From jim at rossberry.com Thu Jun 1 00:56:38 2006 From: jim at rossberry.com (Jim Wildman) Date: Wed, 31 May 2006 20:56:38 -0400 (EDT) Subject: Configuring wireless adapters In-Reply-To: <447CB096.6040003@redfish-solutions.com> References: <447CB096.6040003@redfish-solutions.com> Message-ID: On Tue, 30 May 2006, Philip Prindeville wrote: > solution is. I'd like to be able to get the file as: > > /etc/sysconfig/network-scripts/ifcfg-iwlan0 > > (or whatever) written out in a semi-automatic way, but I > can't figure out how to do that. > cat > /etc/sysconfig/network-scripts/ifcfg-iwlan0 < References: <447CB096.6040003@redfish-solutions.com> Message-ID: <447E5F52.9050401@redfish-solutions.com> Jim Wildman wrote: >On Tue, 30 May 2006, Philip Prindeville wrote: > > > >>solution is. I'd like to be able to get the file as: >> >>/etc/sysconfig/network-scripts/ifcfg-iwlan0 >> >>(or whatever) written out in a semi-automatic way, but I >>can't figure out how to do that. >> >> >> > >cat > /etc/sysconfig/network-scripts/ifcfg-iwlan0 <neat config stuff >ENDOFIWLAN0 > >comes to mind. > >gogole for 'bash "here document"' > > Um.... Look at Network.write() in /usr/lib/anaconda/network.py I was thinking it would be nice to be able to do this in the same fashion without having to replicate code... Goes back to my earlier comment about being able to generate configs for interfaces that might be absent at install time (such as hot-swap PCMCIA cards, etc). -Philip From drkludge at cox.net Thu Jun 1 06:45:09 2006 From: drkludge at cox.net (Greg Morgan) Date: Wed, 31 May 2006 23:45:09 -0700 Subject: Configuring wireless adapters In-Reply-To: <447E5F52.9050401@redfish-solutions.com> References: <447CB096.6040003@redfish-solutions.com> <447E5F52.9050401@redfish-solutions.com> Message-ID: <447E8CF5.9090402@cox.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Philip Prindeville wrote: > Jim Wildman wrote: > >> On Tue, 30 May 2006, Philip Prindeville wrote: >> >> >> >>> solution is. I'd like to be able to get the file as: >>> >>> /etc/sysconfig/network-scripts/ifcfg-iwlan0 >>> >>> (or whatever) written out in a semi-automatic way, but I >>> can't figure out how to do that. >>> >>> >>> >> cat > /etc/sysconfig/network-scripts/ifcfg-iwlan0 <> neat config stuff >> ENDOFIWLAN0 >> >> comes to mind. >> >> gogole for 'bash "here document"' >> >> > > Um.... Look at Network.write() in /usr/lib/anaconda/network.py > > I was thinking it would be nice to be able to do this in the same fashion > without having to replicate code... Goes back to my earlier comment > about being able to generate configs for interfaces that might be absent > at install time (such as hot-swap PCMCIA cards, etc). Philip, I'd agree with you that code reuse is great. network.py is designed to generate these files from a wide assortment of systems. If you have a limited range of systems to support or they all have the same hardware configuration, then a bash here doc makes a good solution too! The end result of network.py for /etc/sysconfig/network-scripts/ifcfg-eth0 as an example is # My card vendor DEVICE=eth0 BOOTPROTO=static HWADDR=variable mac address IPADDR=ip address NETMASK=your network mask ONBOOT=yes Just cause this looked fun to do, I give you mketh0.sh as an example of what Jim is taking about. This kind-of code in anaconda scripts rocks. #!/bin/bash # Ident: mketh0.sh # # Bash here doc for redhat_eth0. # She's copyright 2006 and free under # the GPL v2 and you get to keep # both pieces if it is broken. # Greg Morgan # NOTE: Using short lines # for sending via email. # # Steps: # save code as mketh0.sh # chmod u+x mketh0.sh # ./mketh0.sh # # Uncomment out for production. # redhat_netdb_path='/etc/sysconfig/network-scripts/' # Use this path for practice to see how it works. # Ya'll be careful hear! # # Define working variables. Please change these. # redhat_netdb_path="${HOME}" redhat_netdb_file='ifcfg-eth0' # Don't forget the leading # hash, #, in your string. my_card_vendor='# MY Vendor' my_device='eth0' my_bootproto='static' # I leave the thrill and joy of discovering # how to dig the mac address out of the # system dynamically to you. # Notice the back tics! # my_hwaddr=`some mystry command` # My shoe size? ;-) my_hwaddr=`echo '13:13:13:13:13:13'` my_ipaddr=192.13.13.13 my_netmask=255.255.255.0 my_onboot=yes # # End of user change section. # So now that the variables are set, # let's create the config file. # Make sure the file is gone first. # We expect this error and it is ok if seen. # rm: cannot remove `/home/drkludge/ifcfg-eth0': No such file or directory # rm ${redhat_netdb_path}/${redhat_netdb_file} # Generate the file so that the data looks # like this excluding leading hash for all # lines except the first one. # My card vendor # DEVICE=eth0 # BOOTPROTO=static # HWADDR=variable mac address # IPADDR=ip address # NETMASK=your network mask # ONBOOT=yes # Improvements on code sent # by Jim Wildman 20060530. # Watch the word wrap in email on this line. cat - > ${redhat_netdb_path}/${redhat_netdb_file} << ETH0EOF ${my_card_vendor} DEVICE=${my_device} BOOTPROTO=${my_bootproto} HWADDR=${my_hwaddr} IPADDR=${my_ipaddr} NETMASK=${my_netmask} ONBOOT=${my_onboot} ETH0EOF # Note do not indent the the code ETH0EOF # to make the code look pretty. The uppercase # code that you pick must start at the # beginning of the line! # # end of mketh0.sh The output of this script looks like this more ifcfg-eth0 # MY Vendor DEVICE=eth0 BOOTPROTO=static HWADDR=13:13:13:13:13:13 IPADDR=192.13.13.13 NETMASK=255.255.255.0 ONBOOT=yes Works fine by me. HTH, Greg -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2.2 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFEfoz1xyxe5L6mr7IRAl+MAJ9/FJ/6wetn3bfIpBDJKwVR6USIwQCfSXOn sx8g996OkVJ4mmgO/yzszHw= =M4OU -----END PGP SIGNATURE----- From neeradm at in.niksun.com Thu Jun 1 23:16:04 2006 From: neeradm at in.niksun.com (Neerad Niksun) Date: Thu, 1 Jun 2006 16:16:04 -0700 Subject: Figuring package class References: <44788DDC.4020209@redfish-solutions.com><00f401c6835b$4f51a090$48003c0a@neeradnew> <447CAD7A.1010907@redfish-solutions.com> Message-ID: <054301c685d1$5b03e9c0$48003c0a@neeradnew> This hdlist live on first CD at /RedHat/base (two files hdlist and hdlist2) Answers to the anticipated questions 1) These files are binary files. 2) Generated with the following command. /usr/lib/anaconda-runtime/genhdlist --withnumbers --fileorder --productpath RedHat --hdlist RedHat/base/hdlist `p wd` (you should be in top directory i.e one above RedHat) This will include all the packages present in your RedHat/RPMS/ folder in the hdlist generated with the above command. (but first you have to build the order file) P.S - all this tested on RHEL 4.0 Neerad Mittal. ----- Original Message ----- From: "Philip Prindeville" To: "Discussion list about Kickstart" Sent: Tuesday, May 30, 2006 1:39 PM Subject: Re: Figuring package class > Ok, and where does hdlist live? > > I was trying to figure out, for instance, why "apr" gets installed... > but can't find any chaining or dependencies. And it's not named > explictly in comps.xml or any of the other xml.gz files. > > -Philip > > > Neerad Niksun wrote: > > comps.xml helps you create your own group. > > RHEL 4.0 has lots of rpms not present in comps.xml and yet those get > > installed. > > > > This is simply done through hdlist. If a package is listed in hdlist (and > > not there in any section of comps.xml) it gets installed. > > > > Neerad > > ----- Original Message ----- > > From: "Philip Prindeville" > > To: > > Sent: Saturday, May 27, 2006 10:35 AM > > Subject: Figuring package class > > > > > > > >> Hi. > >> > >> I was trying to come up with a ks.cfg that excluded certain > >> packages but included others that weren't there by default > >> in the package class... for instance, sendmail-cf and -wpa_supplicant, > >> but I can't tell which class they go into (@base-x???). > >> > >> How do you tell? All of this information must be in the .xml.gz > >> files in repodata/ directory... But I can't figure out how to > >> interpret it. > >> > >> Someone care to set me straight? > >> > >> Thanks, > >> > >> -Philip > >> > >> > >> > >> > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > > > > From philipp_subx at redfish-solutions.com Fri Jun 2 20:07:47 2006 From: philipp_subx at redfish-solutions.com (Philip Prindeville) Date: Fri, 02 Jun 2006 14:07:47 -0600 Subject: Adding --onreboot qualifier to %post section Message-ID: <44809A93.90207@redfish-solutions.com> I was wondering if we wanted to have something like: %post --onreboot ... so that users could specify something they wanted to have done the first time that the machine rebooted (and supposedly entered run-level 5). Or it could be called "--firstboot", but that might cause some confusion. Note that if you didn't have "reboot" in your parameter/header section, this would fail. (I was also thinking that the "reboot" parameter could take an option like "--poweroff".) There are a few things that I'd like to do during regular %post processing... and when complete the machine would power itself off, and be handed to the user to cart away. The user would then power the machine back on some time later, and there would be a certain set of steps that he'd have to do. These would be in the "%post --onreboot" section. Or it could be unattended stuff that doesn't require any interaction, but just takes a while to do and we don't want to have our techies tied up with it (such as running "makewhatis" and "updatedb" for the first time, etc). Does this seem reasonable? It fills the gap between simple scripting and having to write firstboot modules. -Philip From fdd.river at gmail.com Mon Jun 5 22:48:14 2006 From: fdd.river at gmail.com (Frozen River) Date: Mon, 5 Jun 2006 15:48:14 -0700 Subject: =?windows-1252?q?AI_=96_Anaconda_Image?= Message-ID: <6da666750606051548r4d578410jccae018ce9c5bdf8@mail.gmail.com> Hello Folks, Made a customization of kickstart/anaconda and I would like to share it with the rest of the anaconda community. It's essentially a way to bypass the rpm installation during a kickstart and instead using file sets images. What differes from other tools such as system imager is that the hardware doesn't need to be the same. Instead as during any kickstart the hardware is detected and the support (kernel etc) is installed as usual. Hence you can e.g. have the same image for a small development server as for the big production server or even snapshot the production server and deploy it on a smaller test machine to investigate bugs. Anyways let me introduce AI i.e. Anaconda Image, please let me first note that it's a work in progress and that the code is far from polished. AI ? Anaconda Image; A way of rapid kickstart deployment with help of file set images on the same hardware architecture but regardless of hardware. Main ideas are to; - By pass the lengthy time it takes for all rpm packages to be installed on a system - Be able to do customizations on the system and have them ready at install time without help of cumbersome post kickstart scripts or other scripts fetching additional configurations or packages at a post install - Be able to snapshot a system either as a backup or for replication and deploy that snapshot on any hardware as long as the architecture is the same To accomplish this kickstart is preformed on the system as usual with the exception of nearly the entire rpm installation portion of kickstart. Instead just before the rpm installation is to take place a cpio archive is extracted complete with the rpm database and all other files (packages) of the system you want to build (basically you have made compleate backup with help of cpio on a system you want to use as your golden system). After the extraction is complete the necessary rpm packages such as kernel, filesystem support packages and boot loader packages will be installed as normal during a kickstart. The system is by that time ready for use and it has the correct initrd image and kernel along with the support of the filesystem such as e.g. ext3, xfs or what ever you will use. The way to customize kickstart without hacking anaconda it self is to use the special word "%installclass" in the kickstart configuration file. This way the kickstart file is split in two parts and the latter part with you custom python code loaded. This way its relatively easy to overload and extend all functions and procedures present in anaconda. What you do is basically defining a "class CustomKickstart(KickstartBase):" in which you skip nearly all steps in a kickstart and add a few of your own namely: flashpre ? preparation to flash mainly getting hw support info (pretty much a copy of existing code) flashinstall ? extraction of the cpio archive flashpackage ? installation of the packages needed for hardware support (also mostly a copy of anaconda code and butchered) flashpost ? post stuff to write hw conf and similar When appending the "%installclass" a new keyword flash can be used in your kickstart config file. It takes the following arguments --server="IP ADDRESS" which is the server from when you will NFS mount your system image --dir=/where/I/have/my/image/ By default it will look for the directory "RedHat/4" in this dir ? but if it doesn't find it will traverse up. The main server image it looks for is "default" but you can have it to look for image names based in MAC address, "desktop", "server" and so forth and so on. The comments in the code are covering those scenarios pretty extensively. And I will try to explain it further when I get the code a bit more polished. The code is however fully functional and I have used to build 66 servers with a average build time of nine minutes per server :) ? that's fully configured servers with third party applications and ready for immediate use. Best luck the Frozen one... PS: Note however that there are a few anomalies in the RedHat Advance Server Anaconda implementation. The key word "%installclass" is not really supported even if it's fully documented in the code. If you use it strait up and down hence I had to change Kickstart.c in loader2 not just look for pre post and packages in the ksReadCommands function since installclass should also taken into consideration if we do a custom kickstart installation (i.e. we should stop parsing when we reach installclass) Changed kickstart.py so that it don't remove the tmp /tmp/ks.conf.new after invoking custom kick start (we will get an error since the file will not excist later on in the execution) i.e. "#os.unlink(newKsFile)" in kickstart.py I mean it's probably a good idea to keep the ksclass.py why remove something looks very meaning less to me? One strange thing is that I didn't find why we could not do ip=dhcp ? there is definitely something wrong in the parser since (ip=xxxxyyyyssss.eeee,) doesn't work while ip, gateway, netmaskworks. PS: code in the next message... -------------- next part -------------- An HTML attachment was scrubbed... URL: From fdd.river at gmail.com Mon Jun 5 23:05:45 2006 From: fdd.river at gmail.com (Frozen River) Date: Mon, 5 Jun 2006 16:05:45 -0700 Subject: =?windows-1252?q?AI_=96_Anaconda_Image?= Message-ID: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> Sorry folks -- looks like my code is to big for the mailing list. If some moderator could approv my "code" posting it would be very nice. Best wishes the frozen one.. -------------- next part -------------- An HTML attachment was scrubbed... URL: From Colin.Coe at woodside.com.au Tue Jun 6 00:36:34 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Tue, 6 Jun 2006 08:36:34 +0800 Subject: Passing boot params from pxelinux to KickStart Message-ID: <17EBC05307239C4896ED25244F42A80101367985@permls05.wde.woodside.com.au> Thanks. That got it. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Jesse Keating Sent: Friday, 26 May 2006 10:02 PM To: kickstart-list at redhat.com Subject: RE: Passing boot params from pxelinux to KickStart On Fri, 2006-05-26 at 13:26 +0800, Coe, Colin C. wrote: > > label el3 > kernel el3u5_x86_64/vmlinux my_param1 my_param2=1 > append ksdevice=eth0 > ks=nfs:10.240.34.254:/export/install/el3u5_x86_64/rhel_3x86_64_lws_tes > ti > ng.cfg load_ramdisk=1 initrd=el3u5_x86_64/initrd.img > ipappend 1 They need to be on the append line. -- Jesse Keating RHCE (geek.j2solutions.net) Fedora Legacy Team (www.fedoralegacy.org) GPG Public Key (geek.j2solutions.net/jkeating.j2solutions.pub) From Colin.Coe at woodside.com.au Tue Jun 6 00:56:55 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Tue, 6 Jun 2006 08:56:55 +0800 Subject: Selecting NFS home via %pre section Message-ID: <17EBC05307239C4896ED25244F42A80101367986@permls05.wde.woodside.com.au> Hi all I want to select a different NFS install path depending on options passed through PXELINUX. Something like: pxelinux.cfg ------------ label el3_lws_safe kernel el3u5_x86_64/vmlinuz append ksdevice=eth0 ks=nfs:a.b.c.d:/export/install/rhel_x86_64_testing.cfg load_ramdisk=1 initrd=el3u5_x86_64/initrd.img ipappend 1 ks.cfg ------ lang en_US langsupport en_US keyboard us mouse genericwheelps/2 timezone --utc Australia/Perth rootpw --iscrypted reboot install %include /tmp/nfs.cfg bootloader --location=mbr %pre #!/bin/sh -x for I in `cat /proc/cmdline`; do echo $I | grep -q initrd && P=`dirname \`echo $I | cut -d= -f2\`` Done echo "nfs --server=a.b.c.d --dir=/export/install/${P}" > /tmp/nfs.cfg --- However, when I do this, anaconda asks for the install type (CDROM, harddisk, ftp, nfs, etc) and refuses to use NFS. Is this possible? Is anyone else doing anything similar? Thanks CC From saetaes at gmail.com Tue Jun 6 01:25:32 2006 From: saetaes at gmail.com (Mike M) Date: Mon, 5 Jun 2006 21:25:32 -0400 Subject: =?windows-1252?q?Re=3A_AI_=96_Anaconda_Image?= In-Reply-To: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> References: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> Message-ID: <8552c3a30606051825q370ff8c4lbebd9179dfa45d7f@mail.gmail.com> > Sorry folks -- looks like my code is to big for the mailing list. If some > moderator could approv my "code" posting it would be very nice. Someone approve, I'm very interested in this code! :) Frozen River: Do you have a website where you can post it? (I wish I did, or I would offer it up). Mike From saugart at mazunetworks.com Tue Jun 6 02:59:45 2006 From: saugart at mazunetworks.com (Steven Augart) Date: Mon, 05 Jun 2006 22:59:45 -0400 Subject: AI =?windows-1251?q?=96?= Anaconda Image In-Reply-To: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> References: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> Message-ID: <1149562786.32123.17.camel@saugart-t41.mazunetworks.com> On Mon, 2006-06-05 at 16:05 -0700, Frozen River wrote: > Sorry folks -- looks like my code is to big for the mailing list. I am willing to put your code up at my personal web site, http://augart.com. My hosting company has multiple independent connections to the backbone. Also, if we do it that way, then the only people who have to download it are the ones who are interested; when we post it to the list, then effectively everyone has to download it. Further, if you've prepared any RPM packages with the new code, I'll be glad to put those up as well. What was the installation time on these servers before you knocked it down to nine minutes? That will help me evaluate whether it's worth exploring this for use in the process here at Mazu Networks, where we build our network security appliances. For the next release, we'll be using a custom boot CD that I make using a slightly modified version of Anaconda. (I'll post a pointer to the diffs and to a writeup on the process of making it, once I've finished making it work on more hardware platforms. -- we have several here at Mazu.) I am glad that you wrote this. Thanks for being a good citizen and offering it back to the community. --Steven -- Steven Augart Mazu Networks Cambridge, Massachusetts, USA From fdd.river at gmail.com Tue Jun 6 10:42:28 2006 From: fdd.river at gmail.com (Frozen River) Date: Tue, 6 Jun 2006 03:42:28 -0700 Subject: =?windows-1252?q?Re=3A_AI_=96_Anaconda_Image?= In-Reply-To: <1149562786.32123.17.camel@saugart-t41.mazunetworks.com> References: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> <1149562786.32123.17.camel@saugart-t41.mazunetworks.com> Message-ID: <6da666750606060342x74180dddxe2fa832249f6138f@mail.gmail.com> Hello Folks, Sent the code to Steven -- Well there is not really any rpm packages to make -- there is a need for a new stage2 image in the base directory and a new loader in the initial initrd image used by RedHat (This due to clear bugs -- or am I out flying maybe).... Maybe I can rpm it -- as a new anaconda package need to check it out but its definitly not on the top of my priority list. The normal rpm based installation time (just barebone system) was around 20 - 25 min. It's to be noted that the build time of nine minutes was to fully configured system. The build time for a bare bone RedHat installation was more in the five - six minutes range. I have developed a system that is self configuring the server depending on what netgroups its part of. Basically embedded in the config files (such as e.g. passwd, syslog, inittab, http.conf etc..) there is statements such as: if InNetgroup("production") your config for production boxes if InNetgroup("development") your config for development boxes or for e.g. startup scripts if InNetgroup("apache") /etc/init.d/httpd start All of the config files and a number of scripts for the server is handeled this way. The files are kept centraly and distributed out to all system. At distribution time or install time (as in building a server) the files are parsed and installed by a special program. This way all config files are kept in one place and you can easily manage all your boxes. The nine minutes build time I'm refering to is with installing all those files managed by the system (and that is a lot of files since virtually every config file and all startup scripts are managed -- i..e more or less the whole etc directory). Anyways it's pretty practical, all customization I need to do to the bare system fileimage is to have NIS enabled and to run the install program after NIS is started the first time.. but this is beyond the scope of AI.... Cheers The Frozen one.... On 6/5/06, Steven Augart wrote: > > On Mon, 2006-06-05 at 16:05 -0700, Frozen River wrote: > > Sorry folks -- looks like my code is to big for the mailing list. > > I am willing to put your code up at my personal web site, > http://augart.com. My hosting company has multiple independent > connections to the backbone. Also, if we do it that way, then the only > people who have to download it are the ones who are interested; when we > post it to the list, then effectively everyone has to download it. > > Further, if you've prepared any RPM packages with the new code, I'll be > glad to put those up as well. > > What was the installation time on these servers before you knocked it > down to nine minutes? That will help me evaluate whether it's worth > exploring this for use in the process here at Mazu Networks, where we > build our network security appliances. For the next release, we'll be > using a custom boot CD that I make using a slightly modified version of > Anaconda. (I'll post a pointer to the diffs and to a writeup on the > process of making it, once I've finished making it work on more hardware > platforms. -- we have several here at Mazu.) > > I am glad that you wrote this. Thanks for being a good citizen and > offering it back to the community. > > --Steven > > -- > Steven Augart > Mazu Networks > Cambridge, Massachusetts, USA > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From fdd.river at gmail.com Tue Jun 6 10:49:32 2006 From: fdd.river at gmail.com (Frozen River) Date: Tue, 6 Jun 2006 03:49:32 -0700 Subject: Selecting NFS home via %pre section In-Reply-To: <17EBC05307239C4896ED25244F42A80101367986@permls05.wde.woodside.com.au> References: <17EBC05307239C4896ED25244F42A80101367986@permls05.wde.woodside.com.au> Message-ID: <6da666750606060349h175eb001rb049ccb53375105f@mail.gmail.com> I think anaconda is not involved at the stage of selection install method its done in the loader if I'm not totally wrong. The loader is actually parsing the kickstart config file (that how I discovered that it doesn't stop parsing when it hits "%installclass" as it should. Hence you can't do a pre thingy to set install type.... since it needs to be present much earlier Cheers The Frozen one.... On 6/5/06, Coe, Colin C. wrote: > > > Hi all > > I want to select a different NFS install path depending on options > passed through PXELINUX. Something like: > > pxelinux.cfg > ------------ > label el3_lws_safe > kernel el3u5_x86_64/vmlinuz > append ksdevice=eth0 > ks=nfs:a.b.c.d:/export/install/rhel_x86_64_testing.cfg load_ramdisk=1 > initrd=el3u5_x86_64/initrd.img > ipappend 1 > > > ks.cfg > ------ > > lang en_US > langsupport en_US > keyboard us > mouse genericwheelps/2 > timezone --utc Australia/Perth > rootpw --iscrypted > reboot > install > %include /tmp/nfs.cfg > bootloader --location=mbr > > %pre > > #!/bin/sh -x > > for I in `cat /proc/cmdline`; do > echo $I | grep -q initrd && P=`dirname \`echo $I | cut -d= -f2\`` > Done > > echo "nfs --server=a.b.c.d --dir=/export/install/${P}" > /tmp/nfs.cfg > > --- > > However, when I do this, anaconda asks for the install type (CDROM, > harddisk, ftp, nfs, etc) and refuses to use NFS. > > Is this possible? Is anyone else doing anything similar? > > Thanks > > CC > > _______________________________________________ > 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 saugart at mazunetworks.com Tue Jun 6 16:07:55 2006 From: saugart at mazunetworks.com (Steven Augart) Date: Tue, 06 Jun 2006 12:07:55 -0400 Subject: Anaconda Image source code online (was: Re: AI =?windows-1251?q?=96?= Anaconda Image) In-Reply-To: <6da666750606060342x74180dddxe2fa832249f6138f@mail.gmail.com> References: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> <1149562786.32123.17.camel@saugart-t41.mazunetworks.com> <6da666750606060342x74180dddxe2fa832249f6138f@mail.gmail.com> Message-ID: <1149610075.1906.21.camel@saugart-t41.mazunetworks.com> On Tue, 2006-06-06 at 03:42 -0700, Frozen River wrote: > Sent the code to Steven -- It's now up at http://www.augart.com/Frozen_River/Anaconda_Image/ . > Well there is not really any rpm packages to make -- Ah, I see that now. > there is a need for a new stage2 image in the base directory and a > new loader in the initial initrd image used by RedHat (This due to > clear bugs -- or am I out flying maybe).... Yes, there's a clear necessity for that. I repacked the initrd and stage2.img for the FC4-based installer that I'm currently working on, since we needed to work around bugs in the stock FC 4 ones. I have some minor Anaconda bug fixes posted ; mostly support for the installation on an IBM blade server. A .spec file and the patches are at http://www.augart.com/Linux/Anaconda-2006-May/ . However, we've switched over to an FC 4 install instead of an FC 5, and the changes to Anaconda are different. I'll post the source code when I get the FC 4 one fully stable. --Steven -- Steven Augart Mazu Networks Cambridge, Massachusetts, USA From fdd.river at gmail.com Tue Jun 6 22:37:19 2006 From: fdd.river at gmail.com (Frozen River) Date: Tue, 6 Jun 2006 15:37:19 -0700 Subject: =?windows-1252?q?Re=3A_Anaconda_Image_source_code_online_=28was?= =?windows-1252?q?=3A_Re=3A_AI_=96_Anaconda_Image=29?= In-Reply-To: <1149610075.1906.21.camel@saugart-t41.mazunetworks.com> References: <6da666750606051605x466c1f7pbaba2138b7aa59c2@mail.gmail.com> <1149562786.32123.17.camel@saugart-t41.mazunetworks.com> <6da666750606060342x74180dddxe2fa832249f6138f@mail.gmail.com> <1149610075.1906.21.camel@saugart-t41.mazunetworks.com> Message-ID: <6da666750606061537s3451c2a7obd018752bb70e0c8@mail.gmail.com> Hello Steven, Big thanks for posting the code, however I get a error accessing the page -- maybe you can take a look :)... Cheers The Frozen one.. On 6/6/06, Steven Augart wrote: > > On Tue, 2006-06-06 at 03:42 -0700, Frozen River wrote: > > > Sent the code to Steven -- > > It's now up at http://www.augart.com/Frozen_River/Anaconda_Image/ . > > > Well there is not really any rpm packages to make -- > Ah, I see that now. > > > there is a need for a new stage2 image in the base directory and a > > new loader in the initial initrd image used by RedHat (This due to > > clear bugs -- or am I out flying maybe).... > Yes, there's a clear necessity for that. I repacked the initrd and > stage2.img for the FC4-based installer that I'm currently working on, > since we needed to work around bugs in the stock FC 4 ones. > > I have some minor Anaconda bug fixes posted ; mostly support for the > installation on an IBM blade server. A .spec file and the patches are > at http://www.augart.com/Linux/Anaconda-2006-May/ . > > However, we've switched over to an FC 4 install instead of an FC 5, and > the changes to Anaconda are different. I'll post the source code when I > get the FC 4 one fully stable. > > --Steven > > -- > Steven Augart > Mazu Networks > Cambridge, Massachusetts, USA > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From Colin.Coe at woodside.com.au Wed Jun 7 00:30:02 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Wed, 7 Jun 2006 08:30:02 +0800 Subject: Selecting NFS home via %pre section Message-ID: <17EBC05307239C4896ED25244F42A80101367995@permls05.wde.woodside.com.au> Hmmm, OK thanks. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Frozen River Sent: Tuesday, 6 June 2006 6:50 PM To: Discussion list about Kickstart Subject: Re: Selecting NFS home via %pre section I think anaconda is not involved at the stage of selection install method its done in the loader if I'm not totally wrong. The loader is actually parsing the kickstart config file (that how I discovered that it doesn't stop parsing when it hits "%installclass" as it should. Hence you can't do a pre thingy to set install type.... since it needs to be present much earlier Cheers The Frozen one.... On 6/5/06, Coe, Colin C. wrote: Hi all I want to select a different NFS install path depending on options passed through PXELINUX. Something like: pxelinux.cfg ------------ label el3_lws_safe kernel el3u5_x86_64/vmlinuz append ksdevice=eth0 ks=nfs:a.b.c.d:/export/install/rhel_x86_64_testing.cfg load_ramdisk=1 initrd=el3u5_x86_64/initrd.img ipappend 1 ks.cfg ------ lang en_US langsupport en_US keyboard us mouse genericwheelps/2 timezone --utc Australia/Perth rootpw --iscrypted reboot install %include /tmp/nfs.cfg bootloader --location=mbr %pre #!/bin/sh -x for I in `cat /proc/cmdline`; do echo $I | grep -q initrd && P=`dirname \`echo $I | cut -d= -f2\`` Done echo "nfs --server= a.b.c.d --dir=/export/install/${P}" > /tmp/nfs.cfg --- However, when I do this, anaconda asks for the install type (CDROM, harddisk, ftp, nfs, etc) and refuses to use NFS. Is this possible? Is anyone else doing anything similar? Thanks CC _______________________________________________ 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 jbrussard at netnumber.net Wed Jun 7 18:18:45 2006 From: jbrussard at netnumber.net (John Brussard) Date: Wed, 7 Jun 2006 14:18:45 -0400 Subject: FW: Newbie to kickstart - question about loading linux from dvd Message-ID: <200606071818.k57IInKR027578@mx3.redhat.com> I want to install ES4U2 from a kickstart DVD, I have loaded the all the RPMS's onto the dvd as well as the ks.cfg file and changed the isolinux.cfg file. My problem is when it looks for the redhat installations disks, it says: "CD NOT FOUND Red Hat Enterprise Linux CD was not found in any of your CDROM drives" Here's my ks.cfg file: root at ncs-ard1 backup]# more ks.cfg #Generated by Kickstart Configurator #platform=x86, AMD64, or Intel EM64T #System language lang en_SG #Language modules to install langsupport --default=en_US.UTF-8 en_US.UTF-8 #System keyboard keyboard us #System mouse mouse #Sytem timezone timezone --utc America/New_York #Root password rootpw --iscrypted $1$1OOUHcE9$nPOink4oNK7B/1vK06RZl1 bootloader --location=mbr --append="rhgb quiet" #Use text mode install text #Install OS instead of upgrade install #Use CDROM installation media cdrom #System bootloader configuration bootloader --location=mbr #Partitioning clearpart --all --drives=sda,sdb part /boot --fstype ext3 --size=256 --ondisk=sda part /home --fstype ext3 --size=4096 --ondisk=sda part /opt/nnmysql/log --fstype ext3 --size=4096 --ondisk=sda part /opt --fstype ext3 --size=4096 --ondisk=sda part /var --fstype ext3 --size=4096 --ondisk=sda part /usr --fstype ext3 --size=4096 --ondisk=sda part / --fstype ext3 --size=4096 --ondisk=sda part swap --size=2048 --ondisk=sda part /backup --fstype ext3 --size=100 --grow --ondisk=sda part /opt/nnmysql/data --fstype ext3 --size=100 --grow --ondisk=sdb #Clear the Master Boot Record #zerombr yes #System authorization infomation auth --useshadow --enablemd5 #Firewall configuration firewall --disabled #Do not configure XWindows skipx #Package install information %packages --resolvedeps @ admin-tools @ system-tools @ gnome-desktop @ dialup @ compat-arch-support @ base-x @ server-cfg kernel-smp e2fsprogs system-config-boot %post My isolinux.cfg file [root at ncs-ard1 isolinux]# more isolinux.cfg default nnks prompt 0 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg F7 snake.msg label nnks kernel vmlinuz append ks=cdrom:/ks.cfg initrd=initrd.img devfs=nomount ramdisk_size=9216 text directory structure, -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexander_rau at yahoo.com Wed Jun 7 18:25:15 2006 From: alexander_rau at yahoo.com (Alexander Rau) Date: Wed, 7 Jun 2006 11:25:15 -0700 (PDT) Subject: Newbie to kickstart - question about loading linux from dvd In-Reply-To: <200606071818.k57IInKR027578@mx3.redhat.com> Message-ID: <20060607182515.83326.qmail@web52108.mail.yahoo.com> Hey: How did get the RPM's to DVD? do you have the install CD's? just copying the the RPM's to a DVD will not make it a valid install media.... Here is a little how to I followed to create an install DVD for the CD's: mount -o ro,loop RHEL4-i386-WS-disc1.iso /mnt/tmp1 # etc. for discs 2,3,4 cp -a /mnt/tmp1/isolinux /mnt/tmp1/.discinfo . chmod +w isolinux/isolinux.bin {Edit ./.discinfo file to list four CDs "1,2,3,4" instead of just "1"} mkisofs -v -pad -o RHEL4-i386-DVD.iso \ -V "RHEL4-i386-DVD.2005-02-16" \ -b isolinux/isolinux.bin -c isolinux/boot.cat \ -no-emul-boot -boot-load-size 4 -boot-info-table \ -r -R -T -m TRANS.TBL \ -x /mnt/tmp1/.discinfo \ -x /mnt/tmp1/isolinux \ -graft-points /mnt/tmp1 .discinfo=.discinfo isolinux/=isolinux \ RedHat=/mnt/tmp2/RedHat \ RedHat=/mnt/tmp3/RedHat \ RedHat=/mnt/tmp4/RedHat /usr/lib/anaconda-runtime/implantisomd5 RHEL4-i386-DVD.iso (This needs the anaconda-runtime rpm installed) If you want to add additonal packages and a kickstart file to the DVD you actually need to copy the 1st CD to /mnt/tmp1 rather than mounting it. In order to get the kickstart install to prompt for IP address and hostname (even thought it is not in the kickstart file - this is a bug in the Anaconda Installer) you need to do the follwing: take the stage2.img from the CD 1, copy it /tmp or whatever. mkdir /mnt/img mkdir /tmp/new_img mount -o loop /tmp/stage2.img /mnt/img rsync -avv /mnt/img/* /tmp/img/ make sure the hidden .buildstamp file gets copied as well... now you want to look at the /tmp/img/usr/lib/anaconda/kickstart.py and comment out the below line: dispatch.skipStep("network") This will force kickstart to prompt for networking. then create your new stage2.img : mkcramfs /tmp/new_img /tmp/stage2.img.new and copy it back (cp /tmp/stage2.img.new /mnt/tmp1/stage2.img) to your temporary CD making directory, re-create your ISO, burn it and you're good to go. Hope this helps AR ----- Original Message ---- From: John Brussard To: kickstart-list at redhat.com Sent: Wednesday, June 7, 2006 2:18:45 PM Subject: FW: Newbie to kickstart - question about loading linux from dvd I want to install ES4U2 from a kickstart DVD, I have loaded the all the RPMS?s onto the dvd as well as the ks.cfg file and changed the isolinux.cfg file. My problem is when it looks for the redhat installations disks, it says: ?CD NOT FOUND Red Hat Enterprise Linux CD was not found in any of your CDROM drives? Here?s my ks.cfg file: root at ncs-ard1 backup]# more ks.cfg #Generated by Kickstart Configurator #platform=x86, AMD64, or Intel EM64T #System language lang en_SG #Language modules to install langsupport --default=en_US.UTF-8 en_US.UTF-8 #System keyboard keyboard us #System mouse mouse #Sytem timezone timezone --utc America/New_York #Root password rootpw --iscrypted $1$1OOUHcE9$nPOink4oNK7B/1vK06RZl1 bootloader --location=mbr --append="rhgb quiet" #Use text mode install text #Install OS instead of upgrade install #Use CDROM installation media cdrom #System bootloader configuration bootloader --location=mbr #Partitioning clearpart --all --drives=sda,sdb part /boot --fstype ext3 --size=256 --ondisk=sda part /home --fstype ext3 --size=4096 --ondisk=sda part /opt/nnmysql/log --fstype ext3 --size=4096 --ondisk=sda part /opt --fstype ext3 --size=4096 --ondisk=sda part /var --fstype ext3 --size=4096 --ondisk=sda part /usr --fstype ext3 --size=4096 --ondisk=sda part / --fstype ext3 --size=4096 --ondisk=sda part swap --size=2048 --ondisk=sda part /backup --fstype ext3 --size=100 --grow --ondisk=sda part /opt/nnmysql/data --fstype ext3 --size=100 --grow --ondisk=sdb #Clear the Master Boot Record #zerombr yes #System authorization infomation auth --useshadow --enablemd5 #Firewall configuration firewall --disabled #Do not configure XWindows skipx #Package install information %packages --resolvedeps @ admin-tools @ system-tools @ gnome-desktop @ dialup @ compat-arch-support @ base-x @ server-cfg kernel-smp e2fsprogs system-config-boot %post My isolinux.cfg file [root at ncs-ard1 isolinux]# more isolinux.cfg default nnks prompt 0 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg F7 snake.msg label nnks kernel vmlinuz append ks=cdrom:/ks.cfg initrd=initrd.img devfs=nomount ramdisk_size=9216 text directory structure, _______________________________________________ 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 Colin.Coe at woodside.com.au Thu Jun 8 05:39:16 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Thu, 8 Jun 2006 13:39:16 +0800 Subject: Selecting NFS home via %pre section Message-ID: <17EBC05307239C4896ED25244F42A8010136799F@permls05.wde.woodside.com.au> Hi all Still struggling with this... Can I pass anaconda the NFS install path from PXELINUX instead of within the kickstart config file? My target platforms are EL3 and EL4. Also, I've been reading that anaconda can be updated by putting Python modules in ./RHupdates/ directory. Where do I find these Python modules? Thanks CC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Frozen River Sent: Tuesday, 6 June 2006 6:50 PM To: Discussion list about Kickstart Subject: Re: Selecting NFS home via %pre section I think anaconda is not involved at the stage of selection install method its done in the loader if I'm not totally wrong. The loader is actually parsing the kickstart config file (that how I discovered that it doesn't stop parsing when it hits "%installclass" as it should. Hence you can't do a pre thingy to set install type.... since it needs to be present much earlier Cheers The Frozen one.... On 6/5/06, Coe, Colin C. wrote: Hi all I want to select a different NFS install path depending on options passed through PXELINUX. Something like: pxelinux.cfg ------------ label el3_lws_safe kernel el3u5_x86_64/vmlinuz append ksdevice=eth0 ks=nfs:a.b.c.d:/export/install/rhel_x86_64_testing.cfg load_ramdisk=1 initrd=el3u5_x86_64/initrd.img ipappend 1 ks.cfg ------ lang en_US langsupport en_US keyboard us mouse genericwheelps/2 timezone --utc Australia/Perth rootpw --iscrypted reboot install %include /tmp/nfs.cfg bootloader --location=mbr %pre #!/bin/sh -x for I in `cat /proc/cmdline`; do echo $I | grep -q initrd && P=`dirname \`echo $I | cut -d= -f2\`` Done echo "nfs --server= a.b.c.d --dir=/export/install/${P}" > /tmp/nfs.cfg --- However, when I do this, anaconda asks for the install type (CDROM, harddisk, ftp, nfs, etc) and refuses to use NFS. Is this possible? Is anyone else doing anything similar? Thanks CC _______________________________________________ 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 jbrussard at netnumber.net Thu Jun 8 17:41:24 2006 From: jbrussard at netnumber.net (John Brussard) Date: Thu, 8 Jun 2006 13:41:24 -0400 Subject: Newbie to kickstart - question about loading linux from dvd In-Reply-To: <20060607182515.83326.qmail@web52108.mail.yahoo.com> Message-ID: <200606081741.k58HfXhp008228@mx1.redhat.com> Alexander, Thank you very much for this info, it worked, except it is asking for disk3, 4 and 5. Below is my .discinfo file. I want to create a minimal Linux Installation, so I don't need all the disks. Any additional advice? root at ncs-ard1 backup]# more .discinfo 1128193362.766911 Red Hat Enterprise Linux 4 x86_64 1,2 RedHat/base RedHat/RPMS RedHat/pixmaps Regards, John _____ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Alexander Rau Sent: Wednesday, June 07, 2006 2:25 PM To: Discussion list about Kickstart Subject: Re: Newbie to kickstart - question about loading linux from dvd Hey: How did get the RPM's to DVD? do you have the install CD's? just copying the the RPM's to a DVD will not make it a valid install media.... Here is a little how to I followed to create an install DVD for the CD's: mount -o ro,loop RHEL4-i386-WS-disc1.iso /mnt/tmp1 # etc. for discs 2,3,4 cp -a /mnt/tmp1/isolinux /mnt/tmp1/.discinfo . chmod +w isolinux/isolinux.bin {Edit ./.discinfo file to list four CDs "1,2,3,4" instead of just "1"} mkisofs -v -pad -o RHEL4-i386-DVD.iso \ -V "RHEL4-i386-DVD.2005-02-16" \ -b isolinux/isolinux.bin -c isolinux/boot.cat \ -no-emul-boot -boot-load-size 4 -boot-info-table \ -r -R -T -m TRANS.TBL \ -x /mnt/tmp1/.discinfo \ -x /mnt/tmp1/isolinux \ -graft-points /mnt/tmp1 .discinfo=.discinfo isolinux/=isolinux \ RedHat=/mnt/tmp2/RedHat \ RedHat=/mnt/tmp3/RedHat \ RedHat=/mnt/tmp4/RedHat /usr/lib/anaconda-runtime/implantisomd5 RHEL4-i386-DVD.iso (This needs the anaconda-runtime rpm installed) If you want to add additonal packages and a kickstart file to the DVD you actually need to copy the 1st CD to /mnt/tmp1 rather than mounting it. In order to get the kickstart install to prompt for IP address and hostname (even thought it is not in the kickstart file - this is a bug in the Anaconda Installer) you need to do the follwing: take the stage2.img from the CD 1, copy it /tmp or whatever. mkdir /mnt/img mkdir /tmp/new_img mount -o loop /tmp/stage2.img /mnt/img rsync -avv /mnt/img/* /tmp/img/ make sure the hidden .buildstamp file gets copied as well... now you want to look at the /tmp/img/usr/lib/anaconda/kickstart.py and comment out the below line: dispatch.skipStep("network") This will force kickstart to prompt for networking. then create your new stage2.img : mkcramfs /tmp/new_img /tmp/stage2.img.new and copy it back (cp /tmp/stage2.img.new /mnt/tmp1/stage2.img) to your temporary CD making directory, re-create your ISO, burn it and you're good to go. Hope this helps AR ----- Original Message ---- From: John Brussard To: kickstart-list at redhat.com Sent: Wednesday, June 7, 2006 2:18:45 PM Subject: FW: Newbie to kickstart - question about loading linux from dvd I want to install ES4U2 from a kickstart DVD, I have loaded the all the RPMS's onto the dvd as well as the ks.cfg file and changed the isolinux.cfg file. My problem is when it looks for the redhat installations disks, it says: "CD NOT FOUND Red Hat Enterprise Linux CD was not found in any of your CDROM drives" Here's my ks.cfg file: root at ncs-ard1 backup]# more ks.cfg #Generated by Kickstart Configurator #platform=x86, AMD64, or Intel EM64T #System language lang en_SG #Language modules to install langsupport --default=en_US.UTF-8 en_US.UTF-8 #System keyboard keyboard us #System mouse mouse #Sytem timezone timezone --utc America/New_York #Root password rootpw --iscrypted $1$1OOUHcE9$nPOink4oNK7B/1vK06RZl1 bootloader --location=mbr --append="rhgb quiet" #Use text mode install text #Install OS instead of upgrade install #Use CDROM installation media cdrom #System bootloader configuration bootloader --location=mbr #Partitioning clearpart --all --drives=sda,sdb part /boot --fstype ext3 --size=256 --ondisk=sda part /home --fstype ext3 --size=4096 --ondisk=sda part /opt/nnmysql/log --fstype ext3 --size=4096 --ondisk=sda part /opt --fstype ext3 --size=4096 --ondisk=sda part /var --fstype ext3 --size=4096 --ondisk=sda part /usr --fstype ext3 --size=4096 --ondisk=sda part / --fstype ext3 --size=4096 --ondisk=sda part swap --size=2048 --ondisk=sda part /backup --fstype ext3 --size=100 --grow --ondisk=sda part /opt/nnmysql/data --fstype ext3 --size=100 --grow --ondisk=sdb #Clear the Master Boot Record #zerombr yes #System authorization infomation auth --useshadow --enablemd5 #Firewall configuration firewall --disabled #Do not configure XWindows skipx #Package install information %packages --resolvedeps @ admin-tools @ system-tools @ gnome-desktop @ dialup @ compat-arch-support @ base-x @ server-cfg kernel-smp e2fsprogs system-config-boot %post My isolinux.cfg file [root at ncs-ard1 isolinux]# more isolinux.cfg default nnks prompt 0 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg F7 snake.msg label nnks kernel vmlinuz append ks=cdrom:/ks.cfg initrd=initrd.img devfs=nomount ramdisk_size=9216 text directory structure, _______________________________________________ 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 mrauschkolb at ebs.com Thu Jun 8 17:51:19 2006 From: mrauschkolb at ebs.com (Rauschkolb, Mark) Date: Thu, 8 Jun 2006 13:51:19 -0400 Subject: Newbie to kickstart - question about loading linux from dvd Message-ID: change the 1,2 line to be 1,2,3,4,5 -----Original Message----- From: John Brussard [mailto:jbrussard at netnumber.net] Sent: Thursday, June 08, 2006 1:41 PM To: 'Alexander Rau'; 'Discussion list about Kickstart' Subject: RE: Newbie to kickstart - question about loading linux from dvd Alexander, Thank you very much for this info, it worked, except it is asking for disk3, 4 and 5. Below is my .discinfo file. I want to create a minimal Linux Installation, so I don't need all the disks. Any additional advice? root at ncs-ard1 backup]# more .discinfo 1128193362.766911 Red Hat Enterprise Linux 4 x86_64 1,2 RedHat/base RedHat/RPMS RedHat/pixmaps Regards, John ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Alexander Rau Sent: Wednesday, June 07, 2006 2:25 PM To: Discussion list about Kickstart Subject: Re: Newbie to kickstart - question about loading linux from dvd Hey: How did get the RPM's to DVD? do you have the install CD's? just copying the the RPM's to a DVD will not make it a valid install media.... Here is a little how to I followed to create an install DVD for the CD's: mount -o ro,loop RHEL4-i386-WS-disc1.iso /mnt/tmp1 # etc. for discs 2,3,4 cp -a /mnt/tmp1/isolinux /mnt/tmp1/.discinfo . chmod +w isolinux/isolinux.bin {Edit ./.discinfo file to list four CDs "1,2,3,4" instead of just "1"} mkisofs -v -pad -o RHEL4-i386-DVD.iso \ -V "RHEL4-i386-DVD.2005-02-16" \ -b isolinux/isolinux.bin -c isolinux/boot.cat \ -no-emul-boot -boot-load-size 4 -boot-info-table \ -r -R -T -m TRANS.TBL \ -x /mnt/tmp1/.discinfo \ -x /mnt/tmp1/isolinux \ -graft-points /mnt/tmp1 .discinfo=.discinfo isolinux/=isolinux \ RedHat=/mnt/tmp2/RedHat \ RedHat=/mnt/tmp3/RedHat \ RedHat=/mnt/tmp4/RedHat /usr/lib/anaconda-runtime/implantisomd5 RHEL4-i386-DVD.iso (This needs the anaconda-runtime rpm installed) If you want to add additonal packages and a kickstart file to the DVD you actually need to copy the 1st CD to /mnt/tmp1 rather than mounting it. In order to get the kickstart install to prompt for IP address and hostname (even thought it is not in the kickstart file - this is a bug in the Anaconda Installer) you need to do the follwing: take the stage2.img from the CD 1, copy it /tmp or whatever. mkdir /mnt/img mkdir /tmp/new_img mount -o loop /tmp/stage2.img /mnt/img rsync -avv /mnt/img/* /tmp/img/ make sure the hidden .buildstamp file gets copied as well... now you want to look at the /tmp/img/usr/lib/anaconda/kickstart.py and comment out the below line: dispatch.skipStep("network") This will force kickstart to prompt for networking. then create your new stage2.img : mkcramfs /tmp/new_img /tmp/stage2.img.new and copy it back (cp /tmp/stage2.img.new /mnt/tmp1/stage2.img) to your temporary CD making directory, re-create your ISO, burn it and you're good to go. Hope this helps AR ----- Original Message ---- From: John Brussard To: kickstart-list at redhat.com Sent: Wednesday, June 7, 2006 2:18:45 PM Subject: FW: Newbie to kickstart - question about loading linux from dvd I want to install ES4U2 from a kickstart DVD, I have loaded the all the RPMS's onto the dvd as well as the ks.cfg file and changed the isolinux.cfg file. My problem is when it looks for the redhat installations disks, it says: "CD NOT FOUND Red Hat Enterprise Linux CD was not found in any of your CDROM drives" Here's my ks.cfg file: root at ncs-ard1 backup]# more ks.cfg #Generated by Kickstart Configurator #platform=x86, AMD64, or Intel EM64T #System language lang en_SG #Language modules to install langsupport --default=en_US.UTF-8 en_US.UTF-8 #System keyboard keyboard us #System mouse mouse #Sytem timezone timezone --utc America/New_York #Root password rootpw --iscrypted $1$1OOUHcE9$nPOink4oNK7B/1vK06RZl1 bootloader --location=mbr --append="rhgb quiet" #Use text mode install text #Install OS instead of upgrade install #Use CDROM installation media cdrom #System bootloader configuration bootloader --location=mbr #Partitioning clearpart --all --drives=sda,sdb part /boot --fstype ext3 --size=256 --ondisk=sda part /home --fstype ext3 --size=4096 --ondisk=sda part /opt/nnmysql/log --fstype ext3 --size=4096 --ondisk=sda part /opt --fstype ext3 --size=4096 --ondisk=sda part /var --fstype ext3 --size=4096 --ondisk=sda part /usr --fstype ext3 --size=4096 --ondisk=sda part / --fstype ext3 --size=4096 --ondisk=sda part swap --size=2048 --ondisk=sda part /backup --fstype ext3 --size=100 --grow --ondisk=sda part /opt/nnmysql/data --fstype ext3 --size=100 --grow --ondisk=sdb #Clear the Master Boot Record #zerombr yes #System authorization infomation auth --useshadow --enablemd5 #Firewall configuration firewall --disabled #Do not configure XWindows skipx #Package install information %packages --resolvedeps @ admin-tools @ system-tools @ gnome-desktop @ dialup @ compat-arch-support @ base-x @ server-cfg kernel-smp e2fsprogs system-config-boot %post My isolinux.cfg file [root at ncs-ard1 isolinux]# more isolinux.cfg default nnks prompt 0 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg F7 snake.msg label nnks kernel vmlinuz append ks=cdrom:/ks.cfg initrd=initrd.img devfs=nomount ramdisk_size=9216 text directory structure, _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list Thank you for being part of it. The information contained in this e-mail is confidential. This e-mail is intended only for the stated addressee. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. if you have received this e-mail in error, please inform us immediately and delete it and all copies from your system. EBS Dealing Resources International Limited. Registered address: 10 Paternoster Square, London EC4M 7DY, United Kingdom. Registered number 2669861. EBS Dealing Resources, Inc, registered in Delaware. Address: 535 Madison Avenue, 24th Floor, New York, NY 10022, USA, and One upper Pond road, Building F - Floor 3, Parsippany, NJ 07054, USA. EBS Dealing Resources Japan Limited, a Japanese Corporation. Address: Asteer Kayabacho Bldg, 6th Floor, 1-6-1, Shinkawa, Chuo-Ku, Tokyo 104-0033, Japan. -------------- next part -------------- An HTML attachment was scrubbed... URL: From fdd.river at gmail.com Thu Jun 8 22:36:34 2006 From: fdd.river at gmail.com (Frozen River) Date: Thu, 8 Jun 2006 15:36:34 -0700 Subject: Selecting NFS home via %pre section In-Reply-To: <17EBC05307239C4896ED25244F42A8010136799F@permls05.wde.woodside.com.au> References: <17EBC05307239C4896ED25244F42A8010136799F@permls05.wde.woodside.com.au> Message-ID: <6da666750606081536m4a36cae6uec74b15a2e38c16e@mail.gmail.com> This is the modules in the stage2 image in the base directory -- i.e. the anaconda modules written in pyton. They are down under the /usr/lib/anaconda in the image. I suggest using %install classes instead although since thats they prefered way of customizing anaconda. Cheers Olof On 6/7/06, Coe, Colin C. wrote: > > Hi all > > Still struggling with this... Can I pass anaconda the NFS install path > from PXELINUX instead of within the kickstart config file? My target > platforms are EL3 and EL4. > > Also, I've been reading that anaconda can be updated by putting Python > modules in ./RHupdates/ directory. Where do I find these Python modules? > > Thanks > > CC > > -----Original Message----- > *From:* kickstart-list-bounces at redhat.com [mailto: > kickstart-list-bounces at redhat.com] *On Behalf Of *Frozen River > *Sent:* Tuesday, 6 June 2006 6:50 PM > *To:* Discussion list about Kickstart > *Subject:* Re: Selecting NFS home via %pre section > > I think anaconda is not involved at the stage of selection install method > its done in the loader if I'm not totally wrong. The loader is actually > parsing the kickstart config file (that how I discovered that it doesn't > stop parsing when it hits "%installclass" as it should. Hence you can't do a > pre thingy to set install type.... since it needs to be present much earlier > > > Cheers The Frozen one.... > > > > > On 6/5/06, Coe, Colin C. wrote: > > > > > > Hi all > > > > I want to select a different NFS install path depending on options > > passed through PXELINUX. Something like: > > > > pxelinux.cfg > > ------------ > > label el3_lws_safe > > kernel el3u5_x86_64/vmlinuz > > append ksdevice=eth0 > > ks=nfs:a.b.c.d:/export/install/rhel_x86_64_testing.cfg load_ramdisk=1 > > initrd=el3u5_x86_64/initrd.img > > ipappend 1 > > > > > > ks.cfg > > ------ > > > > lang en_US > > langsupport en_US > > keyboard us > > mouse genericwheelps/2 > > timezone --utc Australia/Perth > > rootpw --iscrypted > > reboot > > install > > %include /tmp/nfs.cfg > > bootloader --location=mbr > > > > %pre > > > > #!/bin/sh -x > > > > for I in `cat /proc/cmdline`; do > > echo $I | grep -q initrd && P=`dirname \`echo $I | cut -d= -f2\`` > > Done > > > > echo "nfs --server= a.b.c.d --dir=/export/install/${P}" > /tmp/nfs.cfg > > > > --- > > > > However, when I do this, anaconda asks for the install type (CDROM, > > harddisk, ftp, nfs, etc) and refuses to use NFS. > > > > Is this possible? Is anyone else doing anything similar? > > > > Thanks > > > > CC > > > > _______________________________________________ > > 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 fdd.river at gmail.com Sat Jun 10 11:24:42 2006 From: fdd.river at gmail.com (Frozen River) Date: Sat, 10 Jun 2006 04:24:42 -0700 Subject: AI code.. and email website Message-ID: <6da666750606100424n4fd12e8bh416291ae442aa706@mail.gmail.com> Hello Folks, Yes it's not a pretty one but at least you folks can download it. The URL is http://frozenriver00.tripod.com/index.htm Cheers The Frozen one -------------- next part -------------- An HTML attachment was scrubbed... URL: From mariogzuk at technikz.de Mon Jun 12 13:26:22 2006 From: mariogzuk at technikz.de (Mario Gzuk) Date: Mon, 12 Jun 2006 15:26:22 +0200 Subject: Exisiting partitions - ntfs Message-ID: <1150118782.11564.18.camel@alpha.bgc-jena.mpg.de> Hi, is there a way to specify a existing ntfs partition through the kickstart-file? Any hints are welcome. greetings mario gzuk From philipp_subx at redfish-solutions.com Mon Jun 12 16:20:18 2006 From: philipp_subx at redfish-solutions.com (Philip Prindeville) Date: Mon, 12 Jun 2006 10:20:18 -0600 Subject: Issues bringing up wireless in %post script Message-ID: <448D9442.8050706@redfish-solutions.com> I have a Dell Latitude L610 that I'm trying to write a prototype customer installer for FC5 for. It has an Intel 2915ABG NIC. When I try to load up the ipw2200 module from the %post script, I see: % modprobe -v ipw2200 insmod /lib/modules/2.6.15-1.2054_FC5/kernel/net/ieee80211/ie80211_crypt.ko WARNING: Error inserting ieee80211_crypt (...): Operation not permitted ... same for ieee80211.ko and ipw2200.ko ... even though they are present on the system. Any idea why this is happening? I've installed the ipw2200-firmware RPM before trying to bring any of this up... I burned it onto the installer DVD, and added it to the %packages section of the ks.cfg file. Clues? Thanks, -Philip From Chip.Shabazian at bankofamerica.com Mon Jun 12 16:39:57 2006 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 12 Jun 2006 09:39:57 -0700 Subject: Issues bringing up wireless in %post script In-Reply-To: <448D9442.8050706@redfish-solutions.com> Message-ID: You aren't running that kernel until after your reboot. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip Prindeville Sent: Monday, June 12, 2006 9:20 AM To: Discussion list about Kickstart Subject: Issues bringing up wireless in %post script I have a Dell Latitude L610 that I'm trying to write a prototype customer installer for FC5 for. It has an Intel 2915ABG NIC. When I try to load up the ipw2200 module from the %post script, I see: % modprobe -v ipw2200 insmod /lib/modules/2.6.15-1.2054_FC5/kernel/net/ieee80211/ie80211_crypt.ko WARNING: Error inserting ieee80211_crypt (...): Operation not permitted ... same for ieee80211.ko and ipw2200.ko ... even though they are present on the system. Any idea why this is happening? I've installed the ipw2200-firmware RPM before trying to bring any of this up... I burned it onto the installer DVD, and added it to the %packages section of the ks.cfg file. Clues? Thanks, -Philip _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From philipp_subx at redfish-solutions.com Mon Jun 12 18:14:02 2006 From: philipp_subx at redfish-solutions.com (Philip Prindeville) Date: Mon, 12 Jun 2006 12:14:02 -0600 Subject: Issues bringing up wireless in %post script In-Reply-To: References: Message-ID: <448DAEEA.10507@redfish-solutions.com> Yes, I am. "uname -r" done inside the %post script says: 2.6.15-1.2054_FC5 That's the kernel in the FC5 .iso images. -Philip Shabazian, Chip wrote: > You aren't running that kernel until after your reboot. > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Philip > Prindeville > Sent: Monday, June 12, 2006 9:20 AM > To: Discussion list about Kickstart > Subject: Issues bringing up wireless in %post script > > I have a Dell Latitude L610 that I'm trying to write a prototype > customer installer for FC5 for. It has an Intel 2915ABG NIC. > > When I try to load up the ipw2200 module from the %post script, I see: > > % modprobe -v ipw2200 > insmod > /lib/modules/2.6.15-1.2054_FC5/kernel/net/ieee80211/ie80211_crypt.ko > WARNING: Error inserting ieee80211_crypt (...): Operation not permitted > ... > > same for ieee80211.ko and ipw2200.ko ... even though they are present on > the system. > > Any idea why this is happening? > > I've installed the ipw2200-firmware RPM before trying to bring any of > this up... I burned it onto the installer DVD, and added it to the > %packages section of the ks.cfg file. > > Clues? > > Thanks, > > -Philip > > > > _______________________________________________ > 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 treed at ultraviolet.org Mon Jun 12 18:54:33 2006 From: treed at ultraviolet.org (Tracy R Reed) Date: Mon, 12 Jun 2006 11:54:33 -0700 Subject: %pre and include files Message-ID: <448DB869.90105@ultraviolet.org> I have a pretty nice kickstart setup going but now I need two different LVM/RAID schemes: One for our 2 drive systems which we want mirrored and one for our 4 drive systems which we want configured with a RAID 5 and a spare. So I split out the common kickstart configs into a file called ks-standardoptions.cfg so I have one central place to edit the things in common and have only the machine specific LVM/RAID config and an include in each main kickstart config. Eventually I will try to put some smarts in the %pre to detect how many HD's the machine has and load the right one. For now I choose at boot time. The problem is: The machine only tftp's the main kickstart file. How do I get the other file to be included? Googling reveals that most people seem to put a wget in a %pre and then include the file they wget. I have tried that but it seems either it is not successfully retrieving the config file to be included or it is not being included at all. I do not see any way to debug what happens during %pre as I do not see any output, not even errors. Is there some option I am missing? Some way to single step it? I guess I could try making the shell /bin/sh -x and see if it echos the commands to the screen as they execute. Does it matter where the %pre and %post sections are located in the kickstart config? Most people seem to put them at the end. But my %include is at the beginning. I read somewhere that two passes are made processing includes so that if the file to be included is not present the first time around it eventually executes the %pre and it tries again. Is that correct? An example of my main config can be found here: http://paste.plone.org/4669 Anyone see a problem with this? Thanks! -- Tracy R Reed http://ultraviolet.org A: Because we read from top to bottom, left to right Q: Why should I start my reply below the quoted text From mshuler at rackspace.com Mon Jun 12 19:02:28 2006 From: mshuler at rackspace.com (Michael Shuler) Date: Mon, 12 Jun 2006 14:02:28 -0500 Subject: Issues bringing up wireless in %post script In-Reply-To: <448DAEEA.10507@redfish-solutions.com> References: <448DAEEA.10507@redfish-solutions.com> Message-ID: <448DBA44.5050403@rackspace.com> On 06/12/06 13:14, Philip Prindeville wrote: > Yes, I am. "uname -r" done inside the %post script says: > 2.6.15-1.2054_FC5z > > That's the kernel in the FC5 .iso images. `uname -r` reports the same version for the running installer kernel as the rpm in the repository which you are installing to the disk. OK. > Shabazian, Chip wrote: >> You aren't running that kernel until after your reboot. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > Philip Prindeville wrote: >> Any idea why this is happening? You (still) aren't running the kernel that you installed to the disk, along with the firmware package, regardless of the fact that the installer kernel happens to be the same version. >> Clues? Sure, build your own custom installer image, or set up wireless after reboot, and try not to be argumentative with people trying to help you out. Kind Regards, Michael From Chip.Shabazian at bankofamerica.com Mon Jun 12 19:10:41 2006 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 12 Jun 2006 12:10:41 -0700 Subject: %pre and include files In-Reply-To: <448DB869.90105@ultraviolet.org> Message-ID: Here, just fix your HD determining problem. This is what I do in my %pre: # Determine how many drives we have set $(list-harddrives) let numd=$#/2 d1=$1 d2=$3 # This is the LD&E scheme for two or more drives if [ $numd -ge 2 ] ; then cat << EOF >> /tmp/partinfo 2 + Drive Partitioning info EOF else cat << EOF >> /tmp/partinfo 1 Drive Partitioning info EOF Then in the top command section, use: %include /tmp/partinfo As for debugging in %pre, I put a sleep 999999999 in the pre, then simply work at the console. There are other ways as well, but this works for me. When you want to move past the sleep, just kill the process. Doing this, you can step through your pre or post with sleeps at any point you want. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Tracy R Reed Sent: Monday, June 12, 2006 11:55 AM To: kickstart-list at redhat.com Subject: %pre and include files I have a pretty nice kickstart setup going but now I need two different LVM/RAID schemes: One for our 2 drive systems which we want mirrored and one for our 4 drive systems which we want configured with a RAID 5 and a spare. So I split out the common kickstart configs into a file called ks-standardoptions.cfg so I have one central place to edit the things in common and have only the machine specific LVM/RAID config and an include in each main kickstart config. Eventually I will try to put some smarts in the %pre to detect how many HD's the machine has and load the right one. For now I choose at boot time. The problem is: The machine only tftp's the main kickstart file. How do I get the other file to be included? Googling reveals that most people seem to put a wget in a %pre and then include the file they wget. I have tried that but it seems either it is not successfully retrieving the config file to be included or it is not being included at all. I do not see any way to debug what happens during %pre as I do not see any output, not even errors. Is there some option I am missing? Some way to single step it? I guess I could try making the shell /bin/sh -x and see if it echos the commands to the screen as they execute. Does it matter where the %pre and %post sections are located in the kickstart config? Most people seem to put them at the end. But my %include is at the beginning. I read somewhere that two passes are made processing includes so that if the file to be included is not present the first time around it eventually executes the %pre and it tries again. Is that correct? An example of my main config can be found here: http://paste.plone.org/4669 Anyone see a problem with this? Thanks! -- Tracy R Reed http://ultraviolet.org A: Because we read from top to bottom, left to right Q: Why should I start my reply below the quoted text _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From philipp_subx at redfish-solutions.com Mon Jun 12 20:38:06 2006 From: philipp_subx at redfish-solutions.com (Philip Prindeville) Date: Mon, 12 Jun 2006 14:38:06 -0600 Subject: Issues bringing up wireless in %post script In-Reply-To: <448DBA44.5050403@rackspace.com> References: <448DAEEA.10507@redfish-solutions.com> <448DBA44.5050403@rackspace.com> Message-ID: <448DD0AE.4030709@redfish-solutions.com> Michael Shuler wrote: > On 06/12/06 13:14, Philip Prindeville wrote: >> Yes, I am. "uname -r" done inside the %post script says: >> 2.6.15-1.2054_FC5z >> >> That's the kernel in the FC5 .iso images. > > `uname -r` reports the same version for the running installer kernel > as the rpm in the repository which you are installing to the disk. OK. Alright, so the installer kernel tells me it's 2.6.15-1.2054_FC5. The kernel I'll boot into on reboot is 2.6.15-1.2054_FC5. The modules that I'm trying to load up are those for 2.6.15-1.2054_FC5. Where is the incompatibility? >>> You aren't running that kernel until after your reboot. > Shabazian, Chip wrote: > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > > Philip Prindeville wrote: >>> Any idea why this is happening? > > You (still) aren't running the kernel that you installed to the disk, > along with the firmware package, regardless of the fact that the > installer kernel happens to be the same version. Ok, so... How do they differ? There is apparently module support in the installer kernel. >> Clues? > > Sure, build your own custom installer image, or set up wireless after > reboot, and try not to be argumentative with people trying to help you > out. I wasn't aware that I was being argumentative. I understood him to say that they were different versions of the kernel... But they all seem to have the same versioning label. If they differ, it's not by version number that I can tell. There's some other difference that I'm not quite grokking. -Philip > > Kind Regards, > Michael From cleve.cornelius at kroger.com Tue Jun 13 22:00:27 2006 From: cleve.cornelius at kroger.com (cleve.cornelius at kroger.com) Date: Tue, 13 Jun 2006 18:00:27 -0400 Subject: Cleve Cornelius/MIS/CORP/KrogerCo is out of the office. Message-ID: I will be out of the office starting Mon 06/12/2006 and will not return until Mon 06/19/2006. I will be out of the office until 6/26, on vacation from 6/12 to 6/19...I will be checking mail periodically. From mariogzuk at technikz.de Wed Jun 14 09:12:10 2006 From: mariogzuk at technikz.de (Mario Gzuk) Date: Wed, 14 Jun 2006 11:12:10 +0200 Subject: Exisiting partitions - ntfs In-Reply-To: <1150118782.11564.18.camel@alpha.bgc-jena.mpg.de> References: <1150118782.11564.18.camel@alpha.bgc-jena.mpg.de> Message-ID: <1150276330.18549.1.camel@alpha.bgc-jena.mpg.de> Seems to be there is no way.... So maybe its only possible to add a line into /etc/fstab through the post script. .... greetings Am Montag, den 12.06.2006, 15:26 +0200 schrieb Mario Gzuk: > Hi, > is there a way to specify a existing ntfs partition through the > kickstart-file? Any hints are welcome. > > > greetings mario gzuk > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From philipp_subx at redfish-solutions.com Mon Jun 19 16:26:34 2006 From: philipp_subx at redfish-solutions.com (Philip Prindeville) Date: Mon, 19 Jun 2006 10:26:34 -0600 Subject: Issues bringing up wireless in %post script In-Reply-To: <448DBA44.5050403@rackspace.com> References: <448DAEEA.10507@redfish-solutions.com> <448DBA44.5050403@rackspace.com> Message-ID: <4496D03A.2030104@redfish-solutions.com> Michael Shuler wrote: >> Any idea why this is happening? > > You (still) aren't running the kernel that you installed to the disk, > along with the firmware package, regardless of the fact that the > installer kernel happens to be the same version. > >>> Clues? > > Sure, build your own custom installer image, or set up wireless after > reboot, and try not to be argumentative with people trying to help you > out. > > Kind Regards, > Michael Ok, what specifically is different about the kernel bundled with the installer (this isn't documented much in Kickstart)? And what is prohibiting (inhibiting) it from being able to load these modules? -Philip From Elizabeth.Brosch at thomson.com Tue Jun 20 14:35:15 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 10:35:15 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDB@tshuspaphimbx01.ERF.THOMSON.COM> Has anyone successfully used one DHCP to serve hosts on two subnets? If anyone can share their dhcpd.conf syntax, I would appreciate. I think I have it correct because the DHCP server will issue: dhcpd: DHCPDISCOVER...................................... dhcpd: DHCPOFFER.......................................... dhcpd: DHCPREQUEST...................................... dhcpd: DHCPACK............................................... But then I immediately receive: in.tftpd[4254]: RRQ from 10.225.248.1 filename pxelinux.0 in.tftpd[4254]: tftpd: read: Connection refused I am not sure if it is a dhcpd.conf configuration problem or not. My switch has "port fast" enabled. I can Kickstart servers with no problem but when I introduce a second subnet into my dhcpd, tftp fails. How anyone ever seen this? I had tried this a few months ago but was never able to get around it but I have a need to revisit this now and get it working. Thanks. Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Matt.Fahrner at coat.com Tue Jun 20 15:11:53 2006 From: Matt.Fahrner at coat.com (Matt Fahrner) Date: Tue, 20 Jun 2006 11:11:53 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDB@tshuspaphimbx01.ERF.THOMSON.COM> References: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDB@tshuspaphimbx01.ERF.THOMSON.COM> Message-ID: <44981039.3070402@coat.com> Two subnets sharing the same wire and interface on the DHCP server, or two separate wires (LANs) and two separate interfaces on the DHCP server? We've done both. - Matt Elizabeth.Brosch at thomson.com wrote: > Has anyone successfully used one DHCP to serve hosts on two subnets? > > > > If anyone can share their dhcpd.conf syntax, I would appreciate. I > think I have it correct because the DHCP server will issue: > > > > dhcpd: DHCPDISCOVER????????????.. > > dhcpd: DHCPOFFER?????????????? > > dhcpd: DHCPREQUEST????????????.. > dhcpd: DHCPACK???????????????.. > > > > But then I immediately receive: > > > > in.tftpd[4254]: RRQ from 10.225.248.1 filename pxelinux.0 > > in.tftpd[4254]: tftpd: read: Connection refused > > > > I am not sure if it is a dhcpd.conf configuration problem or not. My > switch has ?port fast? enabled. I can Kickstart servers with no problem > but when I introduce a second subnet into my dhcpd, tftp fails. > > > > How anyone ever seen this? > > > > I had tried this a few months ago but was never able to get around it > but I have a need to revisit this now and get it working. > > > > Thanks. > > > > Liz Brosch > > TSH System Services > > Philadelphia, PA 19104 > > office: (215) 823-6144 > > cell: (267) 784-9166 > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- --------------------------------------------------------------------- Matt Fahrner 2 South Park St. Chief Systems Architect Willis House Burlington Coat Factory Warehouse Lebanon, N.H. 03766 Tel: (603) 448-4100 x5150 USA Fax: (603) 443-6190 Matt.Fahrner at COAT.COM --------------------------------------------------------------------- From Elizabeth.Brosch at thomson.com Tue Jun 20 15:14:03 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 11:14:03 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> Two subnets sharing the same wire and interface on the dhcp server. Well actually my dhcp server has one interface configured with a 10.225.248.x IP address. I want to serve hosts on the 10.225.240.x subnet. The 248 and the 240 subnets are on the same vlan. Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 -----Original Message----- From skvidal at linux.duke.edu Tue Jun 20 15:20:08 2006 From: skvidal at linux.duke.edu (seth vidal) Date: Tue, 20 Jun 2006 11:20:08 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> References: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> Message-ID: <1150816809.9213.2.camel@cutter> On Tue, 2006-06-20 at 11:14 -0400, Elizabeth.Brosch at thomson.com wrote: > Two subnets sharing the same wire and interface on the dhcp server. > Well actually my dhcp server has one interface configured with a > 10.225.248.x IP address. I want to serve hosts on the 10.225.240.x > subnet. > > The 248 and the 240 subnets are on the same vlan. 1. wrong list for this question. 2. look at the shared network config for dhcp. -sv From Elizabeth.Brosch at thomson.com Tue Jun 20 15:20:53 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 11:20:53 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDE@tshuspaphimbx01.ERF.THOMSON.COM> Didn't realize wrong list - I am attempting to Kickstart servers. I have the shared network config set up that is why my dhcp server is acknowledging the client. It is the tftp timeout problem I am attempting to get resolved. From cjk at techma.com Tue Jun 20 16:56:38 2006 From: cjk at techma.com (Kovacs, Corey J.) Date: Tue, 20 Jun 2006 12:56:38 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDB@tshuspaphimbx01.ERF.THOMSON.COM> Message-ID: dhcp doesn't cross subnets by default (it's broadcast). You can set up your vlans with a helper address pointing to your dhcp server, ir you might try setting up a dhcp proxy. I have never set up a dhcp proxy, but I use helper addresses all the time. Seems to work ok. Corey ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Elizabeth.Brosch at thomson.com Sent: Tuesday, June 20, 2006 10:35 AM To: kickstart-list at redhat.com Subject: Using a DHCP server for two subnets Has anyone successfully used one DHCP to serve hosts on two subnets? If anyone can share their dhcpd.conf syntax, I would appreciate. I think I have it correct because the DHCP server will issue: dhcpd: DHCPDISCOVER...................................... dhcpd: DHCPOFFER.......................................... dhcpd: DHCPREQUEST...................................... dhcpd: DHCPACK............................................... But then I immediately receive: in.tftpd[4254]: RRQ from 10.225.248.1 filename pxelinux.0 in.tftpd[4254]: tftpd: read: Connection refused I am not sure if it is a dhcpd.conf configuration problem or not. My switch has "port fast" enabled. I can Kickstart servers with no problem but when I introduce a second subnet into my dhcpd, tftp fails. How anyone ever seen this? I had tried this a few months ago but was never able to get around it but I have a need to revisit this now and get it working. Thanks. Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Matt.Fahrner at coat.com Tue Jun 20 17:06:30 2006 From: Matt.Fahrner at coat.com (Matt Fahrner) Date: Tue, 20 Jun 2006 13:06:30 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> References: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> Message-ID: <44982B16.8000203@coat.com> Unfortunately if they share a single broadcast domain, that is, if one subnet can hear the other subnet's broadcasts (which seems to be what you're implying), then in my experience there is no way to hand out the DHCP addresses cleanly on a per subnet basis. The problem is DHCP requests are layer 2 broadcasts and do not contain any IP network level information. Thus it's impossible to tell which of the two subnets the DHCP broadcast request came from. Therefor you couldn't know which pool to allocate from. That's not to say you can't serve DHCP, you can, but the dynamic pool has to be applied to one of the two networks only (or rather, if not, you'll never be able to reliably predict which network pool a given client is going to be allocated from). If you want this form, look into the "shared-network" syntax. You might get it to partially work doing one of the network with hardcoded MAC to host relationships (or perhaps client identifier, I can't remember all the options for keying). Still, there is no way to cleanly dynamically allocate on both networks with some regularity. If you are using VLANs, unless you have some hosts mixed on a hub, why not just split them completely so they don't share broadcast domains? Extra NIC cards are cheap or you could look into using VLANs straight into your server, which Linux supports. - Matt Elizabeth.Brosch at thomson.com wrote: > Two subnets sharing the same wire and interface on the dhcp server. > Well actually my dhcp server has one interface configured with a > 10.225.248.x IP address. I want to serve hosts on the 10.225.240.x > subnet. > > The 248 and the 240 subnets are on the same vlan. > > Liz Brosch > TSH System Services > Philadelphia, PA 19104 > office: (215) 823-6144 > cell: (267) 784-9166 > -----Original Message----- > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- --------------------------------------------------------------------- Matt Fahrner 2 South Park St. Chief Systems Architect Willis House Burlington Coat Factory Warehouse Lebanon, N.H. 03766 Tel: (603) 448-4100 x5150 USA Fax: (603) 443-6190 Matt.Fahrner at COAT.COM --------------------------------------------------------------------- From Elizabeth.Brosch at thomson.com Tue Jun 20 17:48:51 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 13:48:51 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFE0@tshuspaphimbx01.ERF.THOMSON.COM> Thank you Matt for such a detailed response to my problem. For my dhcp configuration, here is an outline of what I have: dhcp/Kickstart/pxe/tftp server: 10.225.248.x dhcpd.conf configuration: shared-network configuration. This lists network information for the 10.225.248.x subnet and the 10.225.240.x subnet. I am attempting to Kickstart servers on the 10.225.240.x subnet. I have it hard coded with MAC to host relationship. Each host I want to install is listed as Node1, Node2, etc. The dhcp seems to work up to the point of requesting tftp. This is coming back with "in.tftpd[22003]: tftpd: read: Connection refused" Does this make sense? Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 From teng at dataway.com Tue Jun 20 19:24:17 2006 From: teng at dataway.com (Tedman Eng) Date: Tue, 20 Jun 2006 12:24:17 -0700 Subject: Using a DHCP server for two subnets Message-ID: <37ED92F9890FAF4BB947613C66FF8B1A08BB2DF1@dw-mail.dataway.com> I haven't seen mention of /etc/hosts.allow, but there's another place to look if you haven't already. > -----Original Message----- > From: Elizabeth.Brosch at thomson.com > [mailto:Elizabeth.Brosch at thomson.com] > Sent: Tuesday, June 20, 2006 10:49 AM > To: kickstart-list at redhat.com > Subject: RE: Using a DHCP server for two subnets > > > Thank you Matt for such a detailed response to my problem. > > For my dhcp configuration, here is an outline of what I have: > > dhcp/Kickstart/pxe/tftp server: 10.225.248.x > dhcpd.conf configuration: shared-network configuration. This > lists network information for the 10.225.248.x subnet and the > 10.225.240.x subnet. I am attempting to Kickstart servers on the > 10.225.240.x subnet. I have it hard coded with MAC to host > relationship. Each host I want to install is listed as Node1, Node2, > etc. > > The dhcp seems to work up to the point of requesting tftp. This is > coming back with "in.tftpd[22003]: tftpd: read: Connection refused" > > Does this make sense? > > Liz Brosch > TSH System Services > Philadelphia, PA 19104 > office: (215) 823-6144 > cell: (267) 784-9166 > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Elizabeth.Brosch at thomson.com Tue Jun 20 19:26:09 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 15:26:09 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFEC@tshuspaphimbx01.ERF.THOMSON.COM> I thought maybe I needed to put an entry into /etc/hosts.allow. Should this be an entry for in.tftpd or tftpd? Would you happen to know? Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 From teng at dataway.com Tue Jun 20 19:32:25 2006 From: teng at dataway.com (Tedman Eng) Date: Tue, 20 Jun 2006 12:32:25 -0700 Subject: Using a DHCP server for two subnets Message-ID: <37ED92F9890FAF4BB947613C66FF8B1A08BB2DF2@dw-mail.dataway.com> In the case that /etc/hosts.allow states: in.tftpd: LOCAL You would need to IP the DHCP server with an alias interface on your other network (so that it's also considered "local"), or change the entry to read: in.tftpd: ALL Yes, ALL is less secure....but if we're concerned about security, why are we using tftp? :) > -----Original Message----- > From: Elizabeth.Brosch at thomson.com > [mailto:Elizabeth.Brosch at thomson.com] > Sent: Tuesday, June 20, 2006 12:26 PM > To: kickstart-list at redhat.com > Subject: RE: Using a DHCP server for two subnets > > > I thought maybe I needed to put an entry into > /etc/hosts.allow. Should > this be an entry for in.tftpd or tftpd? Would you happen to know? > > Liz Brosch > TSH System Services > Philadelphia, PA 19104 > office: (215) 823-6144 > cell: (267) 784-9166 > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Elizabeth.Brosch at thomson.com Tue Jun 20 19:34:03 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Tue, 20 Jun 2006 15:34:03 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8BFED@tshuspaphimbx01.ERF.THOMSON.COM> I have configured an alias on my dhcp server for the other network, so I will use the in.tftpd: LOCAL entry. Thank you for taking the time to write. Liz Brosch TSH System Services Philadelphia, PA 19104 office: (215) 823-6144 cell: (267) 784-9166 -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Tedman Eng Sent: Tuesday, June 20, 2006 3:32 PM To: 'Discussion list about Kickstart' Subject: RE: Using a DHCP server for two subnets In the case that /etc/hosts.allow states: in.tftpd: LOCAL You would need to IP the DHCP server with an alias interface on your other network (so that it's also considered "local"), or change the entry to read: in.tftpd: ALL Yes, ALL is less secure....but if we're concerned about security, why are we using tftp? :) From wcervini at gmail.com Tue Jun 20 19:45:34 2006 From: wcervini at gmail.com (Walter Cervini) Date: Tue, 20 Jun 2006 15:45:34 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: References: Message-ID: <1150832735.3983.22.camel@laptop.volp.local.com> El mar, 20-06-2006 a las 12:56 -0400, Kovacs, Corey J. escribi?: thats right you need configure vlan with sub-interfaces for each subnet > dhcp doesn't cross subnets by default (it's broadcast). You can set up > your vlans with a > helper address pointing to your dhcp server, ir you might try setting > up a dhcp proxy. I > have never set up a dhcp proxy, but I use helper addresses all the > time. Seems to work > ok. > > > Corey > > > > > ______________________________________________________________________ > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of > Elizabeth.Brosch at thomson.com > Sent: Tuesday, June 20, 2006 10:35 AM > To: kickstart-list at redhat.com > Subject: Using a DHCP server for two subnets > > > > Has anyone successfully used one DHCP to serve hosts on two subnets? > > > > If anyone can share their dhcpd.conf syntax, I would appreciate. I > think I have it correct because the DHCP server will issue: > > > > dhcpd: DHCPDISCOVER????????????.. > > dhcpd: DHCPOFFER?????????????? > > dhcpd: DHCPREQUEST????????????.. > dhcpd: DHCPACK???????????????.. > > > > But then I immediately receive: > > > > in.tftpd[4254]: RRQ from 10.225.248.1 filename pxelinux.0 > > in.tftpd[4254]: tftpd: read: Connection refused > > > > I am not sure if it is a dhcpd.conf configuration problem or not. My > switch has ?port fast? enabled. I can Kickstart servers with no > problem but when I introduce a second subnet into my dhcpd, tftp > fails. > > > > How anyone ever seen this? > > > > I had tried this a few months ago but was never able to get around it > but I have a need to revisit this now and get it working. > > > > Thanks. > > > > Liz Brosch > > TSH System Services > > Philadelphia, PA 19104 > > office: (215) 823-6144 > > cell: (267) 784-9166 > > > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list Walter Cervini Soporte Fedora en irc.freenode.net Canal fedora-ve http://www.volp.org.ve -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 191 bytes Desc: This is a digitally signed message part URL: From wcervini at gmail.com Tue Jun 20 19:48:14 2006 From: wcervini at gmail.com (Walter Cervini) Date: Tue, 20 Jun 2006 15:48:14 -0400 Subject: Using a DHCP server for two subnets In-Reply-To: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> References: <2213CAA79DA9094AB52517FDADEF07FA03D8BFDD@tshuspaphimbx01.ERF.THOMSON.COM> Message-ID: <1150832894.3983.26.camel@laptop.volp.local.com> You cas user vconfig for implement vlan. Usage: add [interface-name] [vlan_id] rem [vlan-name] set_flag [interface-name] [flag-num] [0 | 1] set_egress_map [vlan-name] [skb_priority] [vlan_qos] set_ingress_map [vlan-name] [skb_priority] [vlan_qos] set_name_type [name-type] * The [interface-name] is the name of the ethernet card that hosts the VLAN you are talking about. * The vlan_id is the identifier (0-4095) of the VLAN you are operating on. * skb_priority is the priority in the socket buffer (sk_buff). * vlan_qos is the 3 bit priority in the VLAN header * name-type: VLAN_PLUS_VID (vlan0005), VLAN_PLUS_VID_NO_PAD (vlan5), DEV_PLUS_VID (eth0.0005), DEV_PLUS_VID_NO_PAD (eth0.5) * bind-type: PER_DEVICE # Allows vlan 5 on eth0 and eth1 to be unique. PER_KERNEL # Forces vlan 5 to be unique across all devices. * FLAGS: 1 REORDER_HDR When this is set, the VLAN device will move the ethernet header around to make it look exactly like a real ethernet device. This may help programs such as DHCPd which read the raw ethernet packet and make assumptions about the location of bytes. If you don't need it, don't turn it on, because there will be at least a small performance degradation. Default is OFF. El mar, 20-06-2006 a las 11:14 -0400, Elizabeth.Brosch at thomson.com escribi?: > Two subnets sharing the same wire and interface on the dhcp server. > Well actually my dhcp server has one interface configured with a > 10.225.248.x IP address. I want to serve hosts on the 10.225.240.x > subnet. > > The 248 and the 240 subnets are on the same vlan. > > Liz Brosch > TSH System Services > Philadelphia, PA 19104 > office: (215) 823-6144 > cell: (267) 784-9166 > -----Original Message----- > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list Walter Cervini Soporte Fedora en irc.freenode.net Canal fedora-ve http://www.volp.org.ve -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 191 bytes Desc: This is a digitally signed message part URL: From cleve.cornelius at kroger.com Tue Jun 20 20:06:11 2006 From: cleve.cornelius at kroger.com (cleve.cornelius at kroger.com) Date: Tue, 20 Jun 2006 16:06:11 -0400 Subject: Cleve Cornelius/MIS/CORP/KrogerCo is out of the office. Message-ID: I will be out of the office starting Mon 06/19/2006 and will not return until Mon 06/26/2006. I will be out of the office at the Smiths Food & Drug location the week of 6/19. I will be answering email in a delayed fashion due to project requirements at Smith. I will check email periodically through out the day. Thanks. From Elizabeth.Brosch at thomson.com Fri Jun 23 15:53:56 2006 From: Elizabeth.Brosch at thomson.com (Elizabeth.Brosch at thomson.com) Date: Fri, 23 Jun 2006 11:53:56 -0400 Subject: Using a DHCP server for two subnets Message-ID: <2213CAA79DA9094AB52517FDADEF07FA03D8C029@tshuspaphimbx01.ERF.THOMSON.COM> Thanks for the info. I have tried putting in.tftpd in my hosts.allow file but it fails with "invalid option". -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Tedman Eng Sent: Tuesday, June 20, 2006 3:32 PM To: 'Discussion list about Kickstart' Subject: RE: Using a DHCP server for two subnets In the case that /etc/hosts.allow states: in.tftpd: LOCAL You would need to IP the DHCP server with an alias interface on your other network (so that it's also considered "local"), or change the entry to read: in.tftpd: ALL Yes, ALL is less secure....but if we're concerned about security, why are we using tftp? :) From dleske at uvic.ca Mon Jun 26 23:30:51 2006 From: dleske at uvic.ca (Drew Leske) Date: Mon, 26 Jun 2006 16:30:51 -0700 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Message-ID: <44A06E2B.107@uvic.ca> Hi all, We're using KickStart for clusters and Linux deployment in general, and for the most part we are happy with this. Recently however we have acquired some 64-bit machines to be added into our cluster and there are some issues here. We would like to make these resources available for 64-bit computing, and have therefore installed primarily 64-bit libraries on them. However, most of our users run 32-bit applications. At this point there is not enough need for exclusively 64-bit machines, but I would like to make the platform available. According to a colleague there is a way with yum to specify that whenever installing a package, a 32-bit version of the package, if available, will be installed. I am not familiar with yum and am limited to RHEL3, up2date, and rpm, but this is the functionality I am looking for. I have looked at the documentation for KickStart (RHEL3 and RHEL4) as well as browser this mailing list for information, and of course I've searched the web--nada. Does anybody have any suggestions for how to handle this? Thanks, Drew. -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) From Colin.Coe at woodside.com.au Mon Jun 26 23:34:54 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Tue, 27 Jun 2006 07:34:54 +0800 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Message-ID: <17EBC05307239C4896ED25244F42A80101367A14@permls05.wde.woodside.com.au> Under RHEL3&4, this happens automatically. For example, querying the RPM database for info on the 'glibc' package gives: rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 The 'up2date' utility will take care of this for you. HTH CC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske Sent: Tuesday, 27 June 2006 7:31 AM To: kickstart-list at redhat.com Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Hi all, We're using KickStart for clusters and Linux deployment in general, and for the most part we are happy with this. Recently however we have acquired some 64-bit machines to be added into our cluster and there are some issues here. We would like to make these resources available for 64-bit computing, and have therefore installed primarily 64-bit libraries on them. However, most of our users run 32-bit applications. At this point there is not enough need for exclusively 64-bit machines, but I would like to make the platform available. According to a colleague there is a way with yum to specify that whenever installing a package, a 32-bit version of the package, if available, will be installed. I am not familiar with yum and am limited to RHEL3, up2date, and rpm, but this is the functionality I am looking for. I have looked at the documentation for KickStart (RHEL3 and RHEL4) as well as browser this mailing list for information, and of course I've searched the web--nada. Does anybody have any suggestions for how to handle this? Thanks, Drew. -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From dleske at uvic.ca Tue Jun 27 00:01:30 2006 From: dleske at uvic.ca (Drew Leske) Date: Mon, 26 Jun 2006 17:01:30 -0700 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: <17EBC05307239C4896ED25244F42A80101367A14@permls05.wde.woodside.com.au> References: <17EBC05307239C4896ED25244F42A80101367A14@permls05.wde.woodside.com.au> Message-ID: <44A0755A.4050603@uvic.ca> Thanks Colin. I had noticed that some libraries are installed for both, but some definitely are not. One specific example is libGLU, provided by the XFree86-Mesa-libGLU package. [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 Believe it or not our users require X libraries to run their numerical analysis (on non-interactive machines). I wonder if only the 64-bit libGLU is provided because nobody saw a need to have both 32-bit and 64-bit versions of X available. Can you suggest a way around this? At this point it looks like I'll have to take care of this manually or script up my own solution--if the two flavours can even co-exist for some of these packages. Thanks, Drew. Coe, Colin C. wrote: > Under RHEL3&4, this happens automatically. > > For example, querying the RPM database for info on the 'glibc' package > gives: > rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" > glibc 2.3.4 2.19 i686 > glibc 2.3.4 2.19 x86_64 > > The 'up2date' utility will take care of this for you. > > HTH > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 7:31 AM > To: kickstart-list at redhat.com > Subject: Automatically installing 32-bit libraries on 64-bit machine? > (RHEL3) > > > Hi all, > > We're using KickStart for clusters and Linux deployment in general, and > for > the most part we are happy with this. Recently however we have acquired > > some 64-bit machines to be added into our cluster and there are some > issues > here. > > We would like to make these resources available for 64-bit computing, > and > have therefore installed primarily 64-bit libraries on them. However, > most > of our users run 32-bit applications. At this point there is not enough > > need for exclusively 64-bit machines, but I would like to make the > platform > available. > > According to a colleague there is a way with yum to specify that > whenever > installing a package, a 32-bit version of the package, if available, > will be > installed. I am not familiar with yum and am limited to RHEL3, up2date, > and > rpm, but this is the functionality I am looking for. > > I have looked at the documentation for KickStart (RHEL3 and RHEL4) as > well > as browser this mailing list for information, and of course I've > searched > the web--nada. Does anybody have any suggestions for how to handle > this? > > Thanks, > Drew. > -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) From Colin.Coe at woodside.com.au Tue Jun 27 00:16:43 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Tue, 27 Jun 2006 08:16:43 +0800 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Message-ID: <17EBC05307239C4896ED25244F42A80101367A15@permls05.wde.woodside.com.au> Interesting... cat /etc/redhat-release Red Hat Enterprise Linux WS release 3 (Taroon Update 5) rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep -i mesa XFree86-Mesa-libGL 4.3.0 81.EL i386 XFree86-Mesa-libGLU 4.3.0 81.EL x86_64 XFree86-Mesa-libGL 4.3.0 81.EL x86_64 XFree86-Mesa-libGLU 4.3.0 81.EL i386 cat /etc/redhat-release Red Hat Enterprise Linux ES release 4 (Nahant Update 3) rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep -i mesa xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 x86_64 xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 x86_64 xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 i386 xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 i386 You should have both 32 and 64bit packages installed by default. Did perchance someone remove the 'duplicate' packages from the 'broken' installation? I've seen this done.... Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are in the 'Compatibility Arch Support' install group and, I'm pretty sure that all appropriate architectures are installed by default. CC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske Sent: Tuesday, 27 June 2006 8:02 AM To: Discussion list about Kickstart Subject: Re: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Thanks Colin. I had noticed that some libraries are installed for both, but some definitely are not. One specific example is libGLU, provided by the XFree86-Mesa-libGLU package. [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 Believe it or not our users require X libraries to run their numerical analysis (on non-interactive machines). I wonder if only the 64-bit libGLU is provided because nobody saw a need to have both 32-bit and 64-bit versions of X available. Can you suggest a way around this? At this point it looks like I'll have to take care of this manually or script up my own solution--if the two flavours can even co-exist for some of these packages. Thanks, Drew. Coe, Colin C. wrote: > Under RHEL3&4, this happens automatically. > > For example, querying the RPM database for info on the 'glibc' package > gives: > rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" > glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 > > The 'up2date' utility will take care of this for you. > > HTH > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 7:31 AM > To: kickstart-list at redhat.com > Subject: Automatically installing 32-bit libraries on 64-bit machine? > (RHEL3) > > > Hi all, > > We're using KickStart for clusters and Linux deployment in general, > and for the most part we are happy with this. Recently however we > have acquired > > some 64-bit machines to be added into our cluster and there are some > issues here. > > We would like to make these resources available for 64-bit computing, > and have therefore installed primarily 64-bit libraries on them. > However, most > of our users run 32-bit applications. At this point there is not enough > > need for exclusively 64-bit machines, but I would like to make the > platform available. > > According to a colleague there is a way with yum to specify that > whenever installing a package, a 32-bit version of the package, if > available, will be > installed. I am not familiar with yum and am limited to RHEL3, up2date, > and > rpm, but this is the functionality I am looking for. > > I have looked at the documentation for KickStart (RHEL3 and RHEL4) as > well as browser this mailing list for information, and of course I've > searched > the web--nada. Does anybody have any suggestions for how to handle > this? > > Thanks, > Drew. > -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From dleske at uvic.ca Tue Jun 27 00:30:50 2006 From: dleske at uvic.ca (Drew Leske) Date: Mon, 26 Jun 2006 17:30:50 -0700 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: <17EBC05307239C4896ED25244F42A80101367A15@permls05.wde.woodside.com.au> References: <17EBC05307239C4896ED25244F42A80101367A15@permls05.wde.woodside.com.au> Message-ID: <44A07C3A.3010903@uvic.ca> Oddness. I've got Red Hat Enterprise Linux WS release 3 (Taroon Update 7). Are you querying on hosts installed via KickStart? I'm the only one currently using these hosts currently, so interference is not the problem here. If there were some way to turn this OFF in kickstart, that might be the culprit, but the KickStart scriptage all seems pretty plain. The KickStart scripts are managed by xCAT (cluster application toolkit) so it's possible it's doing something nasty. I'll look into that. Thanks for your help. Curiouser and curiouser... Drew. Coe, Colin C. wrote: > Interesting... > > cat /etc/redhat-release > Red Hat Enterprise Linux WS release 3 (Taroon Update 5) > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > -i mesa > XFree86-Mesa-libGL 4.3.0 81.EL i386 > XFree86-Mesa-libGLU 4.3.0 81.EL x86_64 > XFree86-Mesa-libGL 4.3.0 81.EL x86_64 > XFree86-Mesa-libGLU 4.3.0 81.EL i386 > > cat /etc/redhat-release > Red Hat Enterprise Linux ES release 4 (Nahant Update 3) > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > -i mesa > xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 x86_64 > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 x86_64 > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 i386 > xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 i386 > > > You should have both 32 and 64bit packages installed by default. Did > perchance someone remove the 'duplicate' packages from the 'broken' > installation? I've seen this done.... > > Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are in > the 'Compatibility Arch Support' install group and, I'm pretty sure that > all appropriate architectures are installed by default. > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 8:02 AM > To: Discussion list about Kickstart > Subject: Re: Automatically installing 32-bit libraries on 64-bit > machine? (RHEL3) > > > Thanks Colin. I had noticed that some libraries are installed for both, > but > some definitely are not. One specific example is libGLU, provided by > the > XFree86-Mesa-libGLU package. > > [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} > %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm > -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} > %{RELEASE} %{ARCH}\n" > XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 > > Believe it or not our users require X libraries to run their numerical > analysis (on non-interactive machines). I wonder if only the 64-bit > libGLU > is provided because nobody saw a need to have both 32-bit and 64-bit > versions of X available. > > Can you suggest a way around this? At this point it looks like I'll > have to > take care of this manually or script up my own solution--if the two > flavours > can even co-exist for some of these packages. > > Thanks, > Drew. > > Coe, Colin C. wrote: >> Under RHEL3&4, this happens automatically. >> >> For example, querying the RPM database for info on the 'glibc' package >> gives: >> rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" >> glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 >> >> The 'up2date' utility will take care of this for you. >> >> HTH >> >> CC >> >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske >> Sent: Tuesday, 27 June 2006 7:31 AM >> To: kickstart-list at redhat.com >> Subject: Automatically installing 32-bit libraries on 64-bit machine? >> (RHEL3) >> >> >> Hi all, >> >> We're using KickStart for clusters and Linux deployment in general, >> and for the most part we are happy with this. Recently however we >> have acquired >> >> some 64-bit machines to be added into our cluster and there are some >> issues here. >> >> We would like to make these resources available for 64-bit computing, >> and have therefore installed primarily 64-bit libraries on them. >> However, most >> of our users run 32-bit applications. At this point there is not > enough >> need for exclusively 64-bit machines, but I would like to make the >> platform available. >> >> According to a colleague there is a way with yum to specify that >> whenever installing a package, a 32-bit version of the package, if >> available, will be >> installed. I am not familiar with yum and am limited to RHEL3, > up2date, >> and >> rpm, but this is the functionality I am looking for. >> >> I have looked at the documentation for KickStart (RHEL3 and RHEL4) as >> well as browser this mailing list for information, and of course I've >> searched >> the web--nada. Does anybody have any suggestions for how to handle >> this? >> >> Thanks, >> Drew. >> > -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) From Colin.Coe at woodside.com.au Tue Jun 27 00:32:20 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Tue, 27 Jun 2006 08:32:20 +0800 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Message-ID: <17EBC05307239C4896ED25244F42A80101367A16@permls05.wde.woodside.com.au> All our hosts are installed via KickStart. AFAIK KickStart is not granular enuogh to be specifying libraries down to the arch level. Good luck and I'd be interested to hear the culprit & resolution if/when found. CC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske Sent: Tuesday, 27 June 2006 8:31 AM To: Discussion list about Kickstart Subject: Re: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) Oddness. I've got Red Hat Enterprise Linux WS release 3 (Taroon Update 7). Are you querying on hosts installed via KickStart? I'm the only one currently using these hosts currently, so interference is not the problem here. If there were some way to turn this OFF in kickstart, that might be the culprit, but the KickStart scriptage all seems pretty plain. The KickStart scripts are managed by xCAT (cluster application toolkit) so it's possible it's doing something nasty. I'll look into that. Thanks for your help. Curiouser and curiouser... Drew. Coe, Colin C. wrote: > Interesting... > > cat /etc/redhat-release > Red Hat Enterprise Linux WS release 3 (Taroon Update 5) > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > -i mesa XFree86-Mesa-libGL 4.3.0 81.EL i386 > XFree86-Mesa-libGLU 4.3.0 81.EL x86_64 > XFree86-Mesa-libGL 4.3.0 81.EL x86_64 > XFree86-Mesa-libGLU 4.3.0 81.EL i386 > > cat /etc/redhat-release > Red Hat Enterprise Linux ES release 4 (Nahant Update 3) > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > -i mesa xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 x86_64 > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 x86_64 > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 i386 > xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 i386 > > > You should have both 32 and 64bit packages installed by default. Did > perchance someone remove the 'duplicate' packages from the 'broken' > installation? I've seen this done.... > > Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are > in the 'Compatibility Arch Support' install group and, I'm pretty sure > that all appropriate architectures are installed by default. > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 8:02 AM > To: Discussion list about Kickstart > Subject: Re: Automatically installing 32-bit libraries on 64-bit > machine? (RHEL3) > > > Thanks Colin. I had noticed that some libraries are installed for > both, but some definitely are not. One specific example is libGLU, > provided by the > XFree86-Mesa-libGLU package. > > [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} > %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm > -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} %{RELEASE} > %{ARCH}\n" XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 > > Believe it or not our users require X libraries to run their numerical > analysis (on non-interactive machines). I wonder if only the 64-bit > libGLU > is provided because nobody saw a need to have both 32-bit and 64-bit > versions of X available. > > Can you suggest a way around this? At this point it looks like I'll > have to take care of this manually or script up my own solution--if > the two flavours > can even co-exist for some of these packages. > > Thanks, > Drew. > > Coe, Colin C. wrote: >> Under RHEL3&4, this happens automatically. >> >> For example, querying the RPM database for info on the 'glibc' >> package >> gives: >> rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" >> glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 >> >> The 'up2date' utility will take care of this for you. >> >> HTH >> >> CC >> >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske >> Sent: Tuesday, 27 June 2006 7:31 AM >> To: kickstart-list at redhat.com >> Subject: Automatically installing 32-bit libraries on 64-bit machine? >> (RHEL3) >> >> >> Hi all, >> >> We're using KickStart for clusters and Linux deployment in general, >> and for the most part we are happy with this. Recently however we >> have acquired >> >> some 64-bit machines to be added into our cluster and there are some >> issues here. >> >> We would like to make these resources available for 64-bit computing, >> and have therefore installed primarily 64-bit libraries on them. >> However, most >> of our users run 32-bit applications. At this point there is not > enough >> need for exclusively 64-bit machines, but I would like to make the >> platform available. >> >> According to a colleague there is a way with yum to specify that >> whenever installing a package, a 32-bit version of the package, if >> available, will be >> installed. I am not familiar with yum and am limited to RHEL3, > up2date, >> and >> rpm, but this is the functionality I am looking for. >> >> I have looked at the documentation for KickStart (RHEL3 and RHEL4) as >> well as browser this mailing list for information, and of course I've >> searched >> the web--nada. Does anybody have any suggestions for how to handle >> this? >> >> Thanks, >> Drew. >> > -- Drew Leske :: Systems Group/Unix, Computing Services, University of Victoria dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From devnull13 at comcast.net Tue Jun 27 01:47:15 2006 From: devnull13 at comcast.net (John Rothschild) Date: Mon, 26 Jun 2006 21:47:15 -0400 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: <44A06E2B.107@uvic.ca> References: <44A06E2B.107@uvic.ca> Message-ID: On RHEL 4, the following works. I haven't tested it on RHEL 3. Add this to your packages section of your kickstart file. %packages --resolvedeps @Compatibility Arch Support @Compatibility Arch Development Support (I got this information from a March 1 posting to this list.) You can also do this via up2date. You will need to use the undocumented --arch option to get the right version. (--arch=i386 (?) vice --arch=x86_64) I hope this helps. JR On Jun 26, 2006, at 7:30 PM, Drew Leske wrote: > Hi all, > > We're using KickStart for clusters and Linux deployment in general, > and for the most part we are happy with this. Recently however we > have acquired some 64-bit machines to be added into our cluster and > there are some issues here. > > We would like to make these resources available for 64-bit > computing, and have therefore installed primarily 64-bit libraries > on them. However, most of our users run 32-bit applications. At > this point there is not enough need for exclusively 64-bit > machines, but I would like to make the platform available. > > According to a colleague there is a way with yum to specify that > whenever installing a package, a 32-bit version of the package, if > available, will be installed. I am not familiar with yum and am > limited to RHEL3, up2date, and rpm, but this is the functionality I > am looking for. > > I have looked at the documentation for KickStart (RHEL3 and RHEL4) > as well as browser this mailing list for information, and of course > I've searched the web--nada. Does anybody have any suggestions for > how to handle this? > > Thanks, > Drew. > > -- > Drew Leske :: Systems Group/Unix, Computing Services, University of > Victoria > dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From dleske at uvic.ca Tue Jun 27 06:57:26 2006 From: dleske at uvic.ca (Drew Leske) Date: Mon, 26 Jun 2006 23:57:26 -0700 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: <17EBC05307239C4896ED25244F42A80101367A16@permls05.wde.woodside.com.au> References: <17EBC05307239C4896ED25244F42A80101367A16@permls05.wde.woodside.com.au> Message-ID: <44A0D6D6.1020909@uvic.ca> Hi Colin, The solution was this: >> Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are >> in the 'Compatibility Arch Support' install group and, I'm pretty sure >> that all appropriate architectures are installed by default. I misread this the first time around. Adding in this install group fixed the problem. Thanks for your help. Drew. Coe, Colin C. wrote: > All our hosts are installed via KickStart. AFAIK KickStart is not > granular enuogh to be specifying libraries down to the arch level. > > Good luck and I'd be interested to hear the culprit & resolution if/when > found. > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 8:31 AM > To: Discussion list about Kickstart > Subject: Re: Automatically installing 32-bit libraries on 64-bit > machine? (RHEL3) > > > Oddness. > > I've got Red Hat Enterprise Linux WS release 3 (Taroon Update 7). Are > you > querying on hosts installed via KickStart? > > I'm the only one currently using these hosts currently, so interference > is > not the problem here. > > If there were some way to turn this OFF in kickstart, that might be the > culprit, but the KickStart scriptage all seems pretty plain. > > The KickStart scripts are managed by xCAT (cluster application toolkit) > so > it's possible it's doing something nasty. I'll look into that. > > Thanks for your help. Curiouser and curiouser... > Drew. > > > > Coe, Colin C. wrote: >> Interesting... >> >> cat /etc/redhat-release >> Red Hat Enterprise Linux WS release 3 (Taroon Update 5) >> >> rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > >> -i mesa XFree86-Mesa-libGL 4.3.0 81.EL i386 >> XFree86-Mesa-libGLU 4.3.0 81.EL x86_64 >> XFree86-Mesa-libGL 4.3.0 81.EL x86_64 >> XFree86-Mesa-libGLU 4.3.0 81.EL i386 >> >> cat /etc/redhat-release >> Red Hat Enterprise Linux ES release 4 (Nahant Update 3) >> >> rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > >> -i mesa xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 x86_64 >> xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 x86_64 >> xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 i386 >> xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 i386 >> >> >> You should have both 32 and 64bit packages installed by default. Did >> perchance someone remove the 'duplicate' packages from the 'broken' >> installation? I've seen this done.... >> >> Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are >> in the 'Compatibility Arch Support' install group and, I'm pretty sure > >> that all appropriate architectures are installed by default. >> >> CC >> >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske >> Sent: Tuesday, 27 June 2006 8:02 AM >> To: Discussion list about Kickstart >> Subject: Re: Automatically installing 32-bit libraries on 64-bit >> machine? (RHEL3) >> >> >> Thanks Colin. I had noticed that some libraries are installed for >> both, but some definitely are not. One specific example is libGLU, >> provided by the >> XFree86-Mesa-libGLU package. >> >> [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} >> %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm > >> -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} %{RELEASE} >> %{ARCH}\n" XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 >> >> Believe it or not our users require X libraries to run their numerical >> analysis (on non-interactive machines). I wonder if only the 64-bit >> libGLU >> is provided because nobody saw a need to have both 32-bit and 64-bit >> versions of X available. >> >> Can you suggest a way around this? At this point it looks like I'll >> have to take care of this manually or script up my own solution--if >> the two flavours >> can even co-exist for some of these packages. >> >> Thanks, >> Drew. >> >> Coe, Colin C. wrote: >>> Under RHEL3&4, this happens automatically. >>> >>> For example, querying the RPM database for info on the 'glibc' >>> package >>> gives: >>> rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" >>> glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 >>> >>> The 'up2date' utility will take care of this for you. >>> >>> HTH >>> >>> CC >>> >>> -----Original Message----- >>> From: kickstart-list-bounces at redhat.com >>> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske >>> Sent: Tuesday, 27 June 2006 7:31 AM >>> To: kickstart-list at redhat.com >>> Subject: Automatically installing 32-bit libraries on 64-bit machine? >>> (RHEL3) >>> >>> >>> Hi all, >>> >>> We're using KickStart for clusters and Linux deployment in general, >>> and for the most part we are happy with this. Recently however we >>> have acquired >>> >>> some 64-bit machines to be added into our cluster and there are some >>> issues here. >>> >>> We would like to make these resources available for 64-bit computing, >>> and have therefore installed primarily 64-bit libraries on them. >>> However, most >>> of our users run 32-bit applications. At this point there is not >> enough >>> need for exclusively 64-bit machines, but I would like to make the >>> platform available. >>> >>> According to a colleague there is a way with yum to specify that >>> whenever installing a package, a 32-bit version of the package, if >>> available, will be >>> installed. I am not familiar with yum and am limited to RHEL3, >> up2date, >>> and >>> rpm, but this is the functionality I am looking for. >>> >>> I have looked at the documentation for KickStart (RHEL3 and RHEL4) as >>> well as browser this mailing list for information, and of course I've >>> searched >>> the web--nada. Does anybody have any suggestions for how to handle >>> this? >>> >>> Thanks, >>> Drew. >>> > From dleske at uvic.ca Tue Jun 27 06:59:01 2006 From: dleske at uvic.ca (Drew Leske) Date: Mon, 26 Jun 2006 23:59:01 -0700 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: References: <44A06E2B.107@uvic.ca> Message-ID: <44A0D735.6090404@uvic.ca> Hi John, Thanks for the advice. I've already got the --resolvedeps flag but the two Compatibility install groups weren't in there. Now on a clean install I get both flavours and am ready to go back to testing. Thanks all! Drew. John Rothschild wrote: > On RHEL 4, the following works. I haven't tested it on RHEL 3. > > Add this to your packages section of your kickstart file. > > %packages --resolvedeps > > @Compatibility Arch Support > @Compatibility Arch Development Support > > (I got this information from a March 1 posting to this list.) > > You can also do this via up2date. You will need to use the undocumented > --arch option to get the right version. (--arch=i386 (?) vice > --arch=x86_64) I hope this helps. > > JR > > On Jun 26, 2006, at 7:30 PM, Drew Leske wrote: > >> Hi all, >> >> We're using KickStart for clusters and Linux deployment in general, >> and for the most part we are happy with this. Recently however we >> have acquired some 64-bit machines to be added into our cluster and >> there are some issues here. >> >> We would like to make these resources available for 64-bit computing, >> and have therefore installed primarily 64-bit libraries on them. >> However, most of our users run 32-bit applications. At this point >> there is not enough need for exclusively 64-bit machines, but I would >> like to make the platform available. >> >> According to a colleague there is a way with yum to specify that >> whenever installing a package, a 32-bit version of the package, if >> available, will be installed. I am not familiar with yum and am >> limited to RHEL3, up2date, and rpm, but this is the functionality I am >> looking for. >> >> I have looked at the documentation for KickStart (RHEL3 and RHEL4) as >> well as browser this mailing list for information, and of course I've >> searched the web--nada. Does anybody have any suggestions for how to >> handle this? >> >> Thanks, >> Drew. >> >> --Drew Leske :: Systems Group/Unix, Computing Services, University of >> Victoria >> dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) >> >> _______________________________________________ >> 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 jjensen at cisco.com Tue Jun 27 15:38:49 2006 From: jjensen at cisco.com (Joshua Jensen) Date: Tue, 27 Jun 2006 11:38:49 -0400 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: References: <44A06E2B.107@uvic.ca> Message-ID: <20060627153849.GC20466@cisco.com> On Mon, Jun 26, 2006 at 09:47:15PM -0400, John Rothschild wrote: > On RHEL 4, the following works. I haven't tested it on RHEL 3. > > Add this to your packages section of your kickstart file. > > %packages --resolvedeps > > @Compatibility Arch Support > @Compatibility Arch Development Support --resolvedeps is inherent, you can't not automatically resolvedeps The key to those two groups above is that they use the mostly undocumented "biarch" directive, which tells anaconda to do what yum already does... install both arches of any given package. -- Joshua Jensen Work: 919 392 1488 Mobile: 919 454 9679 Fax: 919 392 8902 "Moderation in Moderation" From James_Martin at ao.uscourts.gov Tue Jun 27 19:07:01 2006 From: James_Martin at ao.uscourts.gov (James_Martin at ao.uscourts.gov) Date: Tue, 27 Jun 2006 15:07:01 -0400 Subject: Automatically installing 32-bit libraries on 64-bit machine? (RHEL3) In-Reply-To: <17EBC05307239C4896ED25244F42A80101367A16@permls05.wde.woodside.com.au> Message-ID: kickstart-list-bounces at redhat.com wrote on 06/26/2006 08:32:20 PM: > > All our hosts are installed via KickStart. AFAIK KickStart is not > granular enuogh to be specifying libraries down to the arch level. > Sure it is.. You just say "packagename.i386" under %packages (at least on RHEL4) if you want the 32 bit version. James S. Martin, RHCE Contractor Administrative Office of the United States Courts Washington, DC (202) 502-2394 > Good luck and I'd be interested to hear the culprit & resolution if/when > found. > > CC > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > Sent: Tuesday, 27 June 2006 8:31 AM > To: Discussion list about Kickstart > Subject: Re: Automatically installing 32-bit libraries on 64-bit > machine? (RHEL3) > > > Oddness. > > I've got Red Hat Enterprise Linux WS release 3 (Taroon Update 7). Are > you > querying on hosts installed via KickStart? > > I'm the only one currently using these hosts currently, so interference > is > not the problem here. > > If there were some way to turn this OFF in kickstart, that might be the > culprit, but the KickStart scriptage all seems pretty plain. > > The KickStart scripts are managed by xCAT (cluster application toolkit) > so > it's possible it's doing something nasty. I'll look into that. > > Thanks for your help. Curiouser and curiouser... > Drew. > > > > Coe, Colin C. wrote: > > Interesting... > > > > cat /etc/redhat-release > > Red Hat Enterprise Linux WS release 3 (Taroon Update 5) > > > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > > > -i mesa XFree86-Mesa-libGL 4.3.0 81.EL i386 > > XFree86-Mesa-libGLU 4.3.0 81.EL x86_64 > > XFree86-Mesa-libGL 4.3.0 81.EL x86_64 > > XFree86-Mesa-libGLU 4.3.0 81.EL i386 > > > > cat /etc/redhat-release > > Red Hat Enterprise Linux ES release 4 (Nahant Update 3) > > > > rpm -qa --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" | grep > > > -i mesa xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 x86_64 > > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 x86_64 > > xorg-x11-Mesa-libGL 6.8.2 1.EL.13.25 i386 > > xorg-x11-Mesa-libGLU 6.8.2 1.EL.13.25 i386 > > > > > > You should have both 32 and 64bit packages installed by default. Did > > perchance someone remove the 'duplicate' packages from the 'broken' > > installation? I've seen this done.... > > > > Both of these packages (XFree86-Mesa-libGL & XFree86-Mesa-libGLU) are > > in the 'Compatibility Arch Support' install group and, I'm pretty sure > > > that all appropriate architectures are installed by default. > > > > CC > > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com > > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > > Sent: Tuesday, 27 June 2006 8:02 AM > > To: Discussion list about Kickstart > > Subject: Re: Automatically installing 32-bit libraries on 64-bit > > machine? (RHEL3) > > > > > > Thanks Colin. I had noticed that some libraries are installed for > > both, but some definitely are not. One specific example is libGLU, > > provided by the > > XFree86-Mesa-libGLU package. > > > > [root]# rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} > > %{ARCH}\n" glibc 2.3.2 95.39 i686 glibc 2.3.2 95.39 x86_64 [root]# rpm > > > -q XFree86-Mesa-libGLU --queryformat "%{NAME} %{VERSION} %{RELEASE} > > %{ARCH}\n" XFree86-Mesa-libGLU 4.3.0 98.EL x86_64 > > > > Believe it or not our users require X libraries to run their numerical > > analysis (on non-interactive machines). I wonder if only the 64-bit > > libGLU > > is provided because nobody saw a need to have both 32-bit and 64-bit > > versions of X available. > > > > Can you suggest a way around this? At this point it looks like I'll > > have to take care of this manually or script up my own solution--if > > the two flavours > > can even co-exist for some of these packages. > > > > Thanks, > > Drew. > > > > Coe, Colin C. wrote: > >> Under RHEL3&4, this happens automatically. > >> > >> For example, querying the RPM database for info on the 'glibc' > >> package > >> gives: > >> rpm -q glibc --queryformat "%{NAME} %{VERSION} %{RELEASE} %{ARCH}\n" > >> glibc 2.3.4 2.19 i686 glibc 2.3.4 2.19 x86_64 > >> > >> The 'up2date' utility will take care of this for you. > >> > >> HTH > >> > >> CC > >> > >> -----Original Message----- > >> From: kickstart-list-bounces at redhat.com > >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Drew Leske > >> Sent: Tuesday, 27 June 2006 7:31 AM > >> To: kickstart-list at redhat.com > >> Subject: Automatically installing 32-bit libraries on 64-bit machine? > >> (RHEL3) > >> > >> > >> Hi all, > >> > >> We're using KickStart for clusters and Linux deployment in general, > >> and for the most part we are happy with this. Recently however we > >> have acquired > >> > >> some 64-bit machines to be added into our cluster and there are some > >> issues here. > >> > >> We would like to make these resources available for 64-bit computing, > >> and have therefore installed primarily 64-bit libraries on them. > >> However, most > >> of our users run 32-bit applications. At this point there is not > > enough > >> need for exclusively 64-bit machines, but I would like to make the > >> platform available. > >> > >> According to a colleague there is a way with yum to specify that > >> whenever installing a package, a 32-bit version of the package, if > >> available, will be > >> installed. I am not familiar with yum and am limited to RHEL3, > > up2date, > >> and > >> rpm, but this is the functionality I am looking for. > >> > >> I have looked at the documentation for KickStart (RHEL3 and RHEL4) as > >> well as browser this mailing list for information, and of course I've > >> searched > >> the web--nada. Does anybody have any suggestions for how to handle > >> this? > >> > >> Thanks, > >> Drew. > >> > > > > -- > Drew Leske :: Systems Group/Unix, Computing Services, University of > Victoria > dleske at uvic.ca / +1250 472 5055 (office) / +1250 588 4311 (cel) > > _______________________________________________ > 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 jay at circadence.com Thu Jun 29 14:21:52 2006 From: jay at circadence.com (Jay Hester) Date: Thu, 29 Jun 2006 08:21:52 -0600 Subject: Fedora cd error Message-ID: All, I've been having problems with installing a new cd image i'm creating. After it loads kernel, then the drivers for the hard drive array on the box the install dies with this error: Fedora Core CD was not found in any of your CDROM drives. Please insert the Fedora Core CD and press ok to retry. Here's the commands that i'm using to create the cd. Anyone have any ideas why i'm getting this message? genhdlist --productpath Fedora $CD_DIR pkgorder $CD_DIR i386 Fedora 2>/dev/null | tee /tmp/pkglist.$$ genhdlist --fileorder /tmp/pkglist.$$ --productpath Fedora $CD_DIR Thanks! Jay From Paul.Mallasch at Tectura.com Thu Jun 29 14:54:13 2006 From: Paul.Mallasch at Tectura.com (Mallasch, Paul) Date: Thu, 29 Jun 2006 07:54:13 -0700 Subject: Fedora cd error Message-ID: <5303BDF3F253554E889BE362FA0C604E0110EFA2@MAIL1.TecturaCorp.net> Jay- Are you using FC5 to build your image? If yes, then know that genhdlist is no more: http://fedoraproject.org/wiki/AnacondaCustomization?highlight=%28anacond a%29 -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Jay Hester Sent: Thursday, June 29, 2006 7:22 AM To: kickstart-list at redhat.com Subject: Fedora cd error All, I've been having problems with installing a new cd image i'm creating. After it loads kernel, then the drivers for the hard drive array on the box the install dies with this error: Fedora Core CD was not found in any of your CDROM drives. Please insert the Fedora Core CD and press ok to retry. Here's the commands that i'm using to create the cd. Anyone have any ideas why i'm getting this message? genhdlist --productpath Fedora $CD_DIR pkgorder $CD_DIR i386 Fedora 2>/dev/null | tee /tmp/pkglist.$$ genhdlist --fileorder /tmp/pkglist.$$ --productpath Fedora $CD_DIR Thanks! Jay _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From Colin.Coe at woodside.com.au Thu Jun 29 23:30:07 2006 From: Colin.Coe at woodside.com.au (Coe, Colin C.) Date: Fri, 30 Jun 2006 07:30:07 +0800 Subject: Using an existing LVM Message-ID: <17EBC05307239C4896ED25244F42A80101367A34@permls05.wde.woodside.com.au> Hi all I'm having a problem or two with re-using an existing LVM partition during a KickStart installation. This particular install is supposed to leave the /boot and /local/dsk partitions intact, over-writing /. The partition part of the ks.cfg is: zerombr yes clearpart --none part /boot --noformat --onpart sda1 part swap --onpart sda2 part / --onpart sda3 part swap --onpart sdb1 part pv.10 --noformat --onpart sda5 part pv.11 --noformat --onpart sdb2 volgroup Vol00 pv.10 pv.11 --noformat logvol /local/dsk --vgname=Vol00 --noformat --name=LocalData Any ideas what I'm doing wrong here? Using RHEL3u7 and RHEL3u8BETA. Thanks CC -------------- next part -------------- An HTML attachment was scrubbed... URL: