From s.davison at computer.org Sun Jan 2 16:13:20 2005 From: s.davison at computer.org (S. W. Davison) Date: Sun, 02 Jan 2005 11:13:20 -0500 Subject: 100% network (remote) install -- No direct machine access? Message-ID: <41D81DA0.4070400@computer.org> I'm a kickstart beginner, but it seems to me that this ought to be possible. Maybe someone can point me at the right documentation... We have many sites which are geographically distant from each other. At each site there are 10-20 machines running an old version of Linux. We would like to upgrade all the machines to RHEL 3, but we would like to make the upgrades entirely remotely controlled and entirely automated. Specifically, we don't want to be in the situation where personnel at each site must visit each machine for any purpose. We are trying to avoid: o Any use of floppies or CDs. o Any use of console output from the machines. (Many of them do not have consoles, unless you plug one in especially.) o Any keyboard activity on the machine. (Many of them do not have keyboards attached in their normal service configuration.) o Any need to go into the machine's BIOS setup to enable network booting or PXE. (As far as I know, you can do that only with a local keyboard and console.) Reading the kickstart documentation, it looks as if one way to accomplish a kickstart install is to put a kickstart file (ks.cfg) on an ext2 partition of the machine to be upgraded. The ks.cfg file would specify an NFS host where the installation tree was available. Then you boot from a boot/install floppy, and at the boot prompt you enter: linux ks=hd:sda3:/mydir/ks.cfg OK. If I can do that, it seems to me that it ought to be possible to build the necessary vmlinuz, intitrd, and whatever else I needed (the stuff that would ordinarily be on the boot/install floppy), put them into /boot on the machine to be upgraded, and then change lilo's default boot command to: "linux ks=hd:sda3:/mydir/ks.cfg" Then, if I do a "shutdown -r" over a network connection, the target system should reboot and go straight into the kickstart installation, just as if it had booted from the floppy disk. Is that possible? Can anyone point me at documentation for setting up the necessary boot files? Thanks for your patience in reading this, and many thanks for any suggestions. Stowe Davison s.davison at computer.org From rwcitek at alum.calberkeley.org Sun Jan 2 17:20:21 2005 From: rwcitek at alum.calberkeley.org (Robert Citek) Date: Sun, 2 Jan 2005 11:20:21 -0600 Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: <41D81DA0.4070400@computer.org> Message-ID: <94A7F32C-5CE2-11D9-BCDF-000A95880F88@alum.calberkeley.org> On Sunday, Jan 2, 2005, at 10:13 US/Central, S. W. Davison wrote: > We would like to upgrade all the machines to RHEL 3, but we > would like to make the upgrades entirely remotely controlled > and entirely automated. Yes, it is possible. Here are my notes on upgrading a headless box (p166, 64 MB RAM, 2 GB HDD) from FC1 to FC3: http://www.cwelug.org/cgi-bin/wiki.cgi?FedoraCore3#bootstrap_install BTW, one thing that I have not been able to get working is a kickstart that creates an LVM on a machine with < 128 MB RAM. Good luck and let us know how things go. Regards, - Robert http://www.cwelug.org/downloads Help others get OpenSource. Distribute FLOSS for Windows, Linux, *BSD, and MacOS X with BitTorrent From tdiehl at rogueind.com Sun Jan 2 21:17:16 2005 From: tdiehl at rogueind.com (Tom Diehl) Date: Sun, 2 Jan 2005 16:17:16 -0500 (EST) Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: <41D81DA0.4070400@computer.org> References: <41D81DA0.4070400@computer.org> Message-ID: On Sun, 2 Jan 2005, S. W. Davison wrote: > > > > I'm a kickstart beginner, but it seems to me that this ought > to be possible. Maybe someone can point me at the right > documentation... > > We have many sites which are geographically distant from > each other. At each site there are 10-20 machines running > an old version of Linux. > > We would like to upgrade all the machines to RHEL 3, but we > would like to make the upgrades entirely remotely controlled > and entirely automated. Specifically, we don't want to be > in the situation where personnel at each site must visit > each machine for any purpose. We are trying to avoid: > o Any use of floppies or CDs. > o Any use of console output from the machines. (Many of > them do not have consoles, unless you plug one in > especially.) > o Any keyboard activity on the machine. (Many of them do > not have keyboards attached in their normal service > configuration.) > o Any need to go into the machine's BIOS setup to enable > network booting or PXE. (As far as I know, you can do > that only with a local keyboard and console.) > > Reading the kickstart documentation, it looks as if one way > to accomplish a kickstart install is to put a kickstart file > (ks.cfg) on an ext2 partition of the machine to be upgraded. > The ks.cfg file would specify an NFS host where the > installation tree was available. Then you boot from a > boot/install floppy, and at the boot prompt you enter: > linux ks=hd:sda3:/mydir/ks.cfg > > OK. If I can do that, it seems to me that it ought to be > possible to build the necessary vmlinuz, intitrd, and > whatever else I needed (the stuff that would ordinarily > be on the boot/install floppy), put them into /boot on the > machine to be upgraded, and then change lilo's default boot > command to: "linux ks=hd:sda3:/mydir/ks.cfg" > > Then, if I do a "shutdown -r" over a network connection, the > target system should reboot and go straight into the > kickstart installation, just as if it had booted from the > floppy disk. > > Is that possible? Can anyone point me at documentation for > setting up the necessary boot files? That should work, but another thing you could do is put your ks.cfg file on a web server, then copy the kernel and initrd file from the images/pxe directory on the first install disk to something like /boot/pxe. Add an entry to lilo/grub.conf that looks something like the following: linux ks=http://www.mydomain.com/kickstart/ks.cfg ksdevice=link\ ramdisk=-32768 If you have a properly built ks.cfg it should work just fine. If you want to keep an eye on the progress of the installation simply add a vnc line to your ks.cfg file pointed to a machine you have access too. I have done this many times. It takes a fair amount of fiddeling to get the ks.cfg correct but it can be done. You can also set up a yum repository and use a combiination of kickstart and yum to get the machines installed remotely. Using yum in the %post of the ksconfig along with the yumgroups stuff allows you to basically have the same ks.cfg for multiple machines and just vary the yumgroups.xml file based on the hostname for a particular machine. Again getting this setup initally requires some fiddeling but in the end you end up with a machine that is fully updated on the first boot, without having to build a custom distro. HTH, Tom From mcotton at aptas.com Sun Jan 2 21:58:29 2005 From: mcotton at aptas.com (Mike Cotton) Date: Sun, 2 Jan 2005 14:58:29 -0700 Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: Message-ID: <200501022158.j02LwEt13122@vhost6.atomicservers.com> Tom, I have been having issues getting my post script tweaked in. Do you have one I can compare to? It is an NFS install that works great but when I go to run yum and some other file copies - it doesn't work as expected. I am fairly sure it has something to with the chroot --nochroot options not getting set up correctly. I can send mine to you if that helps. Thanks. Mike -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Tom Diehl Sent: Sunday, January 02, 2005 2:17 PM To: Discussion list about Kickstart Subject: Re: 100% network (remote) install -- No direct machine access? On Sun, 2 Jan 2005, S. W. Davison wrote: > > > > I'm a kickstart beginner, but it seems to me that this ought > to be possible. Maybe someone can point me at the right > documentation... > > We have many sites which are geographically distant from > each other. At each site there are 10-20 machines running > an old version of Linux. > > We would like to upgrade all the machines to RHEL 3, but we > would like to make the upgrades entirely remotely controlled > and entirely automated. Specifically, we don't want to be > in the situation where personnel at each site must visit > each machine for any purpose. We are trying to avoid: > o Any use of floppies or CDs. > o Any use of console output from the machines. (Many of > them do not have consoles, unless you plug one in > especially.) > o Any keyboard activity on the machine. (Many of them do > not have keyboards attached in their normal service > configuration.) > o Any need to go into the machine's BIOS setup to enable > network booting or PXE. (As far as I know, you can do > that only with a local keyboard and console.) > > Reading the kickstart documentation, it looks as if one way > to accomplish a kickstart install is to put a kickstart file > (ks.cfg) on an ext2 partition of the machine to be upgraded. > The ks.cfg file would specify an NFS host where the > installation tree was available. Then you boot from a > boot/install floppy, and at the boot prompt you enter: > linux ks=hd:sda3:/mydir/ks.cfg > > OK. If I can do that, it seems to me that it ought to be > possible to build the necessary vmlinuz, intitrd, and > whatever else I needed (the stuff that would ordinarily > be on the boot/install floppy), put them into /boot on the > machine to be upgraded, and then change lilo's default boot > command to: "linux ks=hd:sda3:/mydir/ks.cfg" > > Then, if I do a "shutdown -r" over a network connection, the > target system should reboot and go straight into the > kickstart installation, just as if it had booted from the > floppy disk. > > Is that possible? Can anyone point me at documentation for > setting up the necessary boot files? That should work, but another thing you could do is put your ks.cfg file on a web server, then copy the kernel and initrd file from the images/pxe directory on the first install disk to something like /boot/pxe. Add an entry to lilo/grub.conf that looks something like the following: linux ks=http://www.mydomain.com/kickstart/ks.cfg ksdevice=link\ ramdisk=-32768 If you have a properly built ks.cfg it should work just fine. If you want to keep an eye on the progress of the installation simply add a vnc line to your ks.cfg file pointed to a machine you have access too. I have done this many times. It takes a fair amount of fiddeling to get the ks.cfg correct but it can be done. You can also set up a yum repository and use a combiination of kickstart and yum to get the machines installed remotely. Using yum in the %post of the ksconfig along with the yumgroups stuff allows you to basically have the same ks.cfg for multiple machines and just vary the yumgroups.xml file based on the hostname for a particular machine. Again getting this setup initally requires some fiddeling but in the end you end up with a machine that is fully updated on the first boot, without having to build a custom distro. HTH, Tom _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From phr at doc.ic.ac.uk Mon Jan 3 20:50:20 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Mon, 3 Jan 2005 20:50:20 +0000 (GMT) Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: <41D81DA0.4070400@computer.org> References: <41D81DA0.4070400@computer.org> Message-ID: On Sun, 2 Jan 2005, S. W. Davison wrote: >We would like to upgrade all the machines to RHEL 3, but we >would like to make the upgrades entirely remotely controlled >and entirely automated. >Reading the kickstart documentation, it looks as if one way >to accomplish a kickstart install is to put a kickstart file >(ks.cfg) on an ext2 partition of the machine to be upgraded. Possible, although unnecessary. The kickstart file can live out on the network (HTTP, FTP, or NFS). >If I can do that, it seems to me that it ought to be possible to build >the necessary vmlinuz, intitrd, and whatever else I needed (the stuff >that would ordinarily be on the boot/install floppy), put them into >/boot on the machine to be upgraded, and then change lilo's default >boot command to: "linux ks=hd:sda3:/mydir/ks.cfg" Yes, and if your requirements aren't too strange the images/pxeboot/{vmlinuz,initrd.img} will handle all supported drivers and install methods. >Then, if I do a "shutdown -r" over a network connection, the target >system should reboot and go straight into the kickstart installation, >just as if it had booted from the floppy disk. Nearly; the bootloader will have to be directed to load the files mentioned above. With remote machines, I'd be sure to test, test, test locally before reinstalling for real; the part to pay the most attention to would be the network setup. Cheers, Phil From rwcitek at alum.calberkeley.org Mon Jan 3 21:50:31 2005 From: rwcitek at alum.calberkeley.org (Robert Citek) Date: Mon, 3 Jan 2005 15:50:31 -0600 Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: Message-ID: <7D83F750-5DD1-11D9-BD5D-000A95880F88@alum.calberkeley.org> On Monday, Jan 3, 2005, at 14:50 US/Central, Philip Rowlands wrote: > With remote machines, I'd be sure to test, test, test locally before > reinstalling for real; the part to pay the most attention to would be > the network setup. Another item to consider is to have either two machines connected via serial connections or an IP-based KVM. In that way if the install fails for whatever reason, you can remotely connect to the other box or the KVM and work on the machine. For a serial connection, you would use two machines and a null modem cable. This assumes that each machine has an available serial port. First, you would connect the null modem cable from the serial port of one machine to the serial of another. Then you would run minicom on the second machine to listen for the first machine. Lastly, you would modify the grub.conf on the first machine to have grub and the kernel output to a serial console: http://www.cwelug.org/cgi-bin/wiki.cgi?SerialConsole You then reboot the first machine. Minicom should then show you the grub console window. In this way you can access the second machine via ssh and then use minicom to connect to the first. Once the first machine is upgraded, you can do the reverse roles to upgrade the second machine (if needed). The serial connections are more limited but cost less. KVM over IP is a much cleaner solution, but significantly more expensive. As always, let us know what you decided and how it went. Regards, - Robert http://www.cwelug.org/downloads Help others get OpenSource. Distribute FLOSS for Windows, Linux, *BSD, and MacOS X with BitTorrent From e_chillon at yahoo.com.mx Tue Jan 4 20:54:38 2005 From: e_chillon at yahoo.com.mx (=?iso-8859-1?q?Edgar=20Chill=F3n=20'Chillux'?=) Date: Tue, 4 Jan 2005 14:54:38 -0600 (CST) Subject: hdstg2.img netstg2.img stage2.img Message-ID: <20050104205438.97794.qmail@web52210.mail.yahoo.com> Is there extreamly necesary to get this files in order to instal the fedora base system ? hdstg2.img netstg2.img stage2.img All my system is based in a kikstart file, so the image are not loaded ... so ? do I need this images ? Regards ... ===== -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- Edgar Alberto Chill?n Esc?rcega Mobile: 04455-1951-9194 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- _________________________________________________________ Do You Yahoo!? La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. http://net.yahoo.com.mx From andrew.w.robinson at mms.gov Tue Jan 4 21:24:51 2005 From: andrew.w.robinson at mms.gov (Robinson, Andrew W.) Date: Tue, 4 Jan 2005 14:24:51 -0700 Subject: How to update anaconda? Message-ID: <379313C94B76D2119AB60008C7A402E409E46309@IMSNOLAA> The problem I'm trying to solve is this. I am attempting a kickstart-directed network installation with RHEL 3 EL update 3 on a Dell 2650 server. At the point when anaconda tries to get files from the server by nfs, I get a dialog box with this message: "That directory could not be mounted from the server." I click OK. Anaconda gives me another dialog to set the server and path. I leave that just as it is and click OK. This process happens anywhere from 2 to 10 times, but eventually anaconda finds the directory and the installation continues. This, however, is not the problem I'm asking about. I think I know the answer to this. What I don't know is how to implement the answer. I believe the problem is related to this bug on bugzilla: kickstart fails to get kickstartfile when using e1000 network driver . The reporter found a problem in the interaction between the e1000 driver and anaconda. The final solution given in this bugzilla was a reference to this web page: Updated anaconda and pump packages . The instructions were to download and install the updated anaconda and pump packages. My first question is what do I do next? Looking at /usr/share/doc/anaconda-9.1.4.1/anaconda-release-notes.txt, it appears that I need to put some updated files in a directory named RHupdates on my repository. That leads to a whole host of new questions, that I'll enumerate for my own thought process: 1. What files go into RHupdates? How do I know which of the anaconda files were updated? Can I just put all the files there that were installed by the anaconda rpms? 2. What is the structure of RHupdates? Do I put all the files directly under RHupdates or do I maintain the same directory structure they enjoy outside of RHupdates? 3. Where do I put RHupdates? I tried putting it parallel to the RedHat directory in my repository and I tried putting it directly under the RedHat directory. I never got my installation to acknowledge the presence of the RHupdates directory despite an afternoon of experimenting. 4. How do I tell anaconda to look for RHupdates? I tried the "updates" directive listed in the command-line.txt file. Anaconda would ask if I wanted to use fd0 or hda as the source of the updates. Again, I never got it to acknowledge the presence of the RHupdates directory. 5. When anaconda did ask me what source I wanted to use for updates, it did so only after the round of questions and answers about the nfs server. Will attempting to update anaconda this way even help? Do I need to use a floppy disk to provide the updates? If so, I suppose I really will need to know which files to use because of the space limitations of a floppy. 6. Finally, does someone know of a source that would answer these questions for me? I spent a couple of hours with Google but didn't come up with anything definitive. Thanks! Andrew Robinson -------------- next part -------------- An HTML attachment was scrubbed... URL: From jim at rossberry.com Tue Jan 4 23:39:15 2005 From: jim at rossberry.com (Jim Wildman) Date: Tue, 4 Jan 2005 18:39:15 -0500 (EST) Subject: hdstg2.img netstg2.img stage2.img In-Reply-To: <20050104205438.97794.qmail@web52210.mail.yahoo.com> References: <20050104205438.97794.qmail@web52210.mail.yahoo.com> Message-ID: On Tue, 4 Jan 2005, [iso-8859-1] Edgar Chill?n 'Chillux' wrote: > hdstg2.img netstg2.img stage2.img > > All my system is based in a kikstart file, so the > image are not loaded ... so ? do I need this images ? > Those are program images (compressed disk images), not graphic images. And yes, you have to have them. ------------------------------------------------------------------------ Jim Wildman, CISSP, RHCE jim at rossberry.com http://www.rossberry.com "Society in every state is a blessing, but Government, even in its best state, is a necessary evil; in its worst state, an intolerable one." Thomas Paine From dadembro at rockwellcollins.com Thu Jan 6 00:21:57 2005 From: dadembro at rockwellcollins.com (dadembro at rockwellcollins.com) Date: Wed, 5 Jan 2005 19:21:57 -0500 Subject: Kickstart fails with multiple ethernet adapters Message-ID: I hope someone can provide some insight with the following problem. I have several duplicate systems I am trying to use a kickstart file to setup with redhat version 9. One of the systems has an additional ethernet adapter and this one fails to use the kickstart file which all the other systems accept. Sample kickstart command: linux ks=http://199.0.0.100/kickstart/u_ios_ks.cfg ip=199.0.0.5 netmask=255.255.255.0 this works for all the systems except the one with two network adapters. linux ks=http://199.0.0.100/kickstart/u_ios_ks.cfg ip=199.0.0.5 netmask=255.255.255.0 ksdevice=eth0 also fails, however, linux ks=hd:fd0/u_ios_ks.cfg ksdevice=eth0 works. Somehow the same kickstart file is read via floppy diskette but not over the network. Can anyone explain what is going on? I am tempted on just using the floppy for now. Thanks, ---d.dembrow From rwcitek at alum.calberkeley.org Thu Jan 6 00:41:06 2005 From: rwcitek at alum.calberkeley.org (Robert Citek) Date: Wed, 5 Jan 2005 18:41:06 -0600 Subject: Kickstart fails with multiple ethernet adapters In-Reply-To: Message-ID: On Wednesday, Jan 5, 2005, at 18:21 US/Central, wrote: > Sample kickstart command: > > linux ks=http://199.0.0.100/kickstart/u_ios_ks.cfg ip=199.0.0.5 > netmask=255.255.255.0 ksdevice=eth0 > > also fails I don't see anything obviously wrong. Here's the isolinux.cfg stanza that we use for kickstarting RHL9 via a local http server: default ks prompt 1 timeout 40 label ks kernel vmlinuz append ksdevice=eth0 ks=http://10.4.0.42/pub/redhat.9.0/ks/ks.cfg initrd=initrd.img text Our machines have either dual 10/100 NICs or dual GigE NICs. Differences between what you have and what we use: - order of the ksdevice option - kickstart file is named ks.cfg - initrd option You might want to play with some of those options. In any case, let us know what you discover. Regards, - Robert http://www.cwelug.org/downloads Help others get OpenSource. Distribute FLOSS for Windows, Linux, *BSD, and MacOS X with BitTorrent From tdiehl at rogueind.com Thu Jan 6 15:15:23 2005 From: tdiehl at rogueind.com (Tom Diehl) Date: Thu, 6 Jan 2005 10:15:23 -0500 (EST) Subject: Kickstart fails with multiple ethernet adapters In-Reply-To: References: Message-ID: On Wed, 5 Jan 2005 dadembro at rockwellcollins.com wrote: > I hope someone can provide some insight with the following problem. > > I have several duplicate systems I am trying to use a kickstart file to > setup with redhat version 9. > One of the systems has an additional ethernet adapter and this one fails to > use the kickstart file > which all the other systems accept. > > Sample kickstart command: > > linux ks=http://199.0.0.100/kickstart/u_ios_ks.cfg ip=199.0.0.5 > netmask=255.255.255.0 > > this works for all the systems except the one with two network adapters. > > linux ks=http://199.0.0.100/kickstart/u_ios_ks.cfg ip=199.0.0.5 > netmask=255.255.255.0 ksdevice=eth0 > > also fails, however, > > linux ks=hd:fd0/u_ios_ks.cfg ksdevice=eth0 Try ksdevice=link. I have seen anaconda's idea of which device is eth0 be different than mine on occasion. HTH, Tom From i_p_a_u_l at yahoo.com Sat Jan 8 00:35:59 2005 From: i_p_a_u_l at yahoo.com (Paul Ionescu) Date: Sat, 08 Jan 2005 02:35:59 +0200 Subject: anaconda network install Message-ID: Hi, I install FC or RH from network, with DHCP settings. Is there an option to send the default ftp/http/nfs server and the directory via DHCP instead of entering manually ? Maybe is not something standard, but we can use some vendor DHCP options for this. Or even better, can we give to anaconda a kickstart file through DHCP ? I don't want to put all the time "ks=http://someaddress/somedirectory/some_kickstartfile" when I need to install a PC. I think, that by default, if we specify "ks" to anaconda, to try and get it first from a dhcp server on first eth, then on second eth, a.s.o, then go to normal behavior. And we try the following files: "info_from_dhcp" + "mac address of that card" "info from dhcp" + "ip address of that card" "info from dhcp" This would ease networked installs and automation of them. Is this doable ? Thanks, Paul From ryan at email.arc.nasa.gov Sat Jan 8 02:23:18 2005 From: ryan at email.arc.nasa.gov (Ryan Nelson) Date: Fri, 07 Jan 2005 18:23:18 -0800 Subject: anaconda network install In-Reply-To: References: Message-ID: <41DF4416.6060908@email.arc.nasa.gov> Paul Ionescu wrote: > Hi, > > I install FC or RH from network, with DHCP settings. > > Is there an option to send the default ftp/http/nfs server and the > directory via DHCP instead of entering manually ? > Maybe is not something standard, but we can use some vendor DHCP options > for this. > > Or even better, can we give to anaconda a kickstart file through DHCP ? > I don't want to put all the time > "ks=http://someaddress/somedirectory/some_kickstartfile" when I need to > install a PC. > > I think, that by default, if we specify "ks" to anaconda, to try and get > it first from a dhcp server on first eth, then on second eth, a.s.o, then > go to normal behavior. And we try the following files: > "info_from_dhcp" + "mac address of that card" > "info from dhcp" + "ip address of that card" > "info from dhcp" > > This would ease networked installs and automation of them. > Is this doable ? > > Thanks, > Paul > Well I wouldn't want my DHCP server giving out the KS file to everyone or I don't want to have to enter a MAC address + IP + KS file into my dhcp server each time. What I have done is modified the isolinux.cfg to contain the ks line as well. This way all I have to have is my build network which gives out DHCP to any system and a boot cd that I tell which version of linux I want to install. Works for me and I think might be a little cleaner. (Although I sure theirs cleaner) Ryan From herrold at owlriver.com Sat Jan 8 03:25:58 2005 From: herrold at owlriver.com (R P Herrold) Date: Fri, 7 Jan 2005 22:25:58 -0500 (EST) Subject: ks] Re: anaconda network install In-Reply-To: <41DF4416.6060908@email.arc.nasa.gov> References: <41DF4416.6060908@email.arc.nasa.gov> Message-ID: On Fri, 7 Jan 2005, Paul Ionescu wrote >> Is there an option to send the default ftp/http/nfs server and the >> directory via DHCP instead of entering manually ? >> Maybe is not something standard, but we can use some vendor DHCP options >> for this. Well, no, not really that early -- The URL with the location of the ks.cfg file can be passed out, but not as part of the DHCP transaction -- dhcp happens too early for there to be a Vendor Option context for it to be meaningful without some major surgery on how pxelinux.0 works. But dhcp identifies the TFTP 'next' server, and the TFTP transfer moves the information of which you ask. As the next server, and the dhcp server are often (although not formally required to be) the same server, you can probably automate pointing to a desired installation path. I have done so since RHL 6.2 days with variants of what is now my 'install-from-iso.sh' mentioned with a URL in the notes at a writeup I did. see: http://www.owlriver.com/tips/pxe-install/ This writeup dates from a couple years ago, detailling, in cookbook fashion, the configuration for identifying in the 'dhcpd.conf' configuration file, (1) the 'next server', and (2) the relevant part of the path from the tftp-server's (which is the 'next server') root path 'filename'. That 'filename' is the 'pxelinux.bin' to be retrieved and executed. That code is smart enough in turn to request, and retrieve the file containing that information (among other things) from the 'next server' identified under DHCP and PXE install practice. Jan 5 23:25:35 ftp tftpd[2796]: Serving /rhl/pxe/pxelinux.bin to 10.16.33.254:2070 By RH (and derived distribution's) PXElinux convention, there is a file of last resort sought [others are probed for first under a nameing convention derived from the hex represesention of the IP address assigned]. In searching for that file (which is called 'default'), it first asks for a series of more genreally named files, each starting 'OA' in this example, [as '0A' is a hex name for '10' decimal, from the A class network 10.x.x.x in which this server and its clients live]. The found file is transferred out to the requesting client, through tftp, using information derived from DHCP. Jan 5 23:25:35 ftp tftpd[2796]: Serving /rhl/pxe/pxelinux.cfg/0A1021FE to 10.16.33.254:57217 Jan 5 23:25:35 ftp tftpd[2796]: File /tftpboot///rhl/pxe/pxelinux.cfg/0A1021FE not found Jan 5 23:25:35 ftp tftpd[2796]: Serving /rhl/pxe/pxelinux.cfg/0A1 to 10.16.33.254:56582 Jan 5 23:25:35 ftp tftpd[2796]: File /tftpboot///rhl/pxe/pxelinux.cfg/0A1 not found Jan 5 23:25:35 ftp tftpd[2796]: Serving /rhl/pxe/pxelinux.cfg/0A to 10.16.33.254:56455 This is also the approach in Debian. Jan 3 02:29:29 ftp tftpd[31242]: File /tftpboot//debian/pxelinux.cfg/0A1 not found Jan 3 02:29:29 ftp tftpd[31242]: Serving debian/pxelinux.cfg/0A to 10.16.33.221:57096 Jan 3 02:29:29 ftp tftpd[31242]: File /tftpboot//debian/pxelinux.cfg/0A not found Jan 3 02:29:29 ftp tftpd[31242]: Serving debian/pxelinux.cfg/0 to 10.16.33.221:57097 Jan 3 02:29:29 ftp tftpd[31242]: File /tftpboot//debian/pxelinux.cfg/0 not found Jan 3 02:29:29 ftp tftpd[31242]: Serving debian/pxelinux.cfg/default to 10.16.33.221:57098 -- and the '0A' file (in the Red Hat example), found before the failsafe 'default' file (in the Debian example) in turn, points to the ks.cfg, which contains and provides the information you mention. [root at ftp log]# cat /tftpboot/rhl/pxe/pxelinux.cfg/0A default linux # serial 0,38400n8 label linux kernel vmlinuz append ksdevice=eth0 local_ramdisk=1 initrd=initrd.img ks=http://10.16.33.105/pub/kickstart/ks.cfg # paragraph 11 of my outline specifically has what you wish to do. -------------- It is perfectly possible for a single dhcp and tftp server to handle multiple platforms and architectures and distributions. -- Here is a cleaned up snapshot from one of my development servers: [root at ftp etc]# cat /tmp/ks-dhcpd.conf.txt ddns-update-style none ; ping-check true ; get-lease-hostnames true ; max-lease-time 120000; default-lease-time 12000; option option-128 code 128 = string; option option-129 code 129 = text; shared-network development { authoritative ; subnet 10.16.33.0 netmask 255.255.255.0 { option domain-name "development.lan"; option domain-name-servers 10.16.33.105, 65.43.29.92, 66.195.224.112 ; option routers 10.16.33.1; option root-path "10.16.33.105:/opt/ltsp/i386"; allow unknown-clients; host blahblah { hardware ethernet 11:22:33:44:55:66 ; } allow booting ; allow bootp ; next-server 10.16.33.105 ; filename "/rhl/pxe/pxelinux.bin" ; range 10.16.33.200 10.16.33.219 ; range 10.16.33.230 10.16.33.254 ; group { host netwinder1 { hardware ethernet 00:10:57:00:07:54 ; fixed-address 10.16.33.90 ; option host-name "netwinder1"; } host p3-220 { hardware ethernet 00:02:B3:1D:8F:76 ; fixed-address 10.16.33.220 ; option host-name "dhcp220"; } } group { next-server 10.16.33.105 ; host debian-tftpclient { hardware ethernet 00:00:DE:AD:BE:EF ; fixed-address 10.16.33.221 ; filename "debian/pxelinux.0" ; option host-name "debian-tftpclient" ; } } } } [root at ftp etc]# This file supports both Red Hat type hands off kickstart installs, and Debian installs; for arches x86, Netwinder mips, and Aurora sparc installs. And an LTSP client as well. ;) -- Russ Herrold From s.davison at computer.org Sun Jan 9 15:59:33 2005 From: s.davison at computer.org (S. W. Davison) Date: Sun, 09 Jan 2005 10:59:33 -0500 Subject: 100% network (remote) install -- No direct machine access? In-Reply-To: <41D81DA0.4070400@computer.org> References: <41D81DA0.4070400@computer.org> Message-ID: <41E154E5.8000709@computer.org> Many thanks to all who responded to this request. All the comments were helpful. The hands-off install works great. I started from Robert Citek's notes (http://www.cwelug.org/cgi-bin/wiki.cgi?FedoraCore3#bootstrap_install) and adapted them for our situation. In case anyone is interested, I have appended a description of the procedure and the lilo.conf and ks.cfg files. Tom Diehl's suggestion about using VNC to monitor the install and using yumgroups.xml to handle platform-to- platform variance looks interesting. I don't think we need to do that -- not many different platform templates -- but it's something to keep in reserve. There's a lot more to be done before we deploy. As one of the responders noted, we have to do a lot of testing. And there will be all the picky, finicky business of getting the various servers running properly and the application software installed. But all that is straightfoward, although time-consuming. Thanks again for your help! Stowe Davison 1. Copy the contents of the RedHat directories from the four RHEL 3 installation CDs on to an NFS server at the site. That is the "installation tree." 2. Copy RHEL 3 isolinux versions of vmlinuz and initrd.img over to the boot partition of the target machine. Those two files are in the isolinux directory of CD #1 of the RHEL 3 installation set. Same two files for all platforms. 3. Copy a kickstart configuration file, ks.cfg, over to the target system. That file contains the IP addresses of the target system and the NFS server from which the RPMs will be copied, so it must be customized for each target machine. It also contains the partitioning and formatting information and the mount list. So we would probably have a template for each platform type and customize them with correct IP addresses for each target machine. 4. Copy a new version of lilo.conf over to the target machine, then run lilo to install it. The new lilo configuration boots the isolinux kernel that you copied over in step 1, and passes the ks option to the boot command line. Same file for all platforms. 5. Run "shutdown -r now" on the target system. When the target system finishes rebooting it will be running RHEL 3 and it will be accessible on the network. 6. Install application software and configure for operations. prompt timeout=50 default=kick_net boot=/dev/hda map=/boot/map install=/boot/boot.b message=/boot/message linear image=/boot/net_ks/vmlinuz label=kick_net initrd=/boot/net_ks/initrd.img read-only root=/dev/hda1 append="ramdisk_size=65536 nofb text ks=hd:hda1/net_ks/ks.cfg" # Kickstart file automatically generated by anaconda and # edited for reinstall. install # I added the nfs... line. nfs --server=192.168.0.29 --dir=/home3/rhel3ws_install_tree lang en_US.UTF-8 langsupport --default en_US.UTF-8 en_US.UTF-8 keyboard us mouse generic3ps/2 --device psaux xconfig --card "ATI Mach64" --videoram 8192 --hsync 30-69 --vsync 50-120 --resolution 1024x768 --depth 24 --startxonboot --defaultdesktop gnome network --device eth0 --bootproto static --ip 192.168.0.30 --netmask 255.255.255.0 --gateway 192.168.0.1 --nameserver 209.183.205.35,209.183.192.65 --hostname swddeb rootpw --iscrypted $1$4Ic/R8dr$Q9v5e7qemk1KYsSLYV4Xb0 firewall --disabled authconfig --enableshadow --enablemd5 timezone America/New_York bootloader --location=mbr --append hdc=ide-scsi --useLilo # 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 # # All the partition stuff was commented out in the anaconda- # generated version. I uncommented it. clearpart --linux part /boot --fstype ext3 --size=1000 --asprimary part raid.13 --size=5000 part raid.11 --size=5000 part raid.9 --size=5000 part raid.7 --size=5000 part raid.6 --size=5000 part swap --size=1000 --asprimary # I added quotes around raid pv.22 --fstype "physical volume (LVM)" --level=RAID0 raid.13 raid.11 raid.9 raid.7 raid.6 volgroup Volume00 pv.22 logvol / --fstype ext3 --name=LogVol00 --vgname=Volume00 --size=3000 logvol /home --fstype ext3 --name=LogVol01 --vgname=Volume00 --size=10000 logvol /tmp --fstype ext3 --name=LogVol02 --vgname=Volume00 --size=3000 logvol /var --fstype ext3 --name=LogVol04 --vgname=Volume00 --size=2000 logvol /usr --fstype ext3 --name=LogVol03 --vgname=Volume00 --size=6000 # That's the end of the section that was commented by anaconda. # I added reboot reboot %packages @ everything kernel lilo %post From Chris.McKenzie at entrust.com Mon Jan 10 22:19:56 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Mon, 10 Jan 2005 17:19:56 -0500 Subject: Kickstart post shutdown and CD-Rom eject? Message-ID: Hi guys. I'm trying to get the post of my kickstart to eject the CD. More importantly, I'd like to have the system shutdown cleanly instead of rebooting, or just sitting there. I haven't been able to eject using any of the methods I've seen around, this one doesn't work, but it the best attempt I've seen... http://linuxgazette.net/105/nielsen.html As for shutting down, can't seem to get it working at all. Any ideas? Thanks. - Chris -------------- next part -------------- An HTML attachment was scrubbed... URL: From phr at doc.ic.ac.uk Tue Jan 11 17:26:49 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Tue, 11 Jan 2005 17:26:49 +0000 (GMT) Subject: Kickstart post shutdown and CD-Rom eject? In-Reply-To: References: Message-ID: On Mon, 10 Jan 2005, Chris Mckenzie wrote: >I'm trying to get the post of my kickstart to eject the CD. More >importantly, I'd like to have the system shutdown cleanly instead of >rebooting, or just sitting there. Tricky... the loader's [fixed] behaviour is either to reboot if the install succeeded, or halt if not. By shutdown, do you mean poweroff? If so, that's dependent on the APM/ACPI interaction between the kernel and your specific hardware - ask Google. Cheers, Phil From e_chillon at yahoo.com.mx Wed Jan 12 01:31:13 2005 From: e_chillon at yahoo.com.mx (=?iso-8859-1?q?Edgar=20Chill=F3n=20'Chillux'?=) Date: Tue, 11 Jan 2005 19:31:13 -0600 (CST) Subject: copy files cdrom to redhat Message-ID: <20050112013113.32619.qmail@web52208.mail.yahoo.com> In the post installarion how can I tell the sistem that copy a file into a especific directory ? I think is something like this: cp /mnt/cdrom/Fedora/files/example.txt /etc/ But I dont know if the system will respect the /mnt/cdrom ... Any ideas ? ===== -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- Edgar Alberto Chill?n Esc?rcega Mobile: 04455-1951-9194 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- _________________________________________________________ Do You Yahoo!? La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. http://net.yahoo.com.mx From e_chillon at yahoo.com.mx Wed Jan 12 18:04:34 2005 From: e_chillon at yahoo.com.mx (=?iso-8859-1?q?Edgar=20Chill=F3n=20'Chillux'?=) Date: Wed, 12 Jan 2005 12:04:34 -0600 (CST) Subject: Fedora Core Legend Message-ID: <20050112180434.9385.qmail@web52201.mail.yahoo.com> When installing the system, to the top of the screen to the left apears this legend: Fedora Core (C) 2004 RedHat, Inc. Does any body know how to move out this legend ? Regards. ===== -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- Edgar Alberto Chill?n Esc?rcega Mobile: 04455-1951-9194 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- _________________________________________________________ Do You Yahoo!? La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. http://net.yahoo.com.mx From peter at boku.net Wed Jan 12 18:24:45 2005 From: peter at boku.net (Peter Eisch) Date: Wed, 12 Jan 2005 12:24:45 -0600 Subject: Fedora Core Legend In-Reply-To: <20050112180434.9385.qmail@web52201.mail.yahoo.com> Message-ID: Depending on which screen, you're probably looking for the strings in the anaconda-10.0 source. You'll probably want to add/rename an install class as well as other misc fixes in anaconda-10.0/po/anaconda.pot. Using grep in that source will help you find the things to change. peter On 1/12/05 12:04 PM, "Edgar Chill?n 'Chillux'" wrote: > When installing the system, to the top of the screen > to the left apears this legend: > > Fedora Core (C) 2004 RedHat, Inc. > > Does any body know how to move out this legend ? > > Regards. > > ===== > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > Edgar Alberto Chill?n Esc?rcega > Mobile: 04455-1951-9194 > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ > FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > > _________________________________________________________ > Do You Yahoo!? > La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. > http://net.yahoo.com.mx > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From dallase at nmgi.com Thu Jan 13 21:18:58 2005 From: dallase at nmgi.com (Dallas L. Engelken) Date: Thu, 13 Jan 2005 15:18:58 -0600 Subject: Fedora core 3 kernel upgrades Message-ID: I'm trying to rebuild core 3 cd's with update kernels on them, specifically 2.6.10-1.737_FC3. When I run through the buildinstall, my isolinux/vmlinuz file and the modules.cgz in initrd.img still shows that its 2.6.9-1.667. Where is it getting this from?? I don't even have 2.6.9-1.667 installed. Thanks, Dallas From Thomas.Cameron at bankofamerica.com Thu Jan 13 21:53:51 2005 From: Thomas.Cameron at bankofamerica.com (Cameron, Thomas) Date: Thu, 13 Jan 2005 15:53:51 -0600 Subject: Is there a way to disable hardware for kickstart? Message-ID: <1C455916B3418145B320711B21A2F045A69E79@ex2k.bankofamerica.com> Howdy all - I am trying to kickstart a RHEL 3 box with both internal disks and SAN attached storage. Unfortunately the internal storage is /dev/cciss/c0d0 and the SAN stuff is seen as /dev/sda, /dev/sdb and so on. The server boots off of internal disks (/dev/cciss/c0d0) but kickstart wants to write the boot record to the SAN (/dev/sda). So I was thinking that I could somehow tell the kernel to not load the drivers for the QLogic fibre channel cards. I think that in modules.conf I could do "alias qla2200 off" but I am not sure how to do this from the Syslinux boot: prompt. Help? Thomas From usergroups at theargoncompany.com Thu Jan 13 22:13:16 2005 From: usergroups at theargoncompany.com (Rishi) Date: Fri, 14 Jan 2005 03:43:16 +0530 Subject: generic ks.cfg to auto-partition raid0 on X number of disks. Message-ID: <200501140343.16909.usergroups@theargoncompany.com> Hi I wanted to know if it is possible to create a generic kickstart file to auto-partition the disk or disks in the following manner - First / partition with 1 Gb - Second partition is 500 MB swap and - The space on the remaining disk or disks be configured as /backup with RAID0 Issues....... - Machine 'A' may have just one 40 GB HDD ... - Machine 'B' may have just two 40 GB HDD ... - Machine 'C' may have just one 40 GB IDE HDD and one 30 GB SCSI disk. - Machine 'D' may have three HDD of 20,40 and 40 GB Is it possible to create a generic ks.cfg to work for all? I've tried but I have not understood the what raid.X means.... Any ideas / tips on how to achieve this? Thanks Regards Rishi From jjensen at cisco.com Thu Jan 13 22:30:17 2005 From: jjensen at cisco.com (Joshua Jensen) Date: Thu, 13 Jan 2005 17:30:17 -0500 Subject: Fedora core 3 kernel upgrades In-Reply-To: References: Message-ID: <20050113223016.GC6285@cisco.com> >From your buildtree. What is the exactly command and args that you are using? Joshua On Thu, Jan 13, 2005 at 03:18:58PM -0600, Dallas L. Engelken wrote: > I'm trying to rebuild core 3 cd's with update kernels on them, > specifically 2.6.10-1.737_FC3. When I run through the buildinstall, my > isolinux/vmlinuz file and the modules.cgz in initrd.img still shows that > its 2.6.9-1.667. Where is it getting this from?? I don't even have > 2.6.9-1.667 installed. > > Thanks, > Dallas > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Joshua Jensen Work: 919 392 1488 Mobile: 919 454 9679 Fax: 919 392 8902 Pager: 866 264 2691 Epage: jjensen ? epage.cisco.com IBM Linux Architect - RHCE, RHCT, etc "If God didn't want us to eat animals, why did he make them out of meat?" From hal at reactivity.com Thu Jan 13 23:05:18 2005 From: hal at reactivity.com (Hal Wine) Date: Thu, 13 Jan 2005 15:05:18 -0800 Subject: RPM limitations at kickstart time? Message-ID: <41E6FEAE.1000903@reactivity.com> with rpm 4.2.3-10 (RHEL 3 ES update 3), I'm seeing rpm database corruption when having an RPM install another RPM in a %post scriptlet during kickstart. Supposedly, this is permitted since 4.1-1, and it works fine outside of kickstart (anaconda). rpm at kickstart time claims to support the feature (as advertised via examination of 'rpm --showrc' output). Any pointers to solutions? Or documentation of the limitations at kickstart time? Thanks, --Hal P.S. Yes, I have my reasons for doing this. From dallase at nmgi.com Thu Jan 13 23:36:06 2005 From: dallase at nmgi.com (Dallas L. Engelken) Date: Thu, 13 Jan 2005 17:36:06 -0600 Subject: Fedora core 3 kernel upgrades Message-ID: well, i'm following http://bipolar.longbros.com/index.php?blog=5&p=29&c=1&page=1&more=1&tb=1&pb=1&disp=single so my buildinstall is [root at fc3 fedora]# echo $FCBASE /fedora [root at fc3 fedora]# buildinstall --pkgorder $FCBASE/fedora/pkgfile.`date +%Y-%m-%d` \ --version 3 --product 'Fedora Core' --release 'Fedora Core 3' \ --prodpath Fedora $FCBASE/fedora/i386 [root at fc3 fedora]# ls -la $FCBASE/fedora/i386 total 148 drwxr-xr-x 5 root root 4096 Jan 13 16:54 . drwxr-xr-x 4 root root 4096 Jan 13 16:39 .. -rwxr-xr-x 1 root root 248 Nov 3 15:28 autorun -rw-r--r-- 1 root root 82 Jan 13 16:54 .discinfo -rw-r--r-- 1 root root 5611 Nov 3 15:28 eula.txt drwxr-xr-x 4 root root 4096 Jan 12 17:09 Fedora -rw-r--r-- 1 root root 18385 Nov 3 15:28 GPL drwxr-xr-x 2 root root 4096 Jan 13 16:54 images drwxr-xr-x 2 root root 4096 Jan 13 17:00 isolinux -rw-r--r-- 1 root root 871 Jan 13 15:08 ks.cfg -rw-r--r-- 1 root root 1910 Jan 12 17:05 RPM-GPG-KEY -rw-r--r-- 1 root root 1706 Jan 12 17:05 RPM-GPG-KEY-beta -rw-r--r-- 1 root root 1519 Jan 12 17:05 RPM-GPG-KEY-fedora -rw-r--r-- 1 root root 1105 Jan 12 17:05 RPM-GPG-KEY-fedora-rawhide -rw-r--r-- 1 root root 1076 Jan 12 17:05 RPM-GPG-KEY-fedora-test -rw-r--r-- 1 root root 1232 Jan 12 17:05 RPM-GPG-KEY-rawhide my rpms directory contains the following things which are needed.. [root at fc3 fedora]# ls -la /fedora/fedora/i386/Fedora/RPMS/anaconda-* -rw-r--r-- 1 root root 4407316 Jan 13 10:56 /fedora/fedora/i386/Fedora/RPMS/anaconda-10.1.0.2-1.i386.rpm -rw-r--r-- 1 root root 820468 Jan 13 12:11 /fedora/fedora/i386/Fedora/RPMS/anaconda-help-10.1.0-1.noarch.rpm -rw-r--r-- 1 root root 526100 Jan 13 10:56 /fedora/fedora/i386/Fedora/RPMS/anaconda-runtime-10.1.0.2-1.i386.rpm [root at fc3 fedora]# ls -la /fedora/fedora/i386/Fedora/RPMS/kernel-* -rw-r--r-- 1 root root 16690097 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-2.6.10-1.737_FC3.i586.rpm -rw-r--r-- 1 root root 16759557 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-2.6.10-1.737_FC3.i686.rpm -rw-r--r-- 1 root root 2158147 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-doc-2.6.10-1.737_FC3.noarch.rpm -rw-r--r-- 1 root root 16246216 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-smp-2.6.10-1.737_FC3.i586.rpm -rw-r--r-- 1 root root 16423709 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-smp-2.6.10-1.737_FC3.i686.rpm -rw-r--r-- 1 root root 541450 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-utils-2.4-13.1.39.i386.rpm based on this, buildinstall should extract a vmlinuz out of the 2.6.10-1.737_FC3 kernel right and put it in the isolinux/ directory? Thanks, Dallas -----Original Message----- From: kickstart-list-bounces at redhat.com on behalf of Joshua Jensen Sent: Thu 1/13/2005 4:30 PM To: Discussion list about Kickstart Subject: Re: Fedora core 3 kernel upgrades >From your buildtree. What is the exactly command and args that you are using? Joshua On Thu, Jan 13, 2005 at 03:18:58PM -0600, Dallas L. Engelken wrote: > I'm trying to rebuild core 3 cd's with update kernels on them, > specifically 2.6.10-1.737_FC3. When I run through the buildinstall, my > isolinux/vmlinuz file and the modules.cgz in initrd.img still shows that > its 2.6.9-1.667. Where is it getting this from?? I don't even have > 2.6.9-1.667 installed. > > Thanks, > Dallas > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Joshua Jensen Work: 919 392 1488 Mobile: 919 454 9679 Fax: 919 392 8902 Pager: 866 264 2691 Epage: jjensen ? epage.cisco.com IBM Linux Architect - RHCE, RHCT, etc "If God didn't want us to eat animals, why did he make them out of meat?" _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From jamesaharrisonuk at yahoo.co.uk Thu Jan 13 23:41:51 2005 From: jamesaharrisonuk at yahoo.co.uk (James Harrison) Date: Thu, 13 Jan 2005 15:41:51 -0800 (PST) Subject: generic ks.cfg to auto-partition raid0 on X number of disks. In-Reply-To: <200501140343.16909.usergroups@theargoncompany.com> Message-ID: <20050113234151.91885.qmail@web25301.mail.ukl.yahoo.com> > I've tried but I have not understood the what raid.X means.... http://www.redhat.com/docs/manuals/enterprise/RHEL-3-Manual/sysadmin-guide/ch-raid-intro.html --- Rishi wrote: > Hi > > I wanted to know if it is possible to create a generic kickstart file to > auto-partition the disk or disks in the following manner > - First / partition with 1 Gb > - Second partition is 500 MB swap and > - The space on the remaining disk or disks be configured as /backup with > RAID0 > > Issues....... > - Machine 'A' may have just one 40 GB HDD ... > - Machine 'B' may have just two 40 GB HDD ... > - Machine 'C' may have just one 40 GB IDE HDD and one 30 GB SCSI disk. > - Machine 'D' may have three HDD of 20,40 and 40 GB > > Is it possible to create a generic ks.cfg to work for all? > > I've tried but I have not understood the what raid.X means.... > > Any ideas / tips on how to achieve this? > > Thanks > > Regards > > Rishi > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > __________________________________ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail From dallase at nmgi.com Thu Jan 13 23:59:27 2005 From: dallase at nmgi.com (Dallas L. Engelken) Date: Thu, 13 Jan 2005 17:59:27 -0600 Subject: Fedora core 3 kernel upgrades Message-ID: not sure if this helps, but the last few lines of my buildinstall report Making images... /tmp/treedir.10482/instimage/usr/share/hwdata/pcitable doesn't exist /tmp/treedir.10482/instimage/usr/share/hwdata/pci.ids doesn't exist Writing .discinfo file timestamp not specified; using the current time [root at fc3 fedora]# dallas -----Original Message----- From: kickstart-list-bounces at redhat.com on behalf of Dallas L. Engelken Sent: Thu 1/13/2005 5:36 PM To: Discussion list about Kickstart; Discussion list about Kickstart Subject: RE: Fedora core 3 kernel upgrades well, i'm following http://bipolar.longbros.com/index.php?blog=5&p=29&c=1&page=1&more=1&tb=1&pb=1&disp=single so my buildinstall is [root at fc3 fedora]# echo $FCBASE /fedora [root at fc3 fedora]# buildinstall --pkgorder $FCBASE/fedora/pkgfile.`date +%Y-%m-%d` \ --version 3 --product 'Fedora Core' --release 'Fedora Core 3' \ --prodpath Fedora $FCBASE/fedora/i386 [root at fc3 fedora]# ls -la $FCBASE/fedora/i386 total 148 drwxr-xr-x 5 root root 4096 Jan 13 16:54 . drwxr-xr-x 4 root root 4096 Jan 13 16:39 .. -rwxr-xr-x 1 root root 248 Nov 3 15:28 autorun -rw-r--r-- 1 root root 82 Jan 13 16:54 .discinfo -rw-r--r-- 1 root root 5611 Nov 3 15:28 eula.txt drwxr-xr-x 4 root root 4096 Jan 12 17:09 Fedora -rw-r--r-- 1 root root 18385 Nov 3 15:28 GPL drwxr-xr-x 2 root root 4096 Jan 13 16:54 images drwxr-xr-x 2 root root 4096 Jan 13 17:00 isolinux -rw-r--r-- 1 root root 871 Jan 13 15:08 ks.cfg -rw-r--r-- 1 root root 1910 Jan 12 17:05 RPM-GPG-KEY -rw-r--r-- 1 root root 1706 Jan 12 17:05 RPM-GPG-KEY-beta -rw-r--r-- 1 root root 1519 Jan 12 17:05 RPM-GPG-KEY-fedora -rw-r--r-- 1 root root 1105 Jan 12 17:05 RPM-GPG-KEY-fedora-rawhide -rw-r--r-- 1 root root 1076 Jan 12 17:05 RPM-GPG-KEY-fedora-test -rw-r--r-- 1 root root 1232 Jan 12 17:05 RPM-GPG-KEY-rawhide my rpms directory contains the following things which are needed.. [root at fc3 fedora]# ls -la /fedora/fedora/i386/Fedora/RPMS/anaconda-* -rw-r--r-- 1 root root 4407316 Jan 13 10:56 /fedora/fedora/i386/Fedora/RPMS/anaconda-10.1.0.2-1.i386.rpm -rw-r--r-- 1 root root 820468 Jan 13 12:11 /fedora/fedora/i386/Fedora/RPMS/anaconda-help-10.1.0-1.noarch.rpm -rw-r--r-- 1 root root 526100 Jan 13 10:56 /fedora/fedora/i386/Fedora/RPMS/anaconda-runtime-10.1.0.2-1.i386.rpm [root at fc3 fedora]# ls -la /fedora/fedora/i386/Fedora/RPMS/kernel-* -rw-r--r-- 1 root root 16690097 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-2.6.10-1.737_FC3.i586.rpm -rw-r--r-- 1 root root 16759557 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-2.6.10-1.737_FC3.i686.rpm -rw-r--r-- 1 root root 2158147 Jan 13 10:57 /fedora/fedora/i386/Fedora/RPMS/kernel-doc-2.6.10-1.737_FC3.noarch.rpm -rw-r--r-- 1 root root 16246216 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-smp-2.6.10-1.737_FC3.i586.rpm -rw-r--r-- 1 root root 16423709 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-smp-2.6.10-1.737_FC3.i686.rpm -rw-r--r-- 1 root root 541450 Jan 13 10:58 /fedora/fedora/i386/Fedora/RPMS/kernel-utils-2.4-13.1.39.i386.rpm based on this, buildinstall should extract a vmlinuz out of the 2.6.10-1.737_FC3 kernel right and put it in the isolinux/ directory? Thanks, Dallas -----Original Message----- From: kickstart-list-bounces at redhat.com on behalf of Joshua Jensen Sent: Thu 1/13/2005 4:30 PM To: Discussion list about Kickstart Subject: Re: Fedora core 3 kernel upgrades >From your buildtree. What is the exactly command and args that you are using? Joshua On Thu, Jan 13, 2005 at 03:18:58PM -0600, Dallas L. Engelken wrote: > I'm trying to rebuild core 3 cd's with update kernels on them, > specifically 2.6.10-1.737_FC3. When I run through the buildinstall, my > isolinux/vmlinuz file and the modules.cgz in initrd.img still shows that > its 2.6.9-1.667. Where is it getting this from?? I don't even have > 2.6.9-1.667 installed. > > Thanks, > Dallas > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Joshua Jensen Work: 919 392 1488 Mobile: 919 454 9679 Fax: 919 392 8902 Pager: 866 264 2691 Epage: jjensen ? epage.cisco.com IBM Linux Architect - RHCE, RHCT, etc "If God didn't want us to eat animals, why did he make them out of meat?" _______________________________________________ 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 James.Oden at tekelec.com Fri Jan 14 03:34:55 2005 From: James.Oden at tekelec.com (Oden, James) Date: Thu, 13 Jan 2005 21:34:55 -0600 Subject: RPM limitations at kickstart time? Message-ID: <58292FA6B3EEFD49AEDAF6597E21E7178DCBAD@DCEVS2.tekelec.com> > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Hal Wine > Sent: Thursday, January 13, 2005 6:05 PM > To: Kickstart-list at redhat.com > Subject: RPM limitations at kickstart time? > > with rpm 4.2.3-10 (RHEL 3 ES update 3), I'm seeing rpm database > corruption when having an RPM install another RPM in a %post scriptlet > during kickstart. Hey Hal, Could you characterize these corruptions? Do you just have stale locks, or is the data itself corrupted, and in what way if you can tell? > > Supposedly, this is permitted since 4.1-1, and it works fine > outside of > kickstart (anaconda). > Permitted and widely used such that it could claimed to be rock solid are two different things, but then how is going to get used if there aren't guinea pigs (-; Cheers...james From usergroups at theargoncompany.com Fri Jan 14 04:55:07 2005 From: usergroups at theargoncompany.com (Rishi) Date: Fri, 14 Jan 2005 10:25:07 +0530 Subject: generic ks.cfg to auto-partition raid0 on X number of disks. In-Reply-To: <20050113234151.91885.qmail@web25301.mail.ukl.yahoo.com> References: <20050113234151.91885.qmail@web25301.mail.ukl.yahoo.com> Message-ID: <200501141025.07487.usergroups@theargoncompany.com> On Friday 14 Jan 2005 5:11 am, James Harrison wrote: > > I've tried but I have not understood the what raid.X means.... > > http://www.redhat.com/docs/manuals/enterprise/RHEL-3-Manual/sysadmin-guide/ >ch-raid-intro.html :-) I meant I know what RAID means but what does raid.X where X is a number that is used to identify the partition number... This is what I'm using in my ks.cfg clearpart --all part / --fstype ext3 --size=1000 part swap --size=512 part raid.1 --size=1000 --grow part raid.2 --size=1000 --grow part raid.3 --size=1000 --grow raid /backup --fstype ext3 --level=RAID0 raid.1 raid.2 raid.3 The above works fine with one or two disks... But when I add a 3rd disk, it does not auto-magically use that space to include in the /backup partition... Regards Rishi From peter at boku.net Fri Jan 14 09:25:54 2005 From: peter at boku.net (Peter Eisch) Date: Fri, 14 Jan 2005 03:25:54 -0600 Subject: generic ks.cfg to auto-partition raid0 on X number of disks. In-Reply-To: <200501141025.07487.usergroups@theargoncompany.com> Message-ID: On 1/13/05 10:55 PM, "Rishi" wrote: ... > > The above works fine with one or two disks... But when I add a 3rd disk, it > does not auto-magically use that space to include in the /backup partition... > You may find what I do useful. Getting it to what you want for the /backups could be a simple exercise for you from here. %pre #!/bin/sh hds=`awk '{print $4}' < /proc/partitions |grep sd[a-z]$` for file in /proc/ide/h* do mymedia=`cat $file/media` if [ $mymedia == "disk" ] ; then hds="$hds `basename $file`" fi done set $hds numhd=`echo $#` if [ $numhd -le "1" ] ; then #1 drive echo "#partitioning scheme generated in %pre for 1 drive" > /tmp/part-include echo "zerombr yes" >> /tmp/part-include echo "clearpart --all --initlabel" >> /tmp/part-include echo "part /boot --fstype ext3 --size 50" >> /tmp/part-include echo "part swap --recommended" >> /tmp/part-include echo "part / --fstype ext3 --size 1 --grow" >> /tmp/part-include else #2 drives echo "#partitioning scheme generated in %pre for 2 drives" > /tmp/part-include echo "zerombr yes" >> /tmp/part-include echo "clearpart --all --initlabel" >> /tmp/part-include drive=0 bootdevs='' rootdevs='' for pdev in $hds do bootp=raid."$drive"1 rootp=raid."$drive"2 bootdevs="$bootdevs $bootp" rootdevs="$rootdevs $rootp" echo "part $bootp --size 50 --ondisk $pdev" >> /tmp/part-include echo "part $rootp --size 1 --grow --ondisk $pdev" >> /tmp/part-include echo "part swap --recommended --ondisk $pdev" >> /tmp/part-include let drive=$drive+1 done # cannot RAID5 /boot, altogether now, "awwwww" echo "raid /boot --level=1 --device=md0 --fstype ext3 $bootdevs" >> /tmp/part-include if [ $numhd -gt 2 ]; then echo "raid / --level=5 --device=md1 --fstype ext3 $rootdevs" >> /tmp/part-include else echo "raid / --level=1 --device=md1 --fstype ext3 $rootdevs" >> /tmp/part-include fi fi -------------- next part -------------- An HTML attachment was scrubbed... URL: From usergroups at theargoncompany.com Fri Jan 14 12:48:13 2005 From: usergroups at theargoncompany.com (Rishi) Date: Fri, 14 Jan 2005 18:18:13 +0530 Subject: generic ks.cfg to auto-partition raid0 on X number of disks. In-Reply-To: References: Message-ID: <200501141818.13850.usergroups@theargoncompany.com> On Friday 14 Jan 2005 2:55 pm, Peter Eisch wrote: > You may find what I do useful. Getting it to what you want for the > /backups could be a simple exercise for you from here. WOW.. you actually wrote a script to get this done. :-) Pretty neat.. I was trying to avoid writing the script and hope that Red Hat or someone in the open source community to have plugged in such functionality to the kickstart binary itself. Thanks for taking out the time to respond though. I love the community... Long live the community... Regards Rishi From brilong at cisco.com Fri Jan 14 15:16:53 2005 From: brilong at cisco.com (Brian Long) Date: Fri, 14 Jan 2005 10:16:53 -0500 Subject: Is there a way to disable hardware for kickstart? In-Reply-To: <1C455916B3418145B320711B21A2F045A69E79@ex2k.bankofamerica.com> References: <1C455916B3418145B320711B21A2F045A69E79@ex2k.bankofamerica.com> Message-ID: <1105715813.24662.54.camel@brilong-lnx2.cisco.com> On Thu, 2005-01-13 at 16:53, Cameron, Thomas wrote: > Howdy all - > > I am trying to kickstart a RHEL 3 box with both internal disks and SAN > attached storage. Unfortunately the internal storage is > /dev/cciss/c0d0 and the SAN stuff is seen as /dev/sda, /dev/sdb and so > on. The server boots off of internal disks (/dev/cciss/c0d0) but > kickstart wants to write the boot record to the SAN (/dev/sda). We've had similar issues and worked around it in another way. We commented out all "qla" entries in the pcitable on our boot disk images. :) This way the kernel cannot match the PCI IDs on the Qlogic adapters to a kernel module and the kernel module will not load. The other way around it is to use the drive ordering to specify cciss first. I've not tested this, however: bootloader --driveorder=cciss,sda /Brian/ -- Brian Long | | | IT Data Center Systems | .|||. .|||. Cisco Linux Developer | ..:|||||||:...:|||||||:.. Phone: (919) 392-7363 | C i s c o S y s t e m s From katzj at redhat.com Fri Jan 14 15:36:18 2005 From: katzj at redhat.com (Jeremy Katz) Date: Fri, 14 Jan 2005 10:36:18 -0500 Subject: Is there a way to disable hardware for kickstart? In-Reply-To: <1105715813.24662.54.camel@brilong-lnx2.cisco.com> References: <1C455916B3418145B320711B21A2F045A69E79@ex2k.bankofamerica.com> <1105715813.24662.54.camel@brilong-lnx2.cisco.com> Message-ID: <1105716978.4479.20.camel@bree.local.net> On Fri, 2005-01-14 at 10:16 -0500, Brian Long wrote: > On Thu, 2005-01-13 at 16:53, Cameron, Thomas wrote: > > I am trying to kickstart a RHEL 3 box with both internal disks and SAN > > attached storage. Unfortunately the internal storage is > > /dev/cciss/c0d0 and the SAN stuff is seen as /dev/sda, /dev/sdb and so > > on. The server boots off of internal disks (/dev/cciss/c0d0) but > > kickstart wants to write the boot record to the SAN (/dev/sda). [snip] > The other way around it is to use the drive ordering to specify cciss > first. I've not tested this, however: > > bootloader --driveorder=cciss,sda This is the recommended way to do it (syntax would be -- driveorder=cciss/c0d0,sda though) In the future, there's some logic to bump the "known" FC controller modules to the end of the list of modules for ordering. This is will be in RHEL4 and enabled by default and also available for use in RHEL3 U5 if you boot with 'linux latefcload'. Should hopefully make some people's lives easier. Cheers, Jeremy From hal at reactivity.com Fri Jan 14 17:09:25 2005 From: hal at reactivity.com (Hal Wine) Date: Fri, 14 Jan 2005 09:09:25 -0800 Subject: RPM limitations at kickstart time? In-Reply-To: <58292FA6B3EEFD49AEDAF6597E21E7178DCBAD@DCEVS2.tekelec.com> References: <58292FA6B3EEFD49AEDAF6597E21E7178DCBAD@DCEVS2.tekelec.com> Message-ID: <41E7FCC5.7040200@reactivity.com> Oden, James wrote, around 01/13/2005 07:34 PM: >>[mailto:kickstart-list-bounces at redhat.com] On Behalf Of Hal Wine >>Subject: RPM limitations at kickstart time? >> >>with rpm 4.2.3-10 (RHEL 3 ES update 3), I'm seeing rpm database >>corruption when having an RPM install another RPM in a %post scriptlet >>during kickstart. > > Could you characterize these corruptions? Do you just have stale locks, > or is the > data itself corrupted, and in what way if you can tell? It's corrupted so bad that packages fail to install :( The log file /root/install.log showed numerous error messages. Starting with the package _after_ the one that sub-installs another RPM, I see repeated sequences like (92 lines total): Installing libcap-1.10-15.1.i386. error: db4 error(-30989) from dbcursor->c_put: DB_PAGE_NOTFOUND: Requested page not found error: db4 error(-30982) from db->sync: DB_RUNRECOVERY: Fatal error, run database recovery error: db4 error(-30982) from dbcursor->c_close: DB_RUNRECOVERY: Fatal error, run database recovery error: db4 error(-30982) from db->sync: DB_RUNRECOVERY: Fatal error, run database recovery error: db4 error(-30982) from db->cursor: DB_RUNRECOVERY: Fatal error, run database recovery error: db4 error(-30982) from db->get: DB_RUNRECOVERY: Fatal error, run database recovery error: error(-30982) getting "libcap" records from Name index error: db4 error(-30982) from db->sync: DB_RUNRECOVERY: Fatal error, run database recovery After the install completes, and the box reboots, /usr/lib/rpm/rpmdb_verify doesn't report any problems with any db file in /var/lib/rpm There are 3 rpms scheduled to be installed after the one that installs the subpackge. The first two (libpcap, ntp) fail, the final one (comps) succeeds. Installations without the wrapper/subinstall package work fine (and have for eons). In this case, the package to be subinstalled is a stock RHEL one: sendmail-cf. It is installed just find, as is the "wrapper" rpm. Both pass 'rpm -V' after installation. > Permitted and widely used such that it could claimed to be rock solid > are two different things, but then how is going to get used if there > aren't guinea pigs (-; Yah, I've got a *very* ugly workaround that I had been using from the AS 2.1 days (which had too old of a version of RPM to consider this approach). The good news it that the feature works outside of anaconda, so my ugly workaround is now more robust :/ Thanks, --Hal From Thomas.Cameron at bankofamerica.com Fri Jan 14 21:19:27 2005 From: Thomas.Cameron at bankofamerica.com (Cameron, Thomas) Date: Fri, 14 Jan 2005 15:19:27 -0600 Subject: Is there a way to disable hardware for kickstart? Message-ID: <1C455916B3418145B320711B21A2F045A69E9A@ex2k.bankofamerica.com> > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com]On Behalf Of Jeremy Katz > Sent: Friday, January 14, 2005 9:36 AM > To: Discussion list about Kickstart > Subject: Re: Is there a way to disable hardware for kickstart? > > > On Fri, 2005-01-14 at 10:16 -0500, Brian Long wrote: > > On Thu, 2005-01-13 at 16:53, Cameron, Thomas wrote: > > > I am trying to kickstart a RHEL 3 box with both internal > disks and SAN > > > attached storage. Unfortunately the internal storage is > > > /dev/cciss/c0d0 and the SAN stuff is seen as /dev/sda, > /dev/sdb and so > > > on. The server boots off of internal disks (/dev/cciss/c0d0) but > > > kickstart wants to write the boot record to the SAN (/dev/sda). > [snip] > > The other way around it is to use the drive ordering to > specify cciss > > first. I've not tested this, however: > > > > bootloader --driveorder=cciss,sda > > This is the recommended way to do it (syntax would be -- > driveorder=cciss/c0d0,sda though) > > In the future, there's some logic to bump the "known" FC controller > modules to the end of the list of modules for ordering. This > is will be > in RHEL4 and enabled by default and also available for use in RHEL3 U5 > if you boot with 'linux latefcload'. Should hopefully make some > people's lives easier. > > Cheers, > > Jeremy Awesome, thanks very much to both of you for the help! Thomas From shortt at cgicafe.com Fri Jan 14 21:47:09 2005 From: shortt at cgicafe.com (Kevin M. Shortt) Date: Fri, 14 Jan 2005 16:47:09 -0500 (EST) Subject: generic ks.cfg to auto-partition raid0 on X number of disks. In-Reply-To: <200501140343.16909.usergroups@theargoncompany.com> References: <200501140343.16909.usergroups@theargoncompany.com> Message-ID: > > Is it possible to create a generic ks.cfg to work for all? I've just joined this group. I am new to Kickstart. I found this site useful. http://www.redhat.com/docs/manuals/linux/RHL-9-Manual/custom-guide/part-install-info.html > I've tried but I have not understood the what raid.X means.... I think the X is just an arbitrary label for mapping purposes. i.e...I built a machine. two 120 GB disks. RAID1 and Vol grouped. I used RAID 1 for "/". the created a VG and with RAID1 on that. Then I diced the VG up for my filesystems. Here's what the anaconda through into my anaconda-ks.cfg. part raid.24 --size=100 --grow --maxsize=500 --ondisk=hda --asprimary part raid.26 --size=100 --grow --maxsize=500 --ondisk=hdc --asprimary part raid.27 --size=100 --grow --ondisk=hdc part raid.25 --size=100 --grow --ondisk=hda raid / --fstype ext3 --level=RAID1 raid.24 raid.26 raid pv.29 --level=RAID1 raid.25 raid.27 volgroup vg00 pv.29 It appears to start at a number (in my case 24) and continue to increment it for labeling purposes only. It does not appear to be a significance to the number itself. RH expert's please correct me if I am wrong. > Any ideas / tips on how to achieve this? This page illustrates how you may script your ks to adjust to the given hardware.... http://www.redhat.com/docs/manuals/linux/RHL-9-Manual/custom-guide/s1-kickstart2-preinstallconfig.html I hope this helps.. Good luck. -k From jdwarner at gmail.com Mon Jan 17 13:24:28 2005 From: jdwarner at gmail.com (Jon Warner) Date: Mon, 17 Jan 2005 13:24:28 +0000 Subject: ASSERT network device problem Message-ID: Hi Guys, I'm having trouble getting kickstart running properly as part of a PXE install on RHEL3. The pxe boot seems to work fine, and the installer starts correctly, but once you try to manually start a http/ftp/nfs install it refuses to find a network driver. Alt-F3 reveals *Getting kickstart file *ASSERT: no network device in chooseNetworkInterface *no network drivers for doing kickstart. *unable to bring up network *trying to mount device hda The nic in question is an intel e1000. The kickstart script looks like this: #Generated by Kickstart Configurator noprobe device scsi megaraid2 device net e1000 #System language lang en_US #Language modules to install langsupport zh_CN.GB2312 zh_TW.Big5 cs_CZ da_DK nl_NL fr_FR de_DE is_IS it_IT ja_JP.eucJP ko_KR.eucKR no_NO pt_PT ru_RU.k0I8r sl_SI es_ES sv_SE uk_UA --default=en_US #System keyboard keyboard uk #System mouse mouse --emulthree generic3ps/2 #Sytem timezone timezone Europe/London #Root password rootpw --iscrypted $1$/Uj0Wrk6$NOLJ8ShowYlV1rexMN2BO. #Reboot after installation reboot #Install Red Hat Linux instead of upgrade install #Use NFS installation Media nfs --server=10.10.158.84 --dir=/home/pxeboot/rhel3 #System bootloader configuration bootloader --location=mbr #Clear the Master Boot Record zerombr yes #Partition clearing information clearpart --all --initlabel #Disk partitioning information part / --fstype ext3 --size 1 --grow part swap --recommended part /boot --fstype ext3 --size 1024 #System authorization infomation auth --useshadow --enablemd5 #Network information network --bootproto=dhcp --device=eth0 #Firewall configuration firewall --disabled #XWindows configuration information xconfig --depth=32 --resolution=1280x1024 --defaultdesktop=GNOME #Package install information %packages --resolvedeps @ X Window System @ GNOME Desktop Environment @ Editors @ Graphical Internet @ Text-based Internet @ Server Configuration Tools @ DNS Name Server @ FTP Server @ Network Servers @ Development Tools @ Kernel Development @ Administration Tools @ System Tools And is called by a pxelinux config file containing: default rhel-3-es label rhel-3-es kernel rhel3/vmlinuz append ksdevice=eth0 ks=nfs:10.10.158.84:/home/pxeboot/rhel3/ks.cfg ip=dhcp console=tty0 console=ttys0,19200 load_ramdisk=1 initrd=rhel3/initrd.img Any help on this one would be much appreciated. Thanks From phr at doc.ic.ac.uk Mon Jan 17 17:36:53 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Mon, 17 Jan 2005 17:36:53 +0000 (GMT) Subject: ASSERT network device problem In-Reply-To: References: Message-ID: On Mon, 17 Jan 2005, Jon Warner wrote: >I'm having trouble getting kickstart running properly as part of a PXE >install on RHEL3. The pxe boot seems to work fine, and the installer >starts correctly, but once you try to manually start a http/ftp/nfs >install it refuses to find a network driver. > >Alt-F3 reveals > >*Getting kickstart file >*ASSERT: no network device in chooseNetworkInterface >*no network drivers for doing kickstart. >*unable to bring up network >*trying to mount device hda > >The nic in question is an intel e1000. I suspect the flavour of e1000 is not known to the driver. Which RHEL3 update are you using? What are the PCI ids of the card (lspci)? Cheers, Phil From drkludge at cox.net Tue Jan 18 05:18:01 2005 From: drkludge at cox.net (Greg Morgan) Date: Mon, 17 Jan 2005 22:18:01 -0700 Subject: ASSERT network device problem In-Reply-To: References: Message-ID: <41EC9C09.4050408@cox.net> Philip Rowlands wrote: > On Mon, 17 Jan 2005, Jon Warner wrote: > > >>I'm having trouble getting kickstart running properly as part of a PXE >>install on RHEL3. The pxe boot seems to work fine, and the installer >>starts correctly, but once you try to manually start a http/ftp/nfs >>install it refuses to find a network driver. >>The nic in question is an intel e1000. > > > I suspect the flavour of e1000 is not known to the driver. Which RHEL3 > update are you using? What are the PCI ids of the card (lspci)? > > > Cheers, > Phil Yep. If RHEL3 is a 2.4 kernel and the e1000 is a late model NIC, then this explains your problem. Please look here https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=115212. The problem was solved for FC1 by moving to FC2 and a 2.6 kernel. It would take them some work to fix it in a 2.4 kernel as noted in the bug report. It was so rude when I experienced the bug ;-) Regards, Greg From jdwarner at gmail.com Tue Jan 18 10:21:43 2005 From: jdwarner at gmail.com (Jon Warner) Date: Tue, 18 Jan 2005 10:21:43 +0000 Subject: ASSERT network device problem In-Reply-To: <41EC9C09.4050408@cox.net> References: <41EC9C09.4050408@cox.net> Message-ID: On Mon, 17 Jan 2005 22:18:01 -0700, Greg Morgan wrote: > Philip Rowlands wrote: > > On Mon, 17 Jan 2005, Jon Warner wrote: > > > > > >>I'm having trouble getting kickstart running properly as part of a PXE > >>install on RHEL3. The pxe boot seems to work fine, and the installer > >>starts correctly, but once you try to manually start a http/ftp/nfs > >>install it refuses to find a network driver. > >>The nic in question is an intel e1000. > > > > > > I suspect the flavour of e1000 is not known to the driver. Which RHEL3 > > update are you using? What are the PCI ids of the card (lspci)? > > > > > > Cheers, > > Phil > Yep. If RHEL3 is a 2.4 kernel and the e1000 is a late model NIC, then > this explains your problem. Please look here > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=115212 . The > problem was solved for FC1 by moving to FC2 and a 2.6 kernel. It would > take them some work to fix it in a 2.4 kernel as noted in the bug > report. It was so rude when I experienced the bug ;-) > > Regards, > Greg > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > Thanks very much guys, you've hit the nail on the head. I downloaded RHEL3U4 and I've just tried using the ramdisk image and boot kernel from RHEL3U4 cd1 rather than the original RHEL3, and it's started kickstart with no problems. From jct at panonet.net Tue Jan 18 11:48:43 2005 From: jct at panonet.net (JCT) Date: Tue, 18 Jan 2005 11:48:43 +0000 Subject: error: cannot open Packages database in /var/lib/rpm Message-ID: <200501181148.43890.jct@panonet.net> Hi All, I wonder if someone can help. My rpm db was corrupted. So I ran rm -f __db* followed rpm rebuild db which ran OK. But I keep getting the following error. error: cannot open Packages database in /var/lib/rpm warning: kernel-source-2.4.21-20.0.1.EL.i386.rpm: V3 DSA signature: NOKEY, key ID 025e513b error: cannot open Packages database in /var/lib/rp -- Regards JC From jct at panonet.net Tue Jan 18 11:51:28 2005 From: jct at panonet.net (JCT) Date: Tue, 18 Jan 2005 11:51:28 +0000 Subject: PCI card-driver not recognised Message-ID: <200501181151.28260.jct@panonet.net> Hi folks, I am having problems to get the SeaLevel IO card/driver to work with Linux CentOS distribution (Note that CentOS is Binary compatible with Red Hat Enterprise 3). Basically the card is being recognised and shown below by lspci. BUT /proc/devices file is not updated with the new devices. Without the latter the driver cannot communicate with the card. The reason I require the /proc/devices infos is because seaio-load runs a mknod command using informations (device major number and device minor number) stored in /proc/devices. Could you tell me what/where infos are needed to get /proc/devices to be updated. This is rather urgent so any help will be much appreciated. ========= [root at machine]# ./seaio-load Warning: kernel-module version mismatch /opt/relay/lib/modules/seaio.o was compiled for kernel version 2.4.21-20.EL.c0smp while this kernel is version 2.4.21-20.0.1.ELsmp Warning: /opt/relay/lib/modules/seaio.o symbol for parameter break_on_load not found Warning: loading /opt/pmx-relay/lib/modules/seaio.o will taint the kernel: no license See http://www.tux.org/lkml/#export-tainted for information about tainted modules Warning: loading /opt/relay/lib/modules/seaio.o will taint the kernel: forced load Module seaio loaded, with warnings mknod: wrong number of arguments Try `mknod --help' for more information. mknod: wrong number of arguments Try `mknod --help' for more information. mknod: wrong number of arguments Try `mknod --help' for more information. mknod: wrong number of arguments Try `mknod --help' for more information. chgrp: failed to get attributes of `/dev/dio[0-3]': No such file or directory chmod: failed to get attributes of `/dev/dio[0-3]': No such file or directory lsmod output ========== Module Size Used by Tainted: PF seaio 8784 0 (unused) Kernel version :- ============ Linux pdc-svr3 2.4.21-20.0.1.ELsmp #1 SMP Thu Dec 2 23:06:40 GMT 2004 i686 i686 i386 GNU/Linux lspci output ========== 03:06.0 Class ff00: Sealevel Systems Inc: Unknown device 8012 (rev 01) /etc/sysconfig/hwconf output ======================= class: OTHER bus: PCI detached: 0 driver: unknown desc: "Sealevel Systems Inc|unknown device 135e:8012" vendorId: 135e deviceId: 8012 subVendorId: 135e subDeviceId: 8012 pciType: 1 pcibus: 3 pcidev: 6 pcifn: 0 Best Regards -- Regards JC From phr at doc.ic.ac.uk Tue Jan 18 17:13:33 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Tue, 18 Jan 2005 17:13:33 +0000 (GMT) Subject: error: cannot open Packages database in /var/lib/rpm In-Reply-To: <200501181148.43890.jct@panonet.net> References: <200501181148.43890.jct@panonet.net> Message-ID: On Tue, 18 Jan 2005, JCT wrote: >I wonder if someone can help. My rpm db was corrupted. During kickstart? >So I ran rm -f __db* followed rpm rebuild db which ran OK. But I keep >getting the following error. > >error: cannot open Packages database in /var/lib/rpm >warning: kernel-source-2.4.21-20.0.1.EL.i386.rpm: V3 DSA signature: NOKEY, key ID 025e513b >error: cannot open Packages database in /var/lib/rp This isn't really kickstart related (nor is your next posting), so please take this question to the correct list [1], where the audience will be better suited and more knowledgable about rpm issues. Cheers, Phil [1] http://www.redhat.com/mailman/listinfo/rpm-list From Chris.McKenzie at entrust.com Tue Jan 18 23:37:55 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Tue, 18 Jan 2005 18:37:55 -0500 Subject: Stupid KS question about CD file copies Message-ID: Hi all. Trying to copy some files from the boot ISO I created to the local system in the %post of my kickstart. What's the path to the CD? Mounting it is out of the question, the RH install is from cdrom. In my isolinux.cfg I use ks=cdrom:/init/ks.cfg and I know the kickstart is loading, but the %post needs some work to copy files from CD to the system. Any suggestions? Searching has been difficult, "CD, kickstart, copying" are all common terms. Thanks! - Chris -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.j.poole at usa-spaceops.com Wed Jan 19 17:05:02 2005 From: john.j.poole at usa-spaceops.com (Poole, John J) Date: Wed, 19 Jan 2005 12:05:02 -0500 Subject: ASSERT network device problem Message-ID: <2CE66A1ABBED8C4B85DFF5E0860987A32CB8F7@usaflcms03.usa-spaceops.ksc.nasa.gov> Phil, I think it possible to make the e1000 work by applying the patches at the bottom of this bug post. https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=110036 John -----Original Message----- From: Jon Warner [mailto:jdwarner at gmail.com] Sent: Tuesday, January 18, 2005 5:22 AM To: Discussion list about Kickstart Subject: Re: ASSERT network device problem On Mon, 17 Jan 2005 22:18:01 -0700, Greg Morgan wrote: > Philip Rowlands wrote: > > On Mon, 17 Jan 2005, Jon Warner wrote: > > > > > >>I'm having trouble getting kickstart running properly as part of a PXE > >>install on RHEL3. The pxe boot seems to work fine, and the installer > >>starts correctly, but once you try to manually start a http/ftp/nfs > >>install it refuses to find a network driver. > >>The nic in question is an intel e1000. > > > > > > I suspect the flavour of e1000 is not known to the driver. Which RHEL3 > > update are you using? What are the PCI ids of the card (lspci)? > > > > > > Cheers, > > Phil > Yep. If RHEL3 is a 2.4 kernel and the e1000 is a late model NIC, then > this explains your problem. Please look here > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=115212 . The > problem was solved for FC1 by moving to FC2 and a 2.6 kernel. It would > take them some work to fix it in a 2.4 kernel as noted in the bug > report. It was so rude when I experienced the bug ;-) > > Regards, > Greg > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > Thanks very much guys, you've hit the nail on the head. I downloaded RHEL3U4 and I've just tried using the ramdisk image and boot kernel from RHEL3U4 cd1 rather than the original RHEL3, and it's started kickstart with no problems. From dallase at nmgi.com Wed Jan 19 17:36:43 2005 From: dallase at nmgi.com (Dallas L. Engelken) Date: Wed, 19 Jan 2005 11:36:43 -0600 Subject: Error 2 mounting none after kickstart Message-ID: I'm trying to rebuild a fedora core 3 install down to a single disk. I've got it to 400mb and the kickstart works fine and everything installs as expected. However, after it reboots I get mount: error 2 mounting none Switching to new root switchroot: mount failed: 22 umount /initrd/dev failed: 2 Kernel panic - not syncing: Attempted to kill init! So I'm stuck now. If I boot a rescue disk, everything is there and it looks fine. My partition table is just 4 primary partitions, /boot, /, swap, and tmp, and they are ext3. Has anyone seen anything like this? I've seen the "error 2 mounting none" message by searching google, but they all seemed to be related to failed upgrades. Thanks, Dallas From phr at doc.ic.ac.uk Wed Jan 19 19:35:46 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Wed, 19 Jan 2005 19:35:46 +0000 (GMT) Subject: Error 2 mounting none after kickstart In-Reply-To: References: Message-ID: On Wed, 19 Jan 2005, Dallas L. Engelken wrote: >I'm trying to rebuild a fedora core 3 install down to a single disk. >I've got it to 400mb and the kickstart works fine and everything >installs as expected. However, after it reboots I get > >mount: error 2 mounting none >Switching to new root >switchroot: mount failed: 22 >umount /initrd/dev failed: 2 >Kernel panic - not syncing: Attempted to kill init! Looks like the kernel cannot find the root filesystem. Did anaconda create and configure an initrd? It should do if your storage requires loadable kernel modules to access (e.g. SCSI). What "root" argument is the bootloader (GRUB, possibly LILO) passing to the kernel? (Actually what's the entire kernel command-line?) Cheers, Phil From teng at dataway.com Wed Jan 19 22:14:51 2005 From: teng at dataway.com (Tedman Eng) Date: Wed, 19 Jan 2005 14:14:51 -0800 Subject: Kickstart via HTTPS? Message-ID: <37ED92F9890FAF4BB947613C66FF8B1AFC7471@dw-mail.dataway.com> Hello, I am wondering if Kickstart supports HTTPS as an installation method. Or, perhaps someone can suggest another method for securely transferring sensitive information (usernames, passwords, etc) during the install. Any help is much appreciated. From timm at fnal.gov Wed Jan 19 23:18:03 2005 From: timm at fnal.gov (Steven Timm) Date: Wed, 19 Jan 2005 17:18:03 -0600 (CST) Subject: Kickstart via HTTPS? In-Reply-To: <37ED92F9890FAF4BB947613C66FF8B1AFC7471@dw-mail.dataway.com> References: <37ED92F9890FAF4BB947613C66FF8B1AFC7471@dw-mail.dataway.com> Message-ID: On Wed, 19 Jan 2005, Tedman Eng wrote: > Hello, > > I am wondering if Kickstart supports HTTPS as an installation method. > > Or, perhaps someone can suggest another method for securely transferring > sensitive information (usernames, passwords, etc) during the install. > > Any help is much appreciated. I know the Rocks project uses ssl-encrypted kickstart, so it can be done, but I think they added code not in the normal installer. If I remember correctly they don't use https because the overhead is too big, they just do ssl encryption across normal http. If it's just a password or two there is the password --iscrypted command which allows you to put in an encrypted password. Steve > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From jct at panonet.net Thu Jan 20 10:28:43 2005 From: jct at panonet.net (JCT) Date: Thu, 20 Jan 2005 10:28:43 +0000 Subject: Kickstart via HTTPS? In-Reply-To: References: <37ED92F9890FAF4BB947613C66FF8B1AFC7471@dw-mail.dataway.com> Message-ID: <200501201028.43148.jct@panonet.net> Just a thought how about seting up and ssh-tunnel and use http kickstart traffic through the tunnel ? On Wednesday 19 Jan 2005 11:18 pm, Steven Timm wrote: > On Wed, 19 Jan 2005, Tedman Eng wrote: > > Hello, > > > > I am wondering if Kickstart supports HTTPS as an installation method. > > > > Or, perhaps someone can suggest another method for securely transferring > > sensitive information (usernames, passwords, etc) during the install. > > > > Any help is much appreciated. > > I know the Rocks project uses ssl-encrypted kickstart, so it > can be done, but I think they added code not in the normal > installer. If I remember correctly they don't use https because > the overhead is too big, they just do ssl encryption across > normal http. > > If it's just a password or two there is the > > password --iscrypted > > command which allows you to put in an encrypted password. > > Steve > > > _______________________________________________ > > 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 -- Regards JC From teng at dataway.com Thu Jan 20 18:16:49 2005 From: teng at dataway.com (Tedman Eng) Date: Thu, 20 Jan 2005 10:16:49 -0800 Subject: Kickstart via HTTPS? Message-ID: <37ED92F9890FAF4BB947613C66FF8B1AFC7473@dw-mail.dataway.com> Unless the kickstart installer can be made to do so, you'd have to use a seperate machine to create the ssh tunnel to pass the kickstart installation through. Although, ssh/scp is perhaps an interesting possibility to use during the %post% section. -----Original Message----- From: JCT [mailto:jct at panonet.net] Sent: Thursday, January 20, 2005 2:29 AM To: Discussion list about Kickstart; Steven Timm Subject: Re: Kickstart via HTTPS? Just a thought how about seting up and ssh-tunnel and use http kickstart traffic through the tunnel ? On Wednesday 19 Jan 2005 11:18 pm, Steven Timm wrote: > On Wed, 19 Jan 2005, Tedman Eng wrote: > > Hello, > > > > I am wondering if Kickstart supports HTTPS as an installation method. > > > > Or, perhaps someone can suggest another method for securely transferring > > sensitive information (usernames, passwords, etc) during the install. > > > > Any help is much appreciated. > > I know the Rocks project uses ssl-encrypted kickstart, so it > can be done, but I think they added code not in the normal > installer. If I remember correctly they don't use https because > the overhead is too big, they just do ssl encryption across > normal http. > > If it's just a password or two there is the > > password --iscrypted > > command which allows you to put in an encrypted password. > > Steve > > > _______________________________________________ > > 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 -- Regards JC _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From ryan at email.arc.nasa.gov Fri Jan 21 18:35:25 2005 From: ryan at email.arc.nasa.gov (Ryan Nelson) Date: Fri, 21 Jan 2005 10:35:25 -0800 Subject: Monitor detection during KS Message-ID: <41F14B6D.50704@email.arc.nasa.gov> Anyone doing anything nifty for detecting highest resolution of monitors during kickstarting. I ask before I start to look into this myself. The installer doesn't get most of the monitors correct resolutions so they are set low, easy enough to fix although one less set for some one to do the better. Any thoughts, thanks. Ryan From wtriest at chemistry.ohio-state.edu Mon Jan 24 23:01:42 2005 From: wtriest at chemistry.ohio-state.edu (Bill (William) Triest) Date: Mon, 24 Jan 2005 18:01:42 -0500 Subject: Group - minus packages Message-ID: <1106607702.2446.1.camel@hassium.chemistry.ohio-state.edu> Is there an easy way to say install this group, except for this package. I've done some quick searching but I haven't come up with anything. Something along the lines of @ Group Name - Group package I don't want Thanks, Bill From mikem.rtp at gmail.com Mon Jan 24 23:06:30 2005 From: mikem.rtp at gmail.com (Mike McLean) Date: Mon, 24 Jan 2005 18:06:30 -0500 Subject: Group - minus packages In-Reply-To: <1106607702.2446.1.camel@hassium.chemistry.ohio-state.edu> References: <1106607702.2446.1.camel@hassium.chemistry.ohio-state.edu> Message-ID: <4f50e068050124150689c5e22@mail.gmail.com> On Mon, 24 Jan 2005 18:01:42 -0500, Bill (William) Triest wrote: > Is there an easy way to say install this group, except for this > package. I've done some quick searching but I haven't come up with > anything. > > Something along the lines of > @ Group Name > - Group package I don't want Yep that's it exactly. However, note that certain things can still cause the package to be pulled in: 1) if the package is marked as mandatory for the group (or some other group) 2) if the package is required by another package that is being installed. From steve.mah at oracle.com Tue Jan 25 01:30:00 2005 From: steve.mah at oracle.com (Stephen Mah) Date: Mon, 24 Jan 2005 17:30:00 -0800 Subject: %pre script problem Message-ID: <41F5A118.3000709@oracle.com> I have a working kickstart cfg file that I'm trying to change to our "expert" cfg file which basically has user intervention in a text console. I use chvt from 7 to 1 for a brief moment to prompt the user for choices. This worked in RHEL3, but for RHEL4 it seems to have problems with the text based consoles. ie: tty1. The text console appear to have squished text in the upper disply area. It's very microscopic to the point where it's eligible. Has anyone experienced this and knows how to fix this. thanks From Paul.Miles at quadriga.com Wed Jan 26 16:50:39 2005 From: Paul.Miles at quadriga.com (Miles, Paul) Date: Wed, 26 Jan 2005 16:50:39 -0000 Subject: Problems with Fedora core 3 Message-ID: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> Hello, We have been successfully kickstarting custom redhat 9, and fedora core 2 servers for a while now. We use a reduced list of RPMS, and a custom kickstart file. We are now attempting to port our kickstart to fedora core 3 but are having a strange problem. Just after Anaconda has formatted the disks, and is 'preparing to install packages', it fatally dies with the message . We see the same message when we both PXE and CD boot. After a lot of trial and error, We discovered that if we use the hdlist and hdlist2 files directly off the Fedora CD's, the install continues and works correctly rather than the ones we manually generate using genhdlist. I would welcome any suggestions or thoughts. Many thanks in advance, Paul This e-mail is the property of Quadriga Worldwide Ltd The message (and any associated files) is intended only for the use of the individual or entity to which it is addressed and may contain information that is confidential, subject to copyright or constitutes a trade secret. If you are not the intended recipient you are hereby notified that any dissemination, copying or distribution of this message, or files associated with this message, is strictly prohibited. If you have received this message in error, please notify us immediately by replying to the message and deleting it from your computer. Messages sent to and from us may be monitored. Internet communications cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. Therefore, we do not accept responsibility for any errors or omissions that are present in this message, or any attachment, that have arisen as a result of e-mail transmission. If verification is required, please request a hard-copy version. Any views or opinions presented are solely those of the author and do not necessarily represent those of the company. -------------- next part -------------- An HTML attachment was scrubbed... URL: From tibbs at math.uh.edu Wed Jan 26 17:45:25 2005 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: Wed, 26 Jan 2005 11:45:25 -0600 Subject: Problems with Fedora core 3 In-Reply-To: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> (Paul Miles's message of "Wed, 26 Jan 2005 16:50:39 -0000") References: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> Message-ID: >>>>> "PM" == Paul Miles writes: PM> After a lot of trial and error, We discovered that if we use the PM> hdlist and hdlist2 files directly off the Fedora CD's, the install PM> continues and works correctly rather than the ones we manually PM> generate using genhdlist. You know, I just had a problem like that this morning. I keep an updated install tree and this had been working fine until I integrated a recent batch of updates; after that the installer would simply die after the "Preparing RPM Transaction" dialog had completed. Going back to the stock set of packages and hdlists from the DVD fixed the problem. I usually run genhdlist without messing with pkgorder because at least for x86 trees it's not needed. (It's required for FC2-x86_64 to get the glibc install ordering correct.) But I went ahead and ran it, at which point it spewed a bunch of messages about dependency loops. After that, everything installed fine. Here's how I call genhdlist/pkgorder from the python script that keeps my tree updates: os.system("/usr/lib/anaconda-runtime/genhdlist --productpath Fedora %s/../.." % updated_release_dir) os.system("PYTHONPATH=/usr/lib/anaconda /usr/lib/anaconda-runtime/pkgorder %s/../.. i386 Fedora > pkgorder" % updated_release_dir) os.system("/usr/lib/anaconda-runtime/genhdlist --productpath Fedora --fileorder pkgorder %s/../.." % updated_release_dir) which is basically, genhdlist once, then pkgorder, then genhdlist again with the generated order file. Install the anaconda package to get pkgorder. Maybe it will work for you. - J< From steve.mah at oracle.com Wed Jan 26 18:00:17 2005 From: steve.mah at oracle.com (Stephen Mah) Date: Wed, 26 Jan 2005 10:00:17 -0800 Subject: Problems with Fedora core 3 In-Reply-To: References: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> Message-ID: <41F7DAB1.1000908@oracle.com> I had similar problems with RHEL4 beta2. I had to run genhdlist twice. #!/bin/bash DISTRO_TREE=/rhel4-beta2 ## CREATE THE PACKAGE ORDER ### echo "Grab package order into pkgorder.txt" pkgorder $DISTRO_TREE i386 >$DISTRO_TREE/pkgorder.txt genhdlist --withnumbers --fileorder $DISTRO_TREE/pkgorder.txt $DISTRO_TREE #run genhdlist again, Doing it once does not work now for some reason. genhdlist --withnumbers --fileorder $DISTRO_TREE/pkgorder.txt $DISTRO_TREE Jason L Tibbitts III wrote: >>>>>>"PM" == Paul Miles writes: >>>>>> >>>>>> > >PM> After a lot of trial and error, We discovered that if we use the >PM> hdlist and hdlist2 files directly off the Fedora CD's, the install >PM> continues and works correctly rather than the ones we manually >PM> generate using genhdlist. > >You know, I just had a problem like that this morning. I keep an >updated install tree and this had been working fine until I integrated >a recent batch of updates; after that the installer would simply die >after the "Preparing RPM Transaction" dialog had completed. Going >back to the stock set of packages and hdlists from the DVD fixed the >problem. > >I usually run genhdlist without messing with pkgorder because at least >for x86 trees it's not needed. (It's required for FC2-x86_64 to get >the glibc install ordering correct.) But I went ahead and ran it, at >which point it spewed a bunch of messages about dependency loops. >After that, everything installed fine. > >Here's how I call genhdlist/pkgorder from the python script that keeps >my tree updates: > >os.system("/usr/lib/anaconda-runtime/genhdlist --productpath Fedora %s/../.." % updated_release_dir) >os.system("PYTHONPATH=/usr/lib/anaconda /usr/lib/anaconda-runtime/pkgorder %s/../.. i386 Fedora > pkgorder" % updated_release_dir) >os.system("/usr/lib/anaconda-runtime/genhdlist --productpath Fedora --fileorder pkgorder %s/../.." % updated_release_dir) > >which is basically, genhdlist once, then pkgorder, then genhdlist >again with the generated order file. Install the anaconda package to >get pkgorder. > >Maybe it will work for you. > > - J< > >_______________________________________________ >Kickstart-list mailing list >Kickstart-list at redhat.com >https://www.redhat.com/mailman/listinfo/kickstart-list > > From phr at doc.ic.ac.uk Wed Jan 26 18:22:44 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Wed, 26 Jan 2005 18:22:44 +0000 (GMT) Subject: Problems with Fedora core 3 In-Reply-To: <41F7DAB1.1000908@oracle.com> References: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> <41F7DAB1.1000908@oracle.com> Message-ID: On Wed, 26 Jan 2005, Stephen Mah wrote: >I had similar problems with RHEL4 beta2. I had to run genhdlist twice. I believe the "correct" order is genhdlist; pkgorder; genhdlist (Google Knows). I also semi-believe that requiring pkgorder at all is the result of a bug introduced in anaconda such that it always assumes the package list is dependency-sorted, whereas it used to check. I'd have to double-check my facts before swearing to this though... Cheers, Phil From christopher.hotchkiss at gmail.com Thu Jan 27 05:01:02 2005 From: christopher.hotchkiss at gmail.com (Christopher Hotchkiss) Date: Thu, 27 Jan 2005 00:01:02 -0500 Subject: Problems with Fedora core 3 In-Reply-To: References: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> <41F7DAB1.1000908@oracle.com> Message-ID: <7f48492a0501262101420be2c5@mail.gmail.com> On Wed, 26 Jan 2005 18:22:44 +0000 (GMT), Philip Rowlands wrote: > On Wed, 26 Jan 2005, Stephen Mah wrote: > > >I had similar problems with RHEL4 beta2. I had to run genhdlist twice. > > I believe the "correct" order is genhdlist; pkgorder; genhdlist (Google > Knows). I also semi-believe that requiring pkgorder at all is the result > of a bug introduced in anaconda such that it always assumes the package > list is dependency-sorted, whereas it used to check. I'd have to > double-check my facts before swearing to this though... > > Cheers, > Phil > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > I was thinking of doing something similar, do you just copy the updated rpms into the install tree and run the three commands? -- Christopher Hotchkiss (813)960-9273 http://www.post227.org From tibbs at math.uh.edu Thu Jan 27 05:35:03 2005 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: Wed, 26 Jan 2005 23:35:03 -0600 Subject: Problems with Fedora core 3 In-Reply-To: <7f48492a0501262101420be2c5@mail.gmail.com> (Christopher Hotchkiss's message of "Thu, 27 Jan 2005 00:01:02 -0500") References: <5D339AE687B0C8438185AE62ED06697D0EA310@ukchmail01.mail.quadriga.com> <41F7DAB1.1000908@oracle.com> <7f48492a0501262101420be2c5@mail.gmail.com> Message-ID: >>>>> "CH" == Christopher Hotchkiss writes: CH> I was thinking of doing something similar, do you just copy the CH> updated rpms into the install tree and run the three commands? I think you need to remove the old versions as well. You can grab the script I use from http://www.math.uh.edu/~tibbs/do_update You'll need the anaconda and anaconda-runtime packages installed and may need to tweak it a little if you're not on i386 arch. I start a fresh tree by rsyncing from the DVD image, then CD into the new tree, copy in the script and make a symlink (called updates) from the appropriate updates directory of my local Fedora mirror. Then just run the script. - J< From Chris.McKenzie at entrust.com Fri Jan 28 02:23:46 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Thu, 27 Jan 2005 21:23:46 -0500 Subject: Kickstart hangs when removing sendmail Message-ID: Hi all. My kickstart occasionally hangs indefinitely intermittedly when from my %post I call: rpm -e sendmail I subsequently install Postfix, and I absolutely require Sendmail removed before hand. It's inexplicable. Rpm never returns, the %post just sits there. Any ideas? I can't pipe STDERR, and the damn thing never returns. When I do get the hang, I'm already on TTY3, and none of the other TTYs offer any clues. Thanks!!! - Chris -------------- next part -------------- An HTML attachment was scrubbed... URL: From phr at doc.ic.ac.uk Fri Jan 28 04:09:02 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Fri, 28 Jan 2005 04:09:02 +0000 (GMT) Subject: Kickstart hangs when removing sendmail In-Reply-To: References: Message-ID: On Thu, 27 Jan 2005, Chris Mckenzie wrote: >Hi all. > >My kickstart occasionally hangs indefinitely intermittedly when from my >%post I call: > > rpm -e sendmail - Definitely indefinitely? Or just a really long time? - Why not exclude sendmail from the package list in the kickstart file? - ps/pstree on the Alt+F2 VT should show at which point the thing's hung - rpm has various verbose/debugging flags which should help pin it down Cheers, Phil From Chris.McKenzie at entrust.com Fri Jan 28 04:22:21 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Thu, 27 Jan 2005 23:22:21 -0500 Subject: Kickstart hangs when removing sendmail Message-ID: Thanks Phil. - Definitely indefinitely? Or just a really long time? Never comes back. I've waited as long as 20 minutes. - Why not exclude sendmail from the package list in the kickstart file? I didn't. Most probably included as part of a RH package I most definately didn't expect to include a sendmail dep, or anaconda insists on installing it despite not including it. - ps/pstree on the Alt+F2 VT should show at which point the thing's hung Don't know if it'll help. - rpm has various verbose/debugging flags which should help pin it down I'll have to take a closer look. Maybe I can get more out of rpm after all. Thanks. - Chris -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip Rowlands Sent: January 27, 2005 11:09 PM To: Discussion list about Kickstart Subject: Re: Kickstart hangs when removing sendmail On Thu, 27 Jan 2005, Chris Mckenzie wrote: >Hi all. > >My kickstart occasionally hangs indefinitely intermittedly when from my >%post I call: > > rpm -e sendmail - Definitely indefinitely? Or just a really long time? - Why not exclude sendmail from the package list in the kickstart file? - ps/pstree on the Alt+F2 VT should show at which point the thing's hung - rpm has various verbose/debugging flags which should help pin it down Cheers, Phil _______________________________________________ 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 micah.parrish at hp.com Fri Jan 28 04:25:50 2005 From: micah.parrish at hp.com (Micah Parrish) Date: Thu, 27 Jan 2005 21:25:50 -0700 Subject: Kickstart hangs when removing sendmail In-Reply-To: References: Message-ID: <1106886350.7613.9.camel@zx> I think you can exclude sendmail by putting -sendmail in the packages section of your kickstart file. --Micah Parrish micah.parrish at hp.com On Thu, 2005-01-27 at 23:22 -0500, Chris Mckenzie wrote: > Thanks Phil. > > - Definitely indefinitely? Or just a really long time? > > Never comes back. I've waited as long as 20 minutes. > > - Why not exclude sendmail from the package list in the kickstart > file? > > I didn't. Most probably included as part of a RH package I most > definately didn't expect to include a sendmail dep, or anaconda > insists on installing it despite not including it. > > - ps/pstree on the Alt+F2 VT should show at which point the thing's > hung > > Don't know if it'll help. > > - rpm has various verbose/debugging flags which should help pin it > down > > I'll have to take a closer look. Maybe I can get more out of rpm after > all. > > Thanks. > > - Chris > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip > Rowlands > Sent: January 27, 2005 11:09 PM > To: Discussion list about Kickstart > Subject: Re: Kickstart hangs when removing sendmail > > > On Thu, 27 Jan 2005, Chris Mckenzie wrote: > > >Hi all. > > > >My kickstart occasionally hangs indefinitely intermittedly when from > my > >%post I call: > > > > rpm -e sendmail > > - Definitely indefinitely? Or just a really long time? > - Why not exclude sendmail from the package list in the kickstart > file? > - ps/pstree on the Alt+F2 VT should show at which point the thing's > hung > - rpm has various verbose/debugging flags which should help pin it > down > > > Cheers, > Phil > > _______________________________________________ > 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 Chris.McKenzie at entrust.com Fri Jan 28 04:30:04 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Thu, 27 Jan 2005 23:30:04 -0500 Subject: Kickstart hangs when removing sendmail Message-ID: I had tried that early on, but noticed it wasn't getting removed. Deamon was still present and startable. Thanks. - Chris -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Micah Parrish Sent: January 27, 2005 11:26 PM To: Discussion list about Kickstart Subject: RE: Kickstart hangs when removing sendmail I think you can exclude sendmail by putting -sendmail in the packages section of your kickstart file. --Micah Parrish micah.parrish at hp.com On Thu, 2005-01-27 at 23:22 -0500, Chris Mckenzie wrote: > Thanks Phil. > > - Definitely indefinitely? Or just a really long time? > > Never comes back. I've waited as long as 20 minutes. > > - Why not exclude sendmail from the package list in the kickstart > file? > > I didn't. Most probably included as part of a RH package I most > definately didn't expect to include a sendmail dep, or anaconda > insists on installing it despite not including it. > > - ps/pstree on the Alt+F2 VT should show at which point the thing's > hung > > Don't know if it'll help. > > - rpm has various verbose/debugging flags which should help pin it > down > > I'll have to take a closer look. Maybe I can get more out of rpm after > all. > > Thanks. > > - Chris > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip > Rowlands > Sent: January 27, 2005 11:09 PM > To: Discussion list about Kickstart > Subject: Re: Kickstart hangs when removing sendmail > > > On Thu, 27 Jan 2005, Chris Mckenzie wrote: > > >Hi all. > > > >My kickstart occasionally hangs indefinitely intermittedly when from > my > >%post I call: > > > > rpm -e sendmail > > - Definitely indefinitely? Or just a really long time? > - Why not exclude sendmail from the package list in the kickstart > file? > - ps/pstree on the Alt+F2 VT should show at which point the thing's > hung > - rpm has various verbose/debugging flags which should help pin it > down > > > Cheers, > Phil > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: From micah.parrish at hp.com Fri Jan 28 04:35:33 2005 From: micah.parrish at hp.com (Micah Parrish) Date: Thu, 27 Jan 2005 21:35:33 -0700 Subject: Kickstart hangs when removing sendmail In-Reply-To: References: Message-ID: <1106886933.7613.12.camel@zx> It would be helpful if you posted your kickstart file, or as much of it as possible. --Micah Parrish On Thu, 2005-01-27 at 23:30 -0500, Chris Mckenzie wrote: > I had tried that early on, but noticed it wasn't getting removed. > Deamon was still present and startable. > > Thanks. > > - Chris > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Micah Parrish > Sent: January 27, 2005 11:26 PM > To: Discussion list about Kickstart > Subject: RE: Kickstart hangs when removing sendmail > > > I think you can exclude sendmail by putting > > -sendmail > > in the packages section of your kickstart file. > > --Micah Parrish > micah.parrish at hp.com > > On Thu, 2005-01-27 at 23:22 -0500, Chris Mckenzie wrote: > > Thanks Phil. > > > > - Definitely indefinitely? Or just a really long time? > > > > Never comes back. I've waited as long as 20 minutes. > > > > - Why not exclude sendmail from the package list in the kickstart > > file? > > > > I didn't. Most probably included as part of a RH package I most > > definately didn't expect to include a sendmail dep, or anaconda > > insists on installing it despite not including it. > > > > - ps/pstree on the Alt+F2 VT should show at which point the > thing's > > hung > > > > Don't know if it'll help. > > > > - rpm has various verbose/debugging flags which should help pin it > > down > > > > I'll have to take a closer look. Maybe I can get more out of rpm > after > > all. > > > > Thanks. > > > > - Chris > > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com > > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip > > Rowlands > > Sent: January 27, 2005 11:09 PM > > To: Discussion list about Kickstart > > Subject: Re: Kickstart hangs when removing sendmail > > > > > > On Thu, 27 Jan 2005, Chris Mckenzie wrote: > > > > >Hi all. > > > > > >My kickstart occasionally hangs indefinitely intermittedly when > from > > my > > >%post I call: > > > > > > rpm -e sendmail > > > > - Definitely indefinitely? Or just a really long time? > > - Why not exclude sendmail from the package list in the kickstart > > file? > > - ps/pstree on the Alt+F2 VT should show at which point the thing's > > hung > > - rpm has various verbose/debugging flags which should help pin it > > down > > > > > > Cheers, > > Phil > > > > _______________________________________________ > > 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 Paul.Miles at quadriga.com Fri Jan 28 12:14:21 2005 From: Paul.Miles at quadriga.com (Miles, Paul) Date: Fri, 28 Jan 2005 12:14:21 -0000 Subject: Problems with Fedora core 3 Message-ID: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> Hey guys, Thanks for all your help, as I think it was Phil that pointed out, prior to Core 3, you didn't need to run pkgorder - as a result we never included this step in our builds. As I think everyone pointed out, by running genhdlist => pkgorder => genhdlist - all my problems were fixed. One interesting thing to mention though, if you happen to have this misfortune of having your RPMS badly named, ie, not having '.' where they should be, for example somehow we had the RPM called : which_2.16.4_i386.rpm - the '_' broke anaconda and generated warnings in the second genhdlist. Just something to be warey off I guess. Thanks again, Paul This e-mail is the property of Quadriga Worldwide Ltd The message (and any associated files) is intended only for the use of the individual or entity to which it is addressed and may contain information that is confidential, subject to copyright or constitutes a trade secret. If you are not the intended recipient you are hereby notified that any dissemination, copying or distribution of this message, or files associated with this message, is strictly prohibited. If you have received this message in error, please notify us immediately by replying to the message and deleting it from your computer. Messages sent to and from us may be monitored. Internet communications cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. Therefore, we do not accept responsibility for any errors or omissions that are present in this message, or any attachment, that have arisen as a result of e-mail transmission. If verification is required, please request a hard-copy version. Any views or opinions presented are solely those of the author and do not necessarily represent those of the company. From marhoy at gmail.com Fri Jan 28 14:42:42 2005 From: marhoy at gmail.com (=?ISO-8859-1?Q?Martin_H=F8y?=) Date: Fri, 28 Jan 2005 15:42:42 +0100 Subject: Problems with Fedora core 3 In-Reply-To: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> References: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> Message-ID: <93072505012806427019ce59@mail.gmail.com> On Fri, 28 Jan 2005 12:14:21 -0000, Miles, Paul wrote: > Hey guys, > > Thanks for all your help, as I think it was Phil that pointed out, prior > to Core 3, you didn't need to run pkgorder - as a result we never > included this step in our builds. > > As I think everyone pointed out, by running genhdlist => pkgorder => > genhdlist - all my problems were fixed. Would you care to give a detailed explanation (command-line options and everything) of how you updated the install-tree? Or at least give me some pointers to a correct, working explanation for FC3? The number of updates to FC3 is growing so large that I would like to include them in the install-tree so I don't have to way for the "yum update" afterwards. (I'm planning a 200+-machines install of FC3). Regards, Martin From elazar at panix.com Fri Jan 28 15:52:25 2005 From: elazar at panix.com (Elazar Rosenthal) Date: Fri, 28 Jan 2005 10:52:25 -0500 (EST) Subject: Kickstart hangs when removing sendmail In-Reply-To: <1106886933.7613.12.camel@zx> References: <1106886933.7613.12.camel@zx> Message-ID: At least on my FC3 system: sendmail is needed by a number of other packages # rpm -e sendmail error: Failed dependencies: /usr/sbin/sendmail is needed by (installed) redhat-lsb-1.3-4.i386 smtpdaemon is needed by (installed) mdadm-1.6.0-2.i386 It is possible that this dependency is keeping it around Elazar Rosenthal On Thu, 27 Jan 2005, Micah Parrish wrote: > It would be helpful if you posted your kickstart file, or as much of it > as possible. > > --Micah Parrish > > > > On Thu, 2005-01-27 at 23:30 -0500, Chris Mckenzie wrote: >> I had tried that early on, but noticed it wasn't getting removed. >> Deamon was still present and startable. >> >> Thanks. >> >> - Chris >> >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Micah Parrish >> Sent: January 27, 2005 11:26 PM >> To: Discussion list about Kickstart >> Subject: RE: Kickstart hangs when removing sendmail >> >> >> I think you can exclude sendmail by putting >> >> -sendmail >> >> in the packages section of your kickstart file. >> >> --Micah Parrish >> micah.parrish at hp.com >> >> On Thu, 2005-01-27 at 23:22 -0500, Chris Mckenzie wrote: >>> Thanks Phil. >>> >>> - Definitely indefinitely? Or just a really long time? >>> >>> Never comes back. I've waited as long as 20 minutes. >>> >>> - Why not exclude sendmail from the package list in the kickstart >>> file? >>> >>> I didn't. Most probably included as part of a RH package I most >>> definately didn't expect to include a sendmail dep, or anaconda >>> insists on installing it despite not including it. >>> >>> - ps/pstree on the Alt+F2 VT should show at which point the >> thing's >>> hung >>> >>> Don't know if it'll help. >>> >>> - rpm has various verbose/debugging flags which should help pin it >>> down >>> >>> I'll have to take a closer look. Maybe I can get more out of rpm >> after >>> all. >>> >>> Thanks. >>> >>> - Chris >>> >>> -----Original Message----- >>> From: kickstart-list-bounces at redhat.com >>> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip >>> Rowlands >>> Sent: January 27, 2005 11:09 PM >>> To: Discussion list about Kickstart >>> Subject: Re: Kickstart hangs when removing sendmail >>> >>> >>> On Thu, 27 Jan 2005, Chris Mckenzie wrote: >>> >>>> Hi all. >>>> >>>> My kickstart occasionally hangs indefinitely intermittedly when >> from >>> my >>>> %post I call: >>>> >>>> rpm -e sendmail >>> >>> - Definitely indefinitely? Or just a really long time? >>> - Why not exclude sendmail from the package list in the kickstart >>> file? >>> - ps/pstree on the Alt+F2 VT should show at which point the thing's >>> hung >>> - rpm has various verbose/debugging flags which should help pin it >>> down >>> >>> >>> Cheers, >>> Phil >>> >>> _______________________________________________ >>> 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 > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From skvidal at phy.duke.edu Fri Jan 28 15:53:37 2005 From: skvidal at phy.duke.edu (seth vidal) Date: Fri, 28 Jan 2005 10:53:37 -0500 Subject: Kickstart hangs when removing sendmail In-Reply-To: References: <1106886933.7613.12.camel@zx> Message-ID: <1106927617.26754.8.camel@opus.phy.duke.edu> On Fri, 2005-01-28 at 10:52 -0500, Elazar Rosenthal wrote: > At least on my FC3 system: > sendmail is needed by a number of other packages > # rpm -e sendmail > error: Failed dependencies: > /usr/sbin/sendmail is needed by (installed) redhat-lsb-1.3-4.i386 > smtpdaemon is needed by (installed) mdadm-1.6.0-2.i386 > > It is possible that this dependency is keeping it around you need to install SOME mta - but sendmail is not the only pkg providing /usr/sbin/sendmail. postfix and exim do also. -sv From ebrown at lanl.gov Fri Jan 28 16:20:19 2005 From: ebrown at lanl.gov (Ed Brown) Date: Fri, 28 Jan 2005 09:20:19 -0700 Subject: Kickstart hangs when removing sendmail In-Reply-To: <1106927617.26754.8.camel@opus.phy.duke.edu> References: <1106886933.7613.12.camel@zx> <1106927617.26754.8.camel@opus.phy.duke.edu> Message-ID: <1106929219.22952.17.camel@edbrown.lanl.gov> So you might try in %packages: postfix -sendmail Or, in %post: rpm -e sendmail --nodeps But it is the hanging behavior that interests me, and I believe it points to a bug in anaconda/kickstart somewhere. I've also seen hangs when rpm is executing within %post and encounters an error situation. -Ed On Fri, 2005-01-28 at 08:53, seth vidal wrote: > On Fri, 2005-01-28 at 10:52 -0500, Elazar Rosenthal wrote: > > At least on my FC3 system: > > sendmail is needed by a number of other packages > > # rpm -e sendmail > > error: Failed dependencies: > > /usr/sbin/sendmail is needed by (installed) redhat-lsb-1.3-4.i386 > > smtpdaemon is needed by (installed) mdadm-1.6.0-2.i386 > > > > It is possible that this dependency is keeping it around > > you need to install SOME mta - but sendmail is not the only pkg > providing /usr/sbin/sendmail. > > postfix and exim do also. > > -sv > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From tibbs at math.uh.edu Fri Jan 28 16:32:52 2005 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: Fri, 28 Jan 2005 10:32:52 -0600 Subject: Problems with Fedora core 3 In-Reply-To: <93072505012806427019ce59@mail.gmail.com> (Martin =?iso-8859-1?q?H=F8y's?= message of "Fri, 28 Jan 2005 15:42:42 +0100") References: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> <93072505012806427019ce59@mail.gmail.com> Message-ID: >>>>> "MH" == Martin H?y writes: MH> Would you care to give a detailed explanation (command-line MH> options and everything) of how you updated the install-tree You can try the script I sent out earlier, if you want something automatic. - J< From e_chillon at yahoo.com.mx Fri Jan 28 17:31:49 2005 From: e_chillon at yahoo.com.mx (=?iso-8859-1?q?Edgar=20Chill=F3n=20'Chillux'?=) Date: Fri, 28 Jan 2005 11:31:49 -0600 (CST) Subject: Kickstart hangs when removing sendmail In-Reply-To: Message-ID: <20050128173150.13855.qmail@web52210.mail.yahoo.com> I dont remember if postfix requires a sendmail as a dependency but if not here is the solution: In the ks.cfg in the package system: @ Base -mdam -sendmail In the %post rpm -ihv postfix If the postfix needs dependencys obviusly you have to install it fist. Regards --- Chris Mckenzie escribi?: > Hi all. > > My kickstart occasionally hangs indefinitely > intermittedly when from my > %post I call: > > rpm -e sendmail > > I subsequently install Postfix, and I absolutely > require Sendmail removed > before hand. > > It's inexplicable. Rpm never returns, the %post just > sits there. > > Any ideas? I can't pipe STDERR, and the damn thing > never returns. When I do > get the hang, I'm already on TTY3, and none of the > other TTYs offer any > clues. > > Thanks!!! > > - Chris > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list ===== -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- Edgar Alberto Chill?n Esc?rcega Mobile: 04455-1951-9194 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- _________________________________________________________ Do You Yahoo!? La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. http://net.yahoo.com.mx From marhoy at gmail.com Mon Jan 31 15:56:33 2005 From: marhoy at gmail.com (=?ISO-8859-1?Q?Martin_H=F8y?=) Date: Mon, 31 Jan 2005 16:56:33 +0100 Subject: Problems with Fedora core 3 In-Reply-To: References: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> <93072505012806427019ce59@mail.gmail.com> Message-ID: <93072505013107564d3103d1@mail.gmail.com> On Fri, 28 Jan 2005 10:32:52 -0600, Jason L Tibbitts III wrote: > >>>>> "MH" == Martin H?y writes: > > MH> Would you care to give a detailed explanation (command-line > MH> options and everything) of how you updated the install-tree > > You can try the script I sent out earlier, if you want something > automatic. > Thanks, that solved the problem, in the sense that the updated .rpms gets installed at install-time :) But the updated install somehow triggers a bug(?), so that when I reboot the machine after installing, kudzu reports the ethernet card, soundcard and sata-controller as "new hardware" which needs to be configured. Thus, I have to reconfigure network, and press "Configure" on the others. This does not happen when I install from the original FC3-tree. Anyone knows why this would happen? Is it due to different kernel-versions on the boot-cd vs. the install-tree? The people doing the installs in rather non-technical, so this is possibly a show-stopper for me... Martin From tibbs at math.uh.edu Mon Jan 31 16:42:35 2005 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: Mon, 31 Jan 2005 10:42:35 -0600 Subject: Problems with Fedora core 3 In-Reply-To: <93072505013107564d3103d1@mail.gmail.com> (Martin =?iso-8859-1?q?H=F8y's?= message of "Mon, 31 Jan 2005 16:56:33 +0100") References: <5D339AE687B0C8438185AE62ED06697D0EA312@ukchmail01.mail.quadriga.com> <93072505012806427019ce59@mail.gmail.com> <93072505013107564d3103d1@mail.gmail.com> Message-ID: >>>>> "MH" == Martin H?y writes: MH> But the updated install somehow triggers a bug(?), so that when I MH> reboot the machine after installing, kudzu reports the ethernet MH> card, soundcard and sata-controller as "new hardware" which needs MH> to be configured. That's new to me; perhaps the updated kernel detects devices in a different order, so that kudzu thinks they're all new? This wouldn't have been a problem before since the kernel in the boot image would have been the same as the freshly installed one. You could try building fresh boot images too, but unfortunately I can't offer any advice on how to do that. As an experiment, you could try updating everything but the kernel. (I.e. start from a fresh tree from the DVD image, pull down the updates and delete the new kernel packages. Then run the update and try an install.) - J< From Chris.McKenzie at entrust.com Mon Jan 31 19:58:58 2005 From: Chris.McKenzie at entrust.com (Chris Mckenzie) Date: Mon, 31 Jan 2005 14:58:58 -0500 Subject: Kickstart hangs when removing sendmail Message-ID: Thanks for the suggestions guys. We're not installing the RH Postfix. We're using PostgreSQL not MySQL and the Postfix that ships with RH doesn't include support for Postgres. I'm installing Postfix with Postgres support as part of the kickstart %post. My problem is no matter what I try I can't not install sendmail when stick strictly to the %packages portion of the kickstart. %packages --resolvedeps gcc make openssh openssh-clients openssh-server patch perl perl-CGI perl-DBI perl-HTML-Parser perl-libwww-perl perl-URI openssl openssl-perl rdate rpm-build rpmdb-redhat sharutils sudo telnet unzip wget up2date eject -sendmail I know I set --resolvedeps, so something in there wants it (or maybe Anaconda itself simply wants it) If I try ot remove it from %post, it sometimes never returns... rpm -e sendmail --nodeps Note the --nodeps, so I don't get the dependancy errors. Once Postfix is installed, those complaints don't mean a thing, it just works as expected. So I tried the following trickery... RETURN=`rpm -e sendmail --nodeps& 2>&1 > /dev/null` ilimit=60 isleep=1 for ((i=0 ; i < ilimit ; i++)); do RETURN=`rpm -q sendmail | grep "not installed"` if [ "$RETURN" ]; then # Sendmail has finally been removed... echo " -> Sendmail has been removed" break fi done Haven't had time to test it yet, but the idea is if `rpm -e sendmail` never retrurns sometimes, I might as well spawn it as a background process and watch the `rpm -q sendmail` for the removal itself. (I noticed that when I get the hang, a reboot and login shows sendmail not installed) Very frustrating, but I wouldn't care so much if Postfix didn't' need sendmail removed. (their docs claim so, and our testing indiciated some incompatibiltiies) Thanks again for the help. - Chris -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Edgar Chill?n 'Chillux' Sent: January 28, 2005 12:32 PM To: Discussion list about Kickstart Subject: Re: Kickstart hangs when removing sendmail I dont remember if postfix requires a sendmail as a dependency but if not here is the solution: In the ks.cfg in the package system: @ Base -mdam -sendmail In the %post rpm -ihv postfix If the postfix needs dependencys obviusly you have to install it fist. Regards --- Chris Mckenzie escribi?: > Hi all. > > My kickstart occasionally hangs indefinitely > intermittedly when from my > %post I call: > > rpm -e sendmail > > I subsequently install Postfix, and I absolutely > require Sendmail removed > before hand. > > It's inexplicable. Rpm never returns, the %post just > sits there. > > Any ideas? I can't pipe STDERR, and the damn thing > never returns. When I do > get the hang, I'm already on TTY3, and none of the > other TTYs offer any > clues. > > Thanks!!! > > - Chris > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list ===== -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- Edgar Alberto Chill?n Esc?rcega Mobile: 04455-1951-9194 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- PublicKeyID: 0x584CF620 [DH/DSS] http://keyserver.noreply.org/ FingerPrint: 1EDE EF26 C6AA 17DC 1A59 97C8 5270 1942 584C F620 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- _________________________________________________________ Do You Yahoo!? La mejor conexi?n a internet y 25MB extra a tu correo por $100 al mes. http://net.yahoo.com.mx _______________________________________________ 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: