at pxe, and static ip at the boot >> > the kernel and the initrd loads from pxe boot, and also stops at >> > download the ks file, but after a few seconds, the links becomes ready >> > and we can get the ks file and the installation goes all well >> > >> > has anyone seen this behavior using bnx2? >> > > > We use HP machines with bnx2(broadcom) and e1000(intel) nics and we have > not seen the same behaviour. Could it be switch related? > > Regards > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > > -- > Anibal Capotorto Sr. Linux Engineer - travelocity global > mob: +44 078 12960970 > 24x7 Tech-SOC number:- +44 (0)20 7866 4500 > acapotorto at lastminute.com > -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image.gif Type: image/gif Size: 3284 bytes Desc: not available URL: From jake at v2gnu.com Mon Apr 28 10:29:57 2008 From: jake at v2gnu.com (Jake Vickers) Date: Mon, 28 Apr 2008 06:29:57 -0400 Subject: Not really a kickstart question..... In-Reply-To: <8E40364C42ABEE4089507070AEB0C81B056F01CD@XCH-NW-8V2.nw.nos.boeing.com> References: <480E1A6F.3070201@v2gnu.com> <8E40364C42ABEE4089507070AEB0C81B056F01CC@XCH-NW-8V2.nw.nos.boeing.com> <4810812C.6070102@v2gnu.com> <8E40364C42ABEE4089507070AEB0C81B056F01CD@XCH-NW-8V2.nw.nos.boeing.com> Message-ID: <4815A725.2010600@v2gnu.com> Meyers, John J wrote: > Hi Jake, > The attached script is called in the following manner: > As you can see I do network based installs; You can modify to your > favorite method(s). > > Thanks! I'll give this a shot later in the week. From Gerrard.Geldenhuis at datacash.com Mon Apr 28 10:50:28 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 28 Apr 2008 11:50:28 +0100 Subject: pxe menu In-Reply-To: References: Message-ID: Hi I received a lot of good replies on my question but also realized that I was putting constraints on myself were none were needed. If you are willing to do a bit more admin then there is no reason why menu items could not be split horizonataly across the screen rather than limiting yourself to 20someting single entry items. Thus 1. Option1 2. Option2 3.Option3 4. Option4 5. Option5 6.Option6 rather than 1. Option1 2. Option2 3. Option3 ... Regards ________________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Gerrard Geldenhuis Sent: 08 April 2008 10:10 To: Discussion list about Kickstart Subject: pxe menu Hi I have been looking through the documentation to see if you can have sub menus in the pxe boot menu but could not find anything that talks specifically about it. I could probably adapt the function key menus but would prefer to use the normal menus. Does anyone have an example I could use or point me to the right documentation please. Regards From Pablo.Iranzo at redhat.com Mon Apr 28 10:59:00 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 28 Apr 2008 12:59:00 +0200 Subject: pxe menu In-Reply-To: References: Message-ID: <1209380340.22171.0.camel@iranzo.users.redhat.com> Are you talking about "menu.c32" or usual menus where a text is displayed and depending on the option, there's something to be loaded? Regards Pablo El lun, 28-04-2008 a las 11:50 +0100, Gerrard Geldenhuis escribi?: > Hi > I received a lot of good replies on my question but also realized that I > was putting constraints on myself were none were needed. If you are > willing to do a bit more admin then there is no reason why menu items > could not be split horizonataly across the screen rather than limiting > yourself to 20someting single entry items. > > Thus > 1. Option1 2. Option2 3.Option3 > 4. Option4 5. Option5 6.Option6 > > rather than > 1. Option1 > 2. Option2 > 3. Option3 > ... > > Regards > > ________________________________________ > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Gerrard > Geldenhuis > Sent: 08 April 2008 10:10 > To: Discussion list about Kickstart > Subject: pxe menu > > Hi > I have been looking through the documentation to see if you can have sub > menus in the pxe boot menu but could not find anything that talks > specifically about it. I could probably adapt the function key menus but > would prefer to use the normal menus. Does anyone have an example I > could use or point me to the right documentation please. > > Regards > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From debian at herakles.homelinux.org Mon Apr 28 11:18:20 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Mon, 28 Apr 2008 19:18:20 +0800 Subject: pxe menu In-Reply-To: References: Message-ID: <4815B27C.7060705@herakles.homelinux.org> Gerrard Geldenhuis wrote: > Hi > I received a lot of good replies on my question but also realized that I > was putting constraints on myself were none were needed. If you are > willing to do a bit more admin then there is no reason why menu items > could not be split horizonataly across the screen rather than limiting > yourself to 20someting single entry items. > > Thus > 1. Option1 2. Option2 3.Option3 > 4. Option4 5. Option5 6.Option6 > > rather than > 1. Option1 > 2. Option2 > 3. Option3 > ... > > Regards centos{4,5}{,k,r} Centos You might understand this to allow to install, manually or by kickstart, CentOS 4 or 5, or to boot in rescue mode. It extrapolates from shell expansions, so it shouldn't be _too_ confusing. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From Gerrard.Geldenhuis at datacash.com Mon Apr 28 11:44:32 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Mon, 28 Apr 2008 12:44:32 +0100 Subject: pxe menu In-Reply-To: <1209380340.22171.0.camel@iranzo.users.redhat.com> References: <1209380340.22171.0.camel@iranzo.users.redhat.com> Message-ID: > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > Sent: 28 April 2008 11:59 > To: Discussion list about Kickstart > Subject: RE: pxe menu > > Are you talking about "menu.c32" or usual menus where a text is > displayed and depending on the option, there's something to be loaded? > > Regards > Pablo > No, I was just talking about stock standard pxe without any additional modules being loaded. Regards From astrand at cendio.se Tue Apr 29 16:51:27 2008 From: astrand at cendio.se (=?UTF-8?Q?Peter_=C3=85strand?=) Date: Tue, 29 Apr 2008 18:51:27 +0200 (CEST) Subject: Anaconda Traceback when using %pre scripts Message-ID: Suddenly, the Anaconda installer gives me tracebacks when using %pre scripts. I can see the last part of the traceback: File "/usr/lib/anaconda/kickstart.py", line 90, os.chmod("%s" % messages, 0600) OSError: [Errno 2] No such file or directory: 'False' From looking at the code, I believe that the problem is that "self.logfile" is False rather than None: if self.logfile is not None: messages = self.logfile elif serial: messages = "%s.log" % path else: messages = "/dev/tty3" Any ideas? Best regards, --- Peter ?strand ThinLinc Chief Developer Cendio AB http://www.cendio.se Wallenbergs gata 4 583 30 Link?ping Phone: +46-13-21 46 00 From prasadtvs_2003 at yahoo.com Tue Apr 29 18:00:16 2008 From: prasadtvs_2003 at yahoo.com (prasad tvs) Date: Tue, 29 Apr 2008 11:00:16 -0700 (PDT) Subject: Anaconda Traceback when using %pre scripts In-Reply-To: Message-ID: <867584.23945.qm@web82701.mail.mud.yahoo.com> May I know which interpreter are you using..?Curios about "Is that interpreter actually supported..?" Thanks, Prasad. Peter ?strand wrote: Suddenly, the Anaconda installer gives me tracebacks when using %pre scripts. I can see the last part of the traceback: File "/usr/lib/anaconda/kickstart.py", line 90, os.chmod("%s" % messages, 0600) OSError: [Errno 2] No such file or directory: 'False' >From looking at the code, I believe that the problem is that "self.logfile" is False rather than None: if self.logfile is not None: messages = self.logfile elif serial: messages = "%s.log" % path else: messages = "/dev/tty3" Any ideas? Best regards, --- Peter ?strand ThinLinc Chief Developer Cendio AB http://www.cendio.se Wallenbergs gata 4 583 30 Link?ping Phone: +46-13-21 46 00_______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list -------------- next part -------------- An HTML attachment was scrubbed... URL: From astrand at cendio.se Tue Apr 29 18:18:05 2008 From: astrand at cendio.se (=?UTF-8?Q?Peter_=C3=85strand?=) Date: Tue, 29 Apr 2008 20:18:05 +0200 (CEST) Subject: Anaconda Traceback when using %pre scripts In-Reply-To: <867584.23945.qm@web82701.mail.mud.yahoo.com> References: <867584.23945.qm@web82701.mail.mud.yahoo.com> Message-ID: > May I know which interpreter are you using..?Curios about "Is that interpreter actually supported..?" I've seen this in two different cases: 1) A Python tool started from a normal shell: %pre /a/python/tool %end 2) A Python interpreter: %pre --interpreter /usr/bin/python import os import subprocess ... Best regards, --- Peter ?strand ThinLinc Chief Developer Cendio AB http://www.cendio.se Wallenbergs gata 4 583 30 Link?ping Phone: +46-13-21 46 00 From clumens at redhat.com Tue Apr 29 20:51:16 2008 From: clumens at redhat.com (Chris Lumens) Date: Tue, 29 Apr 2008 16:51:16 -0400 Subject: Anaconda Traceback when using %pre scripts In-Reply-To: References: Message-ID: <20080429205112.GO27333@localhost.localdomain> > Suddenly, the Anaconda installer gives me tracebacks when using %pre > scripts. I can see the last part of the traceback: > > File "/usr/lib/anaconda/kickstart.py", line 90, > os.chmod("%s" % messages, 0600) > > OSError: [Errno 2] No such file or directory: 'False' What does your kickstart file look like? - Chris From muksyed at stanford.edu Wed Apr 30 05:41:00 2008 From: muksyed at stanford.edu (Mukarram Syed) Date: Tue, 29 Apr 2008 22:41:00 -0700 Subject: Issues with kickstarting RHEL3 with Dell PE 1950 Message-ID: <080701c8aa84$c8ac7710$2e1c42ab@stanford.edu> Hi, Our Kickstart has been working fine with RHEL4 and EL4.no problems. But today I added a RHEL3 menu and it's corresponding ks.cfg file. This is to test our legacy RHEL3 servers which are difficult to get rid of at this time. I could pxe boot and I get the menu. I choose the correct option (17) and it loads up the appropriate vmlinuz and initrd images, then loads up anaconda and instead of getting the network it shows me the Language menu. I have about 16 other options for RHEL4 and EL4 and they all work fine. I have been troubleshooting this for almost 3 hours now and no luck. Suggestions would be appreciated. My take on this is that with RHEL4 it takes all the drivers for the PE1950, for RHEL3 I don't think it contains all the drivers so it just does not load them with anaconda. Can anyone clarify this? Here is my setup: My /tftpboot/linux-install/pxelinux.cfg/default file: Label 16 works but Label 17 does not. label 16 kernel RHELAS32_U6/vmlinuz append nofb ks=http://171.67.39.130/RHELAS32_U6/RedHat/ks_el32_u4_asia-muksyed.cfg initrd=RHELAS32_U6/initrd.img ramdisk_size=10000 lang= devfs=nomount ramdisk_size=9216 ksdevice=eth0 console=ttyS0,9600 ipappend 1 label 17 kernel RHEL3/vmlinuz append nofb ks=http://171.67.39.130/RHEL3/RedHat/ks_el32_u4_asia-muksyed.cfg initrd=RHEL3/initrd.img ramdisk_size=10000 lang= devfs=nomount ramdisk_size=9216 ksdevice=eth0 console=ttyS0,9600 ipappend 1 My /tftpboot/linux-install/msgs/boot.msg file (menu): 16. Base (X Window) ASIA Install for Tripwire Server 17. Base (X Window) RHEL3 My /var/www/html/RHEL3/RedHat/ ks_el32_u4_asia-muksyed.cfg file (ks.cfg equivalent). I could http to: http://171.67.39.130/RHEL3 #System language lang en_US #Language modules to install langsupport en_US #System keyboard keyboard us #System mouse mouse #Sytem timezone timezone America/Los_Angeles #Root password rootpw welcome #Reboot after installation reboot #Install OS instead of upgrade install #Use Web installation url --url http://171.67.39.130/RHEL3 #System bootloader configuration bootloader --location=mbr #Clear the Master Boot Record zerombr yes #Partition clearing information clearpart --all --initlabel #Disk partitioning information part /boot --fstype ext3 --size 100 --asprimary part / --fstype ext3 --size 6000 --asprimary part swap --recommended --asprimary part /tmp --fstype ext3 --size 2000 part /opt --fstype ext3 --size 3000 --grow #System authorization infomation auth --useshadow --enablemd5 #Network information network --bootproto=dhcp --device=eth0 #Firewall configuration firewall --disabled #SELinux configuration selinux --disabled #Do not configure XWindows skipx #Package install information %packages --resolvedeps @ base-x # OEM Agent and Oracle Client Requirements: binutils compat-db compat-libstdc++-296 libaio glibc glibc-common gcc gcc-c++ make gnome-libs libstdc++ libstdc++-devel pdksh xscreensaver setarch control-center # Other RPMs: ntp screen cvs emacs lynx strace sysstat %pre %post -------------- next part -------------- An HTML attachment was scrubbed... URL: From Gerrard.Geldenhuis at datacash.com Wed Apr 30 08:25:21 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 30 Apr 2008 09:25:21 +0100 Subject: Error Message: Volume group defined without any physical volumes Message-ID: Hi I get the following error when trying to build a IBM 336 server: Volume group defined without any physical volumens. Either specify physical volumes or use --existing. The machine contains an old debian system. The same kickstart works fine on HP boxes with redhat or nothing installed. A quick search through my archives of the list suggest people with seemingly related problems, the solution has been to format the disk in a pre-script before commencing install. I am going to try it but am still curious as to why this is happening? What would be an easy/simple way to differentiate between say an HP DL360 G5 and an IBM xSeries 336 in the pre-script? Regards From Pablo.Iranzo at redhat.com Wed Apr 30 08:29:53 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Wed, 30 Apr 2008 10:29:53 +0200 Subject: Error Message: Volume group defined without any physical volumes In-Reply-To: References: Message-ID: <1209544193.8766.63.camel@iranzo.users.redhat.com> Hi El mi?, 30-04-2008 a las 09:25 +0100, Gerrard Geldenhuis escribi?: > What would be an easy/simple way to differentiate between say an HP > DL360 G5 and an IBM xSeries 336 in the pre-script? dmidecode is your friend ;) -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Gerrard.Geldenhuis at datacash.com Wed Apr 30 08:32:48 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 30 Apr 2008 09:32:48 +0100 Subject: Error Message: Volume group defined without any physical volumes In-Reply-To: <1209544193.8766.63.camel@iranzo.users.redhat.com> References: <1209544193.8766.63.camel@iranzo.users.redhat.com> Message-ID: > > Hi > > El mi?, 30-04-2008 a las 09:25 +0100, Gerrard Geldenhuis escribi?: > > > What would be an easy/simple way to differentiate between say an HP > > DL360 G5 and an IBM xSeries 336 in the pre-script? > > dmidecode is your friend ;) > Excellent! you don't perhaps have a small example lying around? ;-) From Pablo.Iranzo at redhat.com Wed Apr 30 08:47:49 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Wed, 30 Apr 2008 10:47:49 +0200 Subject: Error Message: Volume group defined without any physical volumes In-Reply-To: References: <1209544193.8766.63.camel@iranzo.users.redhat.com> Message-ID: <1209545269.8766.72.camel@iranzo.users.redhat.com> What I use at my systems is: TYPE=`dmidecode |grep "Product Name:"|grep "[0-9].*"|head -n 1|cut -d ":" -f 2|cut -d " " -f 2` This provides for me on the systems I've tested one number that I can use for differentiate them: D1419: FSC TX200 D2239: FSC TX150S4 PRIMERGY: FSC TX150S5 1707Y8F: Lenovo X60 ProLiant: Compaq Proliant For me that works, as I'm only differentiating some machines from anothers, but if you've a look at the output, you can choose a better one that fits your needs :) Regards Pablo El mi?, 30-04-2008 a las 09:32 +0100, Gerrard Geldenhuis escribi?: > > > > Hi > > > > El mi?, 30-04-2008 a las 09:25 +0100, Gerrard Geldenhuis escribi?: > > > > > What would be an easy/simple way to differentiate between say an HP > > > DL360 G5 and an IBM xSeries 336 in the pre-script? > > > > dmidecode is your friend ;) > > > > Excellent! you don't perhaps have a small example lying around? ;-) > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Gerrard.Geldenhuis at datacash.com Wed Apr 30 14:32:47 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 30 Apr 2008 15:32:47 +0100 Subject: Error Message: Volume group defined without any physical volumes In-Reply-To: References: Message-ID: > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Gerrard Geldenhuis > Sent: 30 April 2008 09:25 > To: Discussion list about Kickstart > Subject: Error Message: Volume group defined without any physical volumes > > Hi > I get the following error when trying to build a IBM 336 server: > Volume group defined without any physical volumens. Either specify > physical volumes or use --existing. > > The machine contains an old debian system. > > The same kickstart works fine on HP boxes with redhat or nothing > installed. > I've managed to get a different error message which now says not enough space. If I boot the server with a CD then the installation works fine. The kickstart I used is a vanilla kickstart created by Satellite with no frills. I am suspecting that there is a problem with the drivers being loaded for the storage. How would I go about determining the difference? The cd uses kernel 2.6.18-53.el5 and the pxe boot is using the latest kernel which is 2.6.18-53.1.14.el5. Regards From Gerrard.Geldenhuis at datacash.com Wed Apr 30 16:40:24 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 30 Apr 2008 17:40:24 +0100 Subject: Strange error when changing ksdevice Message-ID: Hi I have some old IBM hardware which I must get working nicely with out Satellite system. The first problem is that is switches the onboard and pci network cards around. Thus the onboard cards become eth2 and eth3 instead of eth0 and eth1. That can be overcome by specifying ksdevice=eth2. However I don't want to do that because that would break the install for the HP hardware which is specified as ksdevice=eth0. I thought specifying ksdevice=eth0 would not be a problem because when the server boots it would pxe boot via the onboard card where the pxe boot entry would tell it: ksdevice=eth0 which would then cause it to do the rest via eth0 which is the added pci card. This works and it downloads the kickstart file but it fails when trying to partition the server. I get the following error: Volume group defined without any physical volumes. Either specify physical volumes or use --useexisting. I find it really bizarre because why would ksdevice=eth0 vs ksdevice=eth2 cause a partition failure? Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From Gerrard.Geldenhuis at datacash.com Wed Apr 30 16:42:27 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 30 Apr 2008 17:42:27 +0100 Subject: Error Message: Volume group defined without any physical volumes In-Reply-To: References: Message-ID: > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Gerrard Geldenhuis > Sent: 30 April 2008 15:33 > To: Discussion list about Kickstart > Subject: RE: Error Message: Volume group defined without any physical > volumes > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > > bounces at redhat.com] On Behalf Of Gerrard Geldenhuis > > Sent: 30 April 2008 09:25 > > To: Discussion list about Kickstart > > Subject: Error Message: Volume group defined without any physical > volumes > > > > Hi > > I get the following error when trying to build a IBM 336 server: > > Volume group defined without any physical volumens. Either specify > > physical volumes or use --existing. > > > > The machine contains an old debian system. > > > > The same kickstart works fine on HP boxes with redhat or nothing > > installed. > > > > I've managed to get a different error message which now says not enough > space. If I boot the server with a CD then the installation works fine. > The kickstart I used is a vanilla kickstart created by Satellite with no > frills. I am suspecting that there is a problem with the drivers being > loaded for the storage. How would I go about determining the difference? > > The cd uses kernel 2.6.18-53.el5 and the pxe boot is using the latest > kernel which is 2.6.18-53.1.14.el5. > > Regards Problem solved. I changed my kickstart to a very basic version but neglected to remove the --nostorage kernel parameter that I generally use. Schoolboy error, I will blame tiredness. Regards From brent.clements at gmail.com Wed Apr 30 23:26:48 2008 From: brent.clements at gmail.com (Brent Clements) Date: Wed, 30 Apr 2008 18:26:48 -0500 Subject: RHEL 3 kickstart network retrieval issue Message-ID: I am having the wierdest issue with RHEL 3 network installs. When I boot with a RHEL 3 boot media, I use the following boot commandline: linux ks=http://myhost/RHEL3.ks ksdevice=eth0 ip= netmask= gateway= dns= Anaconda boots up but it fails at installing. I switch over to a terminal and it shows that anaconda has failed to retrieve the kickstart file from the network and also I see a no route to host error. Does not look like the network device is being set up correctly based upon the boot commandline parameters. To Troubleshoot to make sure it wasn't the kickstart file that was the problem, I stuck this same kickstart file as an embedded kickstart on the cd and it worked fine. So that doesn't get me anywhere and validates that my actual kickstart file doesn't have errors and can install an OS So, I try doing the same thing but with RHEL 4 media and a RHEL4 based kickstart using the boot commandline linux ks=http://myhost/RHEL4.ks ksdevice=eth0 ip= netmask= gateway= dns= Everything works fine using RHEL 4 and the command line stated right above. Are their known issues with setting network parameters at the command line with RHEL 3? -------------- next part -------------- An HTML attachment was scrubbed... URL: