From skebi69 at gmail.com Tue Aug 3 18:20:22 2010 From: skebi69 at gmail.com (John Kennedy) Date: Tue, 3 Aug 2010 14:20:22 -0400 Subject: autostep --autosnapshot Message-ID: I just did a kickstart install trying to use autostep --autosnapshot to get images of the install process for training docs. The install stepped through the phases of the install as expected but once the install was complete there was no /root/anaconda-screenshots directory filled with pretty images of my new install process. Has anyone on the list gotten this to work? Thanks, John -- John Kennedy -------------- next part -------------- An HTML attachment was scrubbed... URL: From email at jasonkohles.com Wed Aug 4 15:33:38 2010 From: email at jasonkohles.com (Jason Kohles) Date: Wed, 4 Aug 2010 11:33:38 -0400 Subject: autostep --autosnapshot In-Reply-To: References: Message-ID: The argument is --autoscreenshot, not --autosnapshot... On Aug 3, 2010, at 2:20 PM, John Kennedy wrote: > I just did a kickstart install trying to use autostep --autosnapshot to get images of the install process for training docs. > The install stepped through the phases of the install as expected but once the install was complete there was no /root/anaconda-screenshots directory filled with pretty images of my new install process. > > Has anyone on the list gotten this to work? > > Thanks, > John > > -- > John Kennedy > > _______________________________________________ > 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 skebi69 at gmail.com Wed Aug 4 15:47:33 2010 From: skebi69 at gmail.com (John Kennedy) Date: Wed, 4 Aug 2010 11:47:33 -0400 Subject: autostep --autosnapshot In-Reply-To: References: Message-ID: Sorry, that was a mis-type in my email. In the .ks file it is correctly added as "autostep --autoscreenshot". Thanks for pointing that out... John On Wed, Aug 4, 2010 at 11:33 AM, Jason Kohles wrote: > The argument is --autoscreenshot, not --autosnapshot... > > On Aug 3, 2010, at 2:20 PM, John Kennedy wrote: > > I just did a kickstart install trying to use autostep --autosnapshot to get > images of the install process for training docs. > The install stepped through the phases of the install as expected but once > the install was complete there was no /root/anaconda-screenshots directory > filled with pretty images of my new install process. > > Has anyone on the list gotten this to work? > > Thanks, > John > > -- > John Kennedy > > _______________________________________________ > 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 > -- John Kennedy -------------- next part -------------- An HTML attachment was scrubbed... URL: From joe.hoot at itec.suny.edu Wed Aug 4 16:24:05 2010 From: joe.hoot at itec.suny.edu (Hoot, Joseph) Date: Wed, 4 Aug 2010 12:24:05 -0400 Subject: Assigning VLAN tag during install prior to kickstart being located Message-ID: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> Hi all, I originally posted this yesterday to the anaconda list, but am posting this here too in case others have run into this problem and may have a resolution. I have been digging through google searches and list archives for both kickstart and anaconda for about 45 minutes now and can't seem to find the answer that I am looking for (hoping for). I have been using a kickstart file to install many servers in our environment. The problem that I've run into is when I need to install to a server that is configured with trunk ports. For now, I've lived with recompiling the iso to include the kickstart file. But I would prefer to use the default iso and just use `linux ks=http://etc..etc../ks.cfg`. So my problem now is that, although this works just fine if I have access ports configured, since the `asknetwork` option for anaconda doesn't allow for a VLAN tag to be set. QUESTIONS: =========== 1) Is there an option that I can pass so that anaconda will support a vlan tag prior to kickstarting? 2) If anaconda doesn't currently support it, is there a feature request for it to be supported in the future? REFERENCES: ============ A) https://bugzilla.redhat.com/show_bug.cgi?id=431915 B) RHBA-2009:0164-13: -- http://rhn.redhat.com/errata/RHBA-2009-0164.html -- where is says "431915 - feature request: add 802.1q vlan tagging to anaconda" Thanks Joe =========================== Joseph R. Hoot Lead System Programmer/Analyst (w) 716-878-4832 (c) 716-759-HOOT joe.hoot at itec.suny.edu GPG KEY: 7145F633 =========================== From courchea at net-forces.com Thu Aug 5 12:38:23 2010 From: courchea at net-forces.com (Andre Courchesne) Date: Thu, 5 Aug 2010 08:38:23 -0400 Subject: internet install, kickstart and pppoe Message-ID: Anyone has any recommendation on making a net-install ISO and supporting pppoe connections to the internet? Any hints even far fetched are welcome. --- Andre Courchesne - Consultant http://www.net-forces.com MSN: courchea at net-forces.com Skype: VoipForces L'information contenue dans le pr?sent document est la propri?t? de Andre Courchesne. Et est divulgu?e en toute confidentialit?. Cette information ne doit pas ?tre utilis?e, divulgu?e ? d'autres personnes ou reproduite sans le consentement ?crit explicite de Andre Courchesne. The information contained in this document is confidential and property of Andre Courchesne. It shall not be used, disclosed to others or reproduced without the express written consent of Andre Courchesne. From kiran.sashi at gmail.com Sun Aug 15 10:42:47 2010 From: kiran.sashi at gmail.com (sashi kiran) Date: Sun, 15 Aug 2010 16:12:47 +0530 Subject: Issue in ks.cfg (anaconda-13.21.56-1.el6) Message-ID: *Description of problem:* We use RHEL 6 (snapshot 7) for unattended OS deployment. In the ks.cfg, we used python in %pre section to find hard disk. The python module "kudzu" is not find in RHEL6, which we where using for all our unattended deployments. Also we used doGetBiosDisk function from isys module to find the bios boot disk. This function also seems deprecated. Please suggest the alternate for kudzu module and doGetBiosDisk. *Code Snippet:* >>> %pre --interpreter=/usr/bin/python import kudzu,string,sys ... ... bootDisk=isys.doGetBiosDisk("80") disks=kudzu.probe(kudzu.CLASS_HD,kudzu.BUS_UNSPEC,kudzu.PROBE_ALL) >> *Version-Release number: *RedHat EL 6 (snapshot 7) anaconda-13.21.56-1.el6 Any help is highly appreciated. Thanks, Sashi. -------------- next part -------------- An HTML attachment was scrubbed... URL: From Moray.Henderson at ict-software.org Mon Aug 16 09:01:57 2010 From: Moray.Henderson at ict-software.org (Moray Henderson) Date: Mon, 16 Aug 2010 10:01:57 +0100 Subject: Issue in ks.cfg (anaconda-13.21.56-1.el6) In-Reply-To: References: Message-ID: <000001cb3d21$af864ae0$0e92e0a0$@Henderson@ict-software.org> From: sashi kiran [mailto:kiran.sashi at gmail.com] We use RHEL 6 (snapshot 7) for unattended OS deployment. In the ks.cfg, we used python in %pre section to find hard disk. The python module "kudzu" is not find in RHEL6, which we where using for all our unattended deployments. Also we used doGetBiosDisk function from isys module to find the bios boot disk. This function also seems deprecated. Please suggest the alternate for kudzu module and doGetBiosDisk. There was something about this on the Anaconda list just recently: >-----Original Message----- >From: Hans de Goede [mailto:hdegoede at redhat.com] >On 08/05/2010 07:21 PM, Paniraja_KM at dell.com wrote: >> In old anaconda, if isys is imported and a call doGetBiosDisk like >> isys.doGetBiosDisk("80"), I used to get bios boot disk. >> Any alternative now to get the bios boot disk from python? >> >In the new anaconda code the storage object has an eddDict attribute >which contains for example: >{ "sda": 80, "sdb": 81 } > >This attribute is not available until the disks have been scanned. Moray. "To err is human. To purr, feline" -------------- next part -------------- An HTML attachment was scrubbed... URL: From joe.hoot at itec.suny.edu Mon Aug 16 10:59:55 2010 From: joe.hoot at itec.suny.edu (Hoot, Joseph) Date: Mon, 16 Aug 2010 06:59:55 -0400 Subject: Assigning VLAN tag during install prior to kickstart being located In-Reply-To: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> References: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> Message-ID: Anyone? On Aug 4, 2010, at 12:24 PM, Hoot, Joseph wrote: > Hi all, > > I originally posted this yesterday to the anaconda list, but am posting this here too in case others have run into this problem and may have a resolution. > > I have been digging through google searches and list archives for both kickstart and anaconda for about 45 minutes now and can't seem to find the answer that I am looking for (hoping for). I have been using a kickstart file to install many servers in our environment. The problem that I've run into is when I need to install to a server that is configured with trunk ports. For now, I've lived with recompiling the iso to include the kickstart file. But I would prefer to use the default iso and just use `linux ks=http://etc..etc../ks.cfg`. So my problem now is that, although this works just fine if I have access ports configured, since the `asknetwork` option for anaconda doesn't allow for a VLAN tag to be set. > > QUESTIONS: > =========== > 1) Is there an option that I can pass so that anaconda will support a vlan tag prior to kickstarting? > 2) If anaconda doesn't currently support it, is there a feature request for it to be supported in the future? > > > REFERENCES: > ============ > A) https://bugzilla.redhat.com/show_bug.cgi?id=431915 > B) RHBA-2009:0164-13: > -- http://rhn.redhat.com/errata/RHBA-2009-0164.html > -- where is says "431915 - feature request: add 802.1q vlan tagging to anaconda" > > Thanks > Joe > =========================== > Joseph R. Hoot > Lead System Programmer/Analyst > (w) 716-878-4832 > (c) 716-759-HOOT > joe.hoot at itec.suny.edu > GPG KEY: 7145F633 > =========================== > From angus.clarke at gmail.com Mon Aug 16 12:24:36 2010 From: angus.clarke at gmail.com (Angus Clarke) Date: Mon, 16 Aug 2010 13:24:36 +0100 (BST) Subject: Assigning VLAN tag during install prior to kickstart being located In-Reply-To: Message-ID: <1225255411.561634.1281961476876.JavaMail.root@unx-s-manc3> I cannot answer your query, but this might help. I ask our network guys to assign a default VLAN to switch ports. I don't know the correct networking terminology, so I will describe the process Server boots to perform kickstart sends DHCP query (no VLAN tag) Switch notices untagged packets and assigns them to "default" VLAN I use this "default" VLAN specifically for building servers, and use it across multiple target networks. Final network configuration (derived from kickstart template) can be whatever you want (e.g. include VLAN tag info for other networks) This system works pretty well for me. Hope this helps ~Angus ----- Original Message ----- From: "Joseph Hoot" To: kickstart-list at redhat.com Sent: Monday, 16 August, 2010 12:59:55 PM Subject: Re: Assigning VLAN tag during install prior to kickstart being located Anyone? On Aug 4, 2010, at 12:24 PM, Hoot, Joseph wrote: > Hi all, > > I originally posted this yesterday to the anaconda list, but am posting this here too in case others have run into this problem and may have a resolution. > > I have been digging through google searches and list archives for both kickstart and anaconda for about 45 minutes now and can't seem to find the answer that I am looking for (hoping for). I have been using a kickstart file to install many servers in our environment. The problem that I've run into is when I need to install to a server that is configured with trunk ports. For now, I've lived with recompiling the iso to include the kickstart file. But I would prefer to use the default iso and just use `linux ks=http://etc..etc../ks.cfg`. So my problem now is that, although this works just fine if I have access ports configured, since the `asknetwork` option for anaconda doesn't allow for a VLAN tag to be set. > > QUESTIONS: > =========== > 1) Is there an option that I can pass so that anaconda will support a vlan tag prior to kickstarting? > 2) If anaconda doesn't currently support it, is there a feature request for it to be supported in the future? > > > REFERENCES: > ============ > A) https://bugzilla.redhat.com/show_bug.cgi?id=431915 > B) RHBA-2009:0164-13: > -- http://rhn.redhat.com/errata/RHBA-2009-0164.html > -- where is says "431915 - feature request: add 802.1q vlan tagging to anaconda" > > Thanks > Joe > =========================== > Joseph R. Hoot > Lead System Programmer/Analyst > (w) 716-878-4832 > (c) 716-759-HOOT > joe.hoot at itec.suny.edu > GPG KEY: 7145F633 > =========================== > _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From tony.chamberlain at lemko.com Tue Aug 17 12:47:17 2010 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Tue, 17 Aug 2010 12:47:17 +0000 Subject: driver (5.4) Message-ID: We are installing plenty of machines. For them we have to specify "dd" and then during installation it will ask us for a driver disk. We have on floppy and on USB drive. Floppy works always (it can read it). USB drive for some reason works only the first time. Afterwards we get a "failed to mount disk" message. Anyway I figure there must be a way either to have the driver image on a network somewhere that I can access, or build it right into the Cent OS 5.4 installation disk. That is, in the kick start file somehow specify dd=http:// or ftp:// or something, or create a new installation disk (which I have done many times before) with the additional driver (MegaSR) in it (which I have never done before). Is there some way to locate and specify an additional driver without the necessity of having to put it on a separate floppy or USB thumb drive and needing special hardware (floppy reader, thumb drive) not to mention a free USB port? From jimi at sngx.net Tue Aug 17 12:58:26 2010 From: jimi at sngx.net (James Cammarata) Date: Tue, 17 Aug 2010 07:58:26 -0500 Subject: driver (5.4) In-Reply-To: References: Message-ID: <6d66b4f95c7cf57e529ffc17cf737241@sngx.net> On Tue, 17 Aug 2010 12:47:17 +0000, tony.chamberlain at lemko.com wrote: > We are installing plenty of machines. For them we have to specify > "dd" and then during installation it will ask us for a driver disk. > We have on floppy and on USB drive. Floppy works always (it can read it). > USB drive for some reason works only the first time. Afterwards we get > a "failed to mount disk" message. > > Anyway I figure there must be a way either to have the driver image on > a network somewhere that I can access, or build it right into the Cent OS > 5.4 > installation disk. That is, in the kick start file somehow specify > dd=http:// or ftp:// or something, or create a new installation disk > (which I have done many times before) with the additional driver (MegaSR) > in it (which I have never done before). > > Is there some way to locate and specify an additional driver without the > necessity of having to put it on a separate floppy or USB thumb drive > and needing special hardware (floppy reader, thumb drive) not to mention a > free USB port? Typically we wget the driver file(s) and install them in %pre, and then again maybe in %post depending on the driver and what it does. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From bda20 at cam.ac.uk Tue Aug 17 13:26:42 2010 From: bda20 at cam.ac.uk (Ben) Date: Tue, 17 Aug 2010 14:26:42 +0100 (BST) Subject: Weird kickstart problem regarding not selecting the right cfg Message-ID: We recently acquired four new machines (Sun X4270s). As usual as I've done with dozens of other machines I constructed RHEL5 kickstart .cfg files for them and made additions to the syslinux.cfg of the form: label ct-phy-dom01 kernel vmli5s64 append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom01.cfg console=ttyS0,9600 label ct-phy-dom02 kernel vmli5s64 append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom02.cfg console=ttyS0,9600 label ct-phy-dom03 kernel vmli5s64 append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom03.cfg console=ttyS0,9600 label ct-phy-dom04 kernel vmli5s64 append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom04.cfg console=ttyS0,9600 # vmli5s64 = vmlinuz for RHEL5, server, 64bit # int5s64.img = initrd.img for RHEL5, server, 64bit # (note that the name of the initrd file must conform to 8.3 rules, just # like in DOS or it doesn't work!) I placed the machine-specific .cfg files and the syslinux.cfg on a USB stick as we do for certain installs and powered the first machine (ct-phy-dom01) on. After typing "ct-phy-dom01" at the boot: line the machine installed via HTTP perfectly. However after that all the other machines failed to install. After watching ct-phy-dom02-4 repeatedly stop at the point at which they should load stage2.img from the HTTP server I theorised it was something to do with the network. Cue over an hour of network testing, changing cables, checking VLANs and adding ksdevice= to the syslinux.cfg. Nothing worked. At no point could I ping the IP (in the ks.cfg) of the box I was installing. Eventually, even though I was booting via serial I plugged in a monitor and paged through the other virtual consoles and noticed that at install the machines were all selecting the ct-phy-dom01.cfg file despite what the syslinux.cfg specified. Obviously this meant they were trying to raise the IP of ct-phy-dom01 which had already been built and was up and running. No matter what I tried I could not get any of the other machines (ct-phy-dom02-04) to use their own cfg files until I started commenting out the lines of servers I wasn't building in syslinux.cfg. Want to build ct-phy-dom02? Comment out ct-phy-dom01's lines. ct-phy-dom03? Comment out ct-phy-dom01 _and_ ct-phy-dom02. Etc. Now, can anyone explain why? Ben -- Unix Support, MISD, University of Cambridge, England Plugger of wire, typer of keyboard, imparter of Clue Life Is Short. It's All Good. From Michael.Peterson at duke.edu Tue Aug 17 13:36:47 2010 From: Michael.Peterson at duke.edu (Michael Peterson) Date: Tue, 17 Aug 2010 09:36:47 -0400 Subject: Weird kickstart problem regarding not selecting the right cfg In-Reply-To: References: Message-ID: <4C6A906F.9060105@duke.edu> On 8/17/10 9:26 AM, Ben wrote: > We recently acquired four new machines (Sun X4270s). As usual as I've > done with dozens of other machines I constructed RHEL5 kickstart .cfg > files for them and made additions to the syslinux.cfg of the form: > > label ct-phy-dom01 > kernel vmli5s64 > append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom01.cfg > console=ttyS0,9600 > label ct-phy-dom02 > kernel vmli5s64 > append initrd=int5s64.img ks=hd:sdb1:/configs/ct-phy-dom02.cfg > console=ttyS0,9600 > # vmli5s64 = vmlinuz for RHEL5, server, 64bit > # int5s64.img = initrd.img for RHEL5, server, 64bit > # (note that the name of the initrd file must conform to 8.3 rules, just > # like in DOS or it doesn't work!) > > No matter what I tried I could not get any of the other machines > (ct-phy-dom02-04) to use their own cfg files until I started commenting > out the lines of servers I wasn't building in syslinux.cfg. Want to > build ct-phy-dom02? Comment out ct-phy-dom01's lines. ct-phy-dom03? > Comment out ct-phy-dom01 _and_ ct-phy-dom02. Etc. > > Now, can anyone explain why? > > Ben As you note for the initrd.img filenames, the labels are "mangled" into 8.3 format as well, so all of your labels are effectively the same. See http://syslinux.zytor.com/wiki/index.php/SYSLINUX#LABEL_command Michael -- Michael A. Peterson, PhD. IT Manager / HPC Systems Admin Department of Chemistry Duke University 919-660-1646 "PC" != "Windows" From bda20 at cam.ac.uk Tue Aug 17 13:41:06 2010 From: bda20 at cam.ac.uk (Ben) Date: Tue, 17 Aug 2010 14:41:06 +0100 (BST) Subject: Weird kickstart problem regarding not selecting the right cfg In-Reply-To: References: Message-ID: On Tue, 17 Aug 2010, Ben wrote: > We recently acquired four new machines (Sun X4270s). As usual as I've done > with dozens of other machines I constructed RHEL5 kickstart .cfg files for > them and made additions to the syslinux.cfg of the form: > > label ct-phy-dom01 > [...] So basically it turns out that label directives get mangled DOS-like, too. http://syslinux.zytor.com/wiki/index.php/SYSLINUX#LABEL_command Who knew?* Ben * Well, obviously someone did, and now I do too -- Unix Support, MISD, University of Cambridge, England Plugger of wire, typer of keyboard, imparter of Clue Life Is Short. It's All Good. From jcm at redhat.com Tue Aug 17 16:50:48 2010 From: jcm at redhat.com (Jon Masters) Date: Tue, 17 Aug 2010 12:50:48 -0400 Subject: driver (5.4) In-Reply-To: References: Message-ID: <1282063848.27736.303.camel@constitution.bos.jonmasters.org> On Tue, 2010-08-17 at 12:47 +0000, tony.chamberlain at lemko.com wrote: > Anyway I figure there must be a way either to have the driver image on > a network somewhere that I can access, or build it right into the Cent OS 5.4 > installation disk. That is, in the kick start file somehow specify > dd=http:// or ftp:// or something, or create a new installation disk > (which I have done many times before) with the additional driver (MegaSR) > in it (which I have never done before). There is a way to do updates in RHEL over the network, separate from kickstart, by means of including the update into the TFTP data retrieved over the network and appending to the initrd used by the installer. I see that you're not using RHEL though, so you should ask your vendor what they do about handling this kind of update. Jon. From joe.hoot at itec.suny.edu Wed Aug 18 13:22:21 2010 From: joe.hoot at itec.suny.edu (Hoot, Joseph) Date: Wed, 18 Aug 2010 09:22:21 -0400 Subject: Assigning VLAN tag during install prior to kickstart being located In-Reply-To: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> References: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> Message-ID: OK. So I have only heard back from one person with regards to this. The suggestion was to set a default vlan on the switch such that IP that I set would be able to route over to the kickstart server. I know that this will work, however, after that I would want to set it back to the switch default vlan for compatibility reasons. And my goal in doing this is being able to put in one request to the network admins to setup ports for me and then not have to submit additional requests later. So it sounds like getting anaconda to be able to add on a vlan tag isn't currently doable. What are the proper procedures that I can follow to submit a feature request to be able to add a vlan tag prior to attempting to read the kickstart file? Thanks, Joe On Aug 4, 2010, at 12:24 PM, Hoot, Joseph wrote: > Hi all, > > I originally posted this yesterday to the anaconda list, but am posting this here too in case others have run into this problem and may have a resolution. > > I have been digging through google searches and list archives for both kickstart and anaconda for about 45 minutes now and can't seem to find the answer that I am looking for (hoping for). I have been using a kickstart file to install many servers in our environment. The problem that I've run into is when I need to install to a server that is configured with trunk ports. For now, I've lived with recompiling the iso to include the kickstart file. But I would prefer to use the default iso and just use `linux ks=http://etc..etc../ks.cfg`. So my problem now is that, although this works just fine if I have access ports configured, since the `asknetwork` option for anaconda doesn't allow for a VLAN tag to be set. > > QUESTIONS: > =========== > 1) Is there an option that I can pass so that anaconda will support a vlan tag prior to kickstarting? > 2) If anaconda doesn't currently support it, is there a feature request for it to be supported in the future? > > > REFERENCES: > ============ > A) https://bugzilla.redhat.com/show_bug.cgi?id=431915 > B) RHBA-2009:0164-13: > -- http://rhn.redhat.com/errata/RHBA-2009-0164.html > -- where is says "431915 - feature request: add 802.1q vlan tagging to anaconda" > > Thanks > Joe > =========================== > Joseph R. Hoot > Lead System Programmer/Analyst > joe.hoot at itec.suny.edu > GPG KEY: 7145F633 > =========================== > From hhoffman at ip-solutions.net Wed Aug 18 13:40:19 2010 From: hhoffman at ip-solutions.net (Harry Hoffman) Date: Wed, 18 Aug 2010 09:40:19 -0400 Subject: Assigning VLAN tag during install prior to kickstart being located In-Reply-To: References: <2A61B7CB-32FD-429E-AF9A-3E249970A231@itec.suny.edu> Message-ID: <1282138819.21947.9.camel@n1-20-152.dhcp.drexel.edu> Hi Joe, >From the second link you posted it seems like VLANS are now supported via the kernel module in the initrd as of 11/12/2008. Perhaps the docs haven't been updated to reflect this? Or maybe I'm just reading it wrong. If you want to manually set the switches, then snmp is your best bet. If you have the write community strings you can use the OID that is particular for the switch manuf to set the untagged (native) vlan per switch port. Then when the install is done another snmp call to set it to some other vlan. Cheers, Harry On Wed, 2010-08-18 at 09:22 -0400, Hoot, Joseph wrote: > OK. So I have only heard back from one person with regards to this. The suggestion was to set a default vlan on the switch such that IP that I set would be able to route over to the kickstart server. I know that this will work, however, after that I would want to set it back to the switch default vlan for compatibility reasons. And my goal in doing this is being able to put in one request to the network admins to setup ports for me and then not have to submit additional requests later. > > So it sounds like getting anaconda to be able to add on a vlan tag isn't currently doable. What are the proper procedures that I can follow to submit a feature request to be able to add a vlan tag prior to attempting to read the kickstart file? > > Thanks, > Joe > > > On Aug 4, 2010, at 12:24 PM, Hoot, Joseph wrote: > > > Hi all, > > > > I originally posted this yesterday to the anaconda list, but am posting this here too in case others have run into this problem and may have a resolution. > > > > I have been digging through google searches and list archives for both kickstart and anaconda for about 45 minutes now and can't seem to find the answer that I am looking for (hoping for). I have been using a kickstart file to install many servers in our environment. The problem that I've run into is when I need to install to a server that is configured with trunk ports. For now, I've lived with recompiling the iso to include the kickstart file. But I would prefer to use the default iso and just use `linux ks=http://etc..etc../ks.cfg`. So my problem now is that, although this works just fine if I have access ports configured, since the `asknetwork` option for anaconda doesn't allow for a VLAN tag to be set. > > > > QUESTIONS: > > =========== > > 1) Is there an option that I can pass so that anaconda will support a vlan tag prior to kickstarting? > > 2) If anaconda doesn't currently support it, is there a feature request for it to be supported in the future? > > > > > > REFERENCES: > > ============ > > A) https://bugzilla.redhat.com/show_bug.cgi?id=431915 > > B) RHBA-2009:0164-13: > > -- http://rhn.redhat.com/errata/RHBA-2009-0164.html > > -- where is says "431915 - feature request: add 802.1q vlan tagging to anaconda" > > > > Thanks > > Joe > > =========================== > > Joseph R. Hoot > > Lead System Programmer/Analyst > > joe.hoot at itec.suny.edu > > GPG KEY: 7145F633 > > =========================== > > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Floydsmith at aol.com Tue Aug 24 19:45:31 2010 From: Floydsmith at aol.com (Floydsmith at aol.com) Date: Tue, 24 Aug 2010 15:45:31 EDT Subject: need to know what repositories are needed for revisor in fc 13 Message-ID: I am trying to run revisor with the following cmdline: [u at floydsmi ~]$ sudo revisor --cli --model=f13-i386 --kickstart=/dosc/anaconda-ks.cfg --install-dvd And I get an error message saying it can't communicate with the repositories. Here are my yum repositories: # yum -- repolist Loaded plugins: presto, refresh-packagekit repo id repo name status adobe-linux-i386 Adobe Systems Incorporated 18 fedora Fedora 13 - i386 16,814 updates Fedora 13 - i386 - Updates 4,827 repolist: 21,659 What lines do I have to add and to what file? I have never had to add anything in the past. Any help greatly appreciated in advance. Floyd, -------------- next part -------------- An HTML attachment was scrubbed... URL: From streamscalelance at gmail.com Wed Aug 25 19:55:19 2010 From: streamscalelance at gmail.com (Lance Gropper) Date: Wed, 25 Aug 2010 12:55:19 -0700 Subject: Identifying boot device Message-ID: Hello All: I am wondering if there is a way to identify what device is running the kickstart file? I am working on a kickstart USB key, and on some systems, it comes up as /dev/sdb, sometimes /dev/sda, and sometimes /dev/hda...Knowing what device the kickstart is running from will tell me what device I am installing to...i.e. if I am booting from /dev/sdb, I'm installing to /dev/sda. If I'm booting from /dev/sda, I'm installing to /dev/hda. If I am booting from /dev/hda, I'm installing to /dev/sda. Lance -------------- next part -------------- An HTML attachment was scrubbed... URL: From linux at alteeve.com Wed Aug 25 21:08:15 2010 From: linux at alteeve.com (Digimer) Date: Wed, 25 Aug 2010 17:08:15 -0400 Subject: Odd problem with RAID -> LVM Message-ID: <4C75863F.9010106@alteeve.com> Hi all, I've been bashing my head all day trying to get a kickstart working that takes two disks, creates two RAID1 devices, puts an LVM on one and then creates two LVs. The server has 2x 500GB hard drives. Here is the script, thus far: http://wiki.alteeve.com/index.php/Fedora13_KS_xenmaster001.ks As it is now, anaconda fails and the trace says that there is not enough space for the LV. This made me think that "--grow" wasn't being used. So I changed the "--size=500 --grow" to "--size=50000 --grow" and it installed. However! This is what pvscan shows me post install: ------------------------------------------------------------------------- [root at xenmaster001 ~]# pvscan PV /dev/md1 VG vg01 lvm2 [232.75 GiB / 191.69 GiB free] PV /dev/sdb3 lvm2 [232.76 GiB] Total: 2 [465.51 GiB] / in use: 1 [232.75 GiB] / in no VG: 1 [232.76 GiB] ------------------------------------------------------------------------- It looks like it created TWO physical volumes. If I look at the actual drives, it gets even weirder: ------------------------------------------------------------------------- [root at xenmaster001 ~]# fdisk /dev/sda WARNING: DOS-compatible mode is deprecated. It's strongly recommended to switch off the mode (command 'c') and change display units to sectors (command 'u'). Command (m for help): p Disk /dev/sda: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x00064e85 Device Boot Start End Blocks Id System /dev/sda1 * 1 32 256000 fd Linux raid autodetect Partition 1 does not end on cylinder boundary. /dev/sda2 32 60802 488129536 fd Linux raid autodetect Command (m for help): q [root at xenmaster001 ~]# fdisk /dev/sdb WARNING: DOS-compatible mode is deprecated. It's strongly recommended to switch off the mode (command 'c') and change display units to sectors (command 'u'). Command (m for help): p Disk /dev/sdb: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x00053b7f Device Boot Start End Blocks Id System /dev/sdb1 * 1 32 256000 fd Linux raid autodetect Partition 1 does not end on cylinder boundary. /dev/sdb2 32 30417 244064256 fd Linux raid autodetect /dev/sdb3 30417 60802 244064256 8e Linux LVM Command (m for help): q ------------------------------------------------------------------------- How screwed up is that? lol! So, I decided to run a manual install to see what kickstart script was created. That came out really weird, too: ------------------------------------------------------------------------- #clearpart --none --drives=sda,sdb #ignoredisk --only-use=sda,sdb #raid /boot --fstype=ext3 --level=1 --device=md0 raid.2db68162-2542-41c7-baca-d2b76328d459 raid.None #raid pv.84iuZQ-4p9R-iOK1-XfJr-pfPz-1QJQ-UqLPYW --level=1 --device=md1 raid.None raid.None #part raid.None --onpart=sda1 #part raid.None --grow --asprimary --size=500 #part raid.None --asprimary --size=250 #part raid.None --grow --asprimary --size=500 #volgroup vg_01 --pesize=32768 pv.84iuZQ-4p9R-iOK1-XfJr-pfPz-1QJQ-UqLPYW #logvol / --fstype=ext3 --name=lv_root --vgname=vg_01 --size=40864 #logvol swap --name=lv_swap --vgname=vg_01 --size=2048 bootloader --location=mbr --driveorder=sda,sdb --append="rhgb quiet" ------------------------------------------------------------------------- Am I doing something terribly daft, or is there a bug? Thanks! -- Digimer E-Mail: linux at alteeve.com AN!Whitepapers: http://alteeve.com Node Assassin: http://nodeassassin.org From streamscalelance at gmail.com Fri Aug 27 02:11:16 2010 From: streamscalelance at gmail.com (Lance Gropper) Date: Thu, 26 Aug 2010 19:11:16 -0700 Subject: harddrive in %pre Message-ID: Hello Ya'all: I am writing a script to install from a USB key, however it appears as a different device on several different systems, so I put the harddisk command into an include, generated by the %pre area, however the system seems to be ignoring this - is there a way to make it work, or are there commands which just can't be put into includes? Lance -------------- next part -------------- An HTML attachment was scrubbed... URL: From AndrewJohnson at HBOSplc.com Fri Aug 27 14:10:59 2010 From: AndrewJohnson at HBOSplc.com (AndrewJohnson at HBOSplc.com) Date: Fri, 27 Aug 2010 15:10:59 +0100 Subject: Dynamically building on spare device / boot device. Message-ID: <201008271411.o7REBokx014772@mx1.redhat.com> Hi, I'm trying to identify the best approach to dynamically identify whether a client has a spare disk to build Logical Volumes against or whether I need to deploy them onto the rootdisk. At present i've built the post script that deliveries the following: Validate Volume Groups, If no additional Volume Groups available Built on RootVG Validate Available Space and Add LV's However I'm wondering on how to accuratly identify which disk is free and what approach should be taken to build a VG on it with removing the wrong disk, i'm new to this so any help would be greatly appreciated, i've tried a couple of %pre scripts but assuming rootvg is always going to be on sda isn't always true. Kind Regards Andy ----------------------------------------------------------------------------------------------------- Lloyds TSB Bank plc. Registered Office: 25 Gresham Street, London EC2V 7HN. Registered in England and Wales, number 2065. Telephone: 020 7626 1500. Bank of Scotland plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 327000. Telephone: 0870 600 5000 Lloyds TSB Scotland plc. Registered Office: Henry Duncan House, 120 George Street, Edinburgh EH2 4LH. Registered in Scotland, number 95237. Telephone: 0131 225 4555. Cheltenham & Gloucester plc. Registered Office: Barnett Way, Gloucester GL4 3RL. Registered in England and Wales, number 2299428. Telephone: 01452 372372. Lloyds TSB Bank plc, Lloyds TSB Scotland plc, Bank of Scotland plc and Cheltenham & Gloucester plc are authorised and regulated by the Financial Services Authority. Halifax is a division of Bank of Scotland plc. Cheltenham & Gloucester Savings is a division of Lloyds TSB Bank plc. HBOS plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 218813. Telephone: 0870 600 5000 Lloyds Banking Group plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 95000. Telephone: 0131 225 4555 This e-mail (including any attachments) is private and confidential and may contain privileged material. If you have received this e-mail in error, please notify the sender and delete it (including any attachments) immediately. You must not copy, distribute, disclose or use any of the information in it or any attachments. Telephone calls may be monitored or recorded. ============================================================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: From clumens at redhat.com Fri Aug 27 14:51:59 2010 From: clumens at redhat.com (Chris Lumens) Date: Fri, 27 Aug 2010 10:51:59 -0400 Subject: harddrive in %pre In-Reply-To: References: Message-ID: <20100827145158.GU30596@exeter.usersys.redhat.com> > I am writing a script to install from a USB key, however it appears as a > different device on several different systems, so I put the harddisk command > into an include, generated by the %pre area, however the system seems to be > ignoring this - is there a way to make it work, or are there commands which > just can't be put into includes? Any of the method selection commands (harddrive/nfs/url/cdrom) are the only ones that cannot be put in an include. There are ways to do what you're trying to do, but I'll need a little more context to make a suggestion. Can you post your kickstart file, minus any sensitive information? - Chris From AndrewJohnson at HBOSplc.com Fri Aug 27 15:11:10 2010 From: AndrewJohnson at HBOSplc.com (AndrewJohnson at HBOSplc.com) Date: Fri, 27 Aug 2010 16:11:10 +0100 Subject: harddrive in %pre In-Reply-To: References: Message-ID: <201008271511.o7RFBiAe020028@mx1.redhat.com> Hi There... I've stripped it down as much as possible, as you can see I'm defining 2 disks however I cannot gaurentee that these 2 disks will always exist and am looking for something more dynamic, i.e find a way of scanning for them at pre-setup time and manually setting it up if it exists, as these are scsi devices I'm trying to associate the sda / sdb to what would be the correct luns etc which would be used in defining the rootvg and datavg etc.? Excuse the perl not had time to clean it up, but this would be used to validate / build websphere volumes based on build time parameters I pick up from /proc/cmdline and if there is the existance of an additional volume group build on that otherwise build on rootvg. As always any help would be greatly appreciated. install text network --bootproto dhcp lang en_US keyboard uk zerombr clearpart --all part /boot --fstype ext3 --size=100 --ondisk=sda part pv.00 --size=0 --grow --ondisk=sda part pv.01 --size=0 --grow --ondisk=sdb volgroup datavg --pesize=32768 pv.01 volgroup rootvg --pesize=32768 pv.00 logvol / --fstype ext3 --name=rootlv --vgname=rootvg --size=2048 logvol /home --fstype ext3 --name=homelv --vgname=rootvg --size=2048 logvol /opt --fstype ext3 --name=optlv --vgname=rootvg --size=1280 logvol /tmp --fstype ext3 --name=tmplv --vgname=rootvg --size=128 logvol /usr --fstype ext3 --name=usrlv --vgname=rootvg --size=5120 logvol /var --fstype ext3 --name=varlv --vgname=rootvg --size=2048 logvol swap --fstype swap --name=paginglv --vgname=rootvg --size=2048 bootloader --location mbr timezone Europe/London auth --enablemd5 --enableshadow rootpw --iscrypted XXXXXXXX selinux --disabled reboot firewall --disabled skipx key --skip %packages bridge-utils libXp @ Base compat-libstdc++-33 device-mapper-multipath %pre %post --nochroot ( set -x mkdir /mnt/sysimage/tmp/ks-tree-copy if [ -d /oldtmp/ks-tree-shadow ]; then cp -fa /oldtmp/ks-tree-shadow/* /mnt/sysimage/tmp/ks-tree-copy elif [ -d /tmp/ks-tree-shadow ]; then cp -fa /tmp/ks-tree-shadow/* /mnt/sysimage/tmp/ks-tree-copy fi cp /etc/resolv.conf /mnt/sysimage/etc/resolv.conf cp -f /tmp/ks-pre.log /mnt/sysimage/root/ ) >> /mnt/sysimage/root/ks-post.log 2>&1 %post ( # Log %post errors rhn_check ) >> /root/ks-post.log 2>&1 # Start post_install_network_config generated code # End post_install_network_config generated code # MOTD echo >> /etc/motd echo "RHN Satellite kickstart on $(date +'%Y-%m-%d')" >> /etc/motd echo >> /etc/motd # end of generated kickstart file %post --interpreter /usr/bin/perl my $vg_data = "datavg"; my $backup_vg = "rootvg"; my $vgs = '/usr/sbin/vgs'; sub extract_bootparam{ open (DATA, "/proc/cmdline") or die "cannot open file $!\n"; my @item=split(/[ ]+/,); for (@item){ chomp; next if !/^name|role/; push(@bp_array,$_); } close (DATA); return (@bp_array); } sub validate_bootparam{ (my $item1,my $item2)=@_; if ( ($item1 eq "") || ($item2 eq "")){ die "No Defined Boot Parameters [ name= role= ]";} (my $item1a,my $item1b)=split('=',$item1); (my $item2a,my $item2b)=split('=',$item2); if (($item1a =~ /name/)&&($item2a =~ /role/)){return ($item1b,$item2b);} elsif (($item2a =~ /name/) && ($item1a =~ /role/)){return ($item2b,$item1b);} else{die "No Defined Boot Parameters [ name= role= ]";} } sub validate_vg{ print "Validate Volume Group : $vg_data\n"; my $result_vg=`$vgs --options vg_name --noheadings $vg_data 2>&1`; if ( $result_vg =~ /not found/ ){ print "ERROR!! $result_vg Attempting to validate rootvg\n"; $vg_data=$backup_vg; my $result_vg=`$vgs --options vg_name --noheadings $vg_data 2>&1`; if ( $result_vg =~ /not found/ ){ die "ERROR!! $result_vg.. Exiting...\n"; }else{ print "$vg_data has been validated... continuing\n"; } } open(VG_DATA,"$vgs --options vg_name,vg_free --noheadings --units M $vg_data 2>/dev/null|") or die "LVM Error: $!"; for (){; chomp; s/^\s+//; (my $item1, my $item2)=split(/[ ]+/,$_); return((split(/\./,$item2))[0]); } close (VG_DATA); } sub create_group{ (my $item1, my $item2)=@_; print "Create Group\n"; print "============\n"; my $item1_gname=(getgrnam($item1))[0]; my $item2_gid=(getgrgid($item2))[2]; if ( $item1_gname eq $item1 ){ print "ERROR!! Group $item1 not created, group already exists...\n"; }elsif ( $item2_gid == $item2 ){ print "ERROR!! Group $item1 not created, group-id already exists ...\n"; }else{ my $result_grp=`/usr/sbin/groupadd -g $item2 $item1 2>&1`; if ( $result_grp =~ /exists|unique/){print "ERROR!! Creating group $item1 : $result_grp";} else {print "Group $item1 has been created \n";} } } sub create_user{ (my $item1, my $item2, my $item3, my $item4)=@_; print "Create User\n"; print "===========\n"; my $item1_uname=(getpwnam($item1))[0]; my $item2_uid=(getpwuid($item2))[2]; my $item3_gid=(getgrgid($item3))[2]; if ( $item1_uname eq $item1){ print "ERROR!! User $item1 not created, user already exists...\n"; }elsif ( $item2_uid == $item2){ print "ERROR!! User $item1 not created, user-id validation problem...\n"; }elsif ( $item3_gid != $item3){ print "ERROR!! User $item1 not created, group-id validation problem...\n"; }else{ my $result_usr=`/usr/sbin/useradd -u $item2 -g $item3 -c $item4 -b /home -m $item1 2>&1`; if ( ($result_usr =~ /exists|unique/ )&&( $result_usr !~ /mailbox/) ){print "Cannot add user $item1 : $result_usr";} else{print "User $item1 has been created\n";} } } sub create_perms{ (my $item1, my $item2, my $item3)=@_; print "Create Perms\n"; print "============\n"; my $item1_uname=(getpwnam($item2))[0]; my $item2_gid=(getgrgid($item3))[2]; if ( -d $item1){ if ( ($item1_uname eq $item2) && ($item2_uid == $item3) ){ print "user $item2 and group $item3 exist...\n"; my $result_perm=`/bin/chown -R ${item2}:${$item3} ${item1}`; if ( $result_perm =~/invalid [user|group]/){print "ERROR!! Unable to change permissions\n";} else {print "Permissions have been changed \n";} }else{print "ERROR!! User/Group ($item2:$item3) validation has occured, unable to change permissions\n";} }else{print "ERROR!! Permissions for $item1 unchanged directory doesn't exist..\n";} } sub create_clientlv{ (my $item1, my $item2,my $item3,my $item4,my $item5,my $item6)=@_; my $lv_curr_free=&validate_vg(); if ( $item4 > $lv_curr_free ){ $item4=($lv_curr_free - ($lv_curr_free % 2))/2;} if ( (length($item5)==1) && ($item5=~/\//)){$item5="";} if ( $item6 eq "" ){ print "Creating Logical Volume /dev/${vg_data}/${item3}61${item1}\n"; my $result_lv=`/usr/sbin/lvcreate --size ${item4} --name ${item3}61${item1} $vg_data 2>&1`; if ( $result_lv =~ /create/){ print "Formatting Logical Volume ${item3}61${item1}\n"; my $result_mk=`/sbin/mkfs.ext3 /dev/${vg_data}/${item3}61${item1} 2>&1`; print "Validating Filesystem ${item5}/${item3}61${item1}\n"; if ( -d "${item5}/${item3}61${item1}" ){ print "ERROR!! ${item5}/${item3}61${item1} already exists...\n";} else{ print "Creating Filesystem ${item5}/${item3}61${item1}\n"; my $result_fstab=`/bin/echo -e "/dev/${vg_data}/${item3}61${item1}\t${item5}/${item3}61${item1}\t\text3 \tdefaults\t1 2" >> /etc/fstab`; my $result_dir=`/bin/mkdir -p ${item5}/${item3}61${item1} 2>&1`; if ( $result_dir =~ /cannot create directory/ ){print "ERROR!! $result_dir\n";} else{ my $result_mnt=`/bin/mount ${item5}/${item3}61${item1}`; printf "Exit code %d",$? >> 8;} } }else {print "ERROR!! ${item3}61${item1} already exists...\n";} }elsif( $item6 eq "-" ){ print "Creating Logical Volume /dev/${vg_data}/${item3}\n"; my $result_lv=`/usr/sbin/lvcreate --size ${item4} --name ${item3} $vg_data 2>&1`; if ( $result_lv =~ /create/ ){ print "Formatting Logical Volume ${item3}\n"; my $result_mk=`/sbin/mkfs.ext3 /dev/${vg_data}/${item3} 2>&1`; print "Validating Filesystem ${item5}/${item3}\n"; if ( -d "${item5}/${item3}" ){ print "ERROR!! ${item5}/${item3} already exists...\n"; } else{ print "Creating Filesystem ${item5}/${item3}\n"; my $result_fstab=`/bin/echo -e "/dev/${vg_data}/${item3}\t${item5}/${item3}\t\text3\tdefaults\t1 2" >> /etc/fstab`; my $result_dir=`/bin/mkdir -p ${item5}/${item3} 2>&1`; if ( $result_dir =~ /cannot create directory/ ){print "ERROR!! $result_dir\n";} else{my $result_mnt=`/bin/mount ${item5}/${item3}`;printf "Exit Code %d",$? >>8;} } }else {print "ERROR!! ${item5}/${item3} already exists...\n";} }else{ print "Creating Logical Volume /dev/${vg_data}/${item3}61${item1}${item6}\n"; my $result_lv=`/usr/sbin/lvcreate --size ${item4} --name ${item3}61${item1}_${item6} $vg_data 2>&1`; if ( $result_lv =~ /create/ ){ print "Formatting Logical Volume ${item3}61${item1}${item6}\n"; my $result_mk=`/sbin/mkfs.ext3 /dev/${vg_data}/${item3}61${item1}_${item6} 2>&1`; print "Validating Filesystem ${item5}/${item3}61${item1}/${item6}\n"; if ( -d "${item5}/${item3}61${item1}/${item6}" ){print "ERROR!! ${item5}/${item3}61${item1}/${item6} already exists...\n"} else{ print "Creating Filesystem ${item5}/${item3}61${item1}/${item6}\n"; my $result_fstab=`/bin/echo -e "/dev/${vg_data}/${item3}61${item1}_${item6}\t${item5}/${item3}61${item1 }/${item6}\t\text3\tdefaults\t1 2" >> /etc/fstab`; my $result_dir=`/bin/mkdir -p ${item5}/${item3}61${item1}/${item6}`; if ( $result_dir =~ /cannot create directory/ ){print "ERROR!! $result_dir\n";} else{my $result_mnt=`/bin/mount ${item5}/${item3}61${item1}/${item6}`;printf "Exit Code %d",$? >> 8;} } }else {print "ERROR!! ${item3}61${item1}${item6} already exists...\n"; } } } sub define_clientlv{ (my $item1, my $item2)=@_; print "Create Standard Filesystems\n"; &create_clientlv($item1,$item2,"logs",2048,"/","-"); &create_clientlv($item1,$item2,"scripts",256,"/","-"); &create_clientlv($item1,$item2,"scratch",2048,"/","-"); if ( $item2 eq "IHS" ){ &create_clientlv($item1,$item2,"plg",10240,"/usr/websphere"); &create_perms("/usr/websphere","webadmin","webadms"); }elsif ( $item2 eq "WAS" ){ &create_group("dba",209); &create_user("oracle",1001,40112,"Oracle"); &create_clientlv($item1,$item2,"ndb",10240,"/usr/websphere"); &create_clientlv($item1,$item2,"was",10240,"/usr/websphere","AppServer") ; &create_clientlv($item1,$item2,"u01",5128,"/","-"); &create_perms("/u01","oracle","dba"); }elsif ( $item2 eq "WDM" ){ &create_clientlv($item1,$item2,"ndb",4096); &create_clientlv($item1,$item2,"was",4096,"/","DeploymentManager"); }else{die "No Valid Role Defined : $item2\n";} } (my @bootparam_val1)=&extract_bootparam(); (my $name,my $role)=&validate_bootparam(@bootparam_val1); &define_clientlv($name,$role); -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Chris Lumens Sent: 27 August 2010 15:52 To: kickstart-list at redhat.com Subject: Re: harddrive in %pre > I am writing a script to install from a USB key, however it appears as > a different device on several different systems, so I put the harddisk > command into an include, generated by the %pre area, however the > system seems to be ignoring this - is there a way to make it work, or > are there commands which just can't be put into includes? Any of the method selection commands (harddrive/nfs/url/cdrom) are the only ones that cannot be put in an include. There are ways to do what you're trying to do, but I'll need a little more context to make a suggestion. Can you post your kickstart file, minus any sensitive information? - Chris _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list ----------------------------------------------------------------------------------------------------- Lloyds TSB Bank plc. Registered Office: 25 Gresham Street, London EC2V 7HN. Registered in England and Wales, number 2065. Telephone: 020 7626 1500. Bank of Scotland plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 327000. Telephone: 0870 600 5000 Lloyds TSB Scotland plc. Registered Office: Henry Duncan House, 120 George Street, Edinburgh EH2 4LH. Registered in Scotland, number 95237. Telephone: 0131 225 4555. Cheltenham & Gloucester plc. Registered Office: Barnett Way, Gloucester GL4 3RL. Registered in England and Wales, number 2299428. Telephone: 01452 372372. Lloyds TSB Bank plc, Lloyds TSB Scotland plc, Bank of Scotland plc and Cheltenham & Gloucester plc are authorised and regulated by the Financial Services Authority. Halifax is a division of Bank of Scotland plc. Cheltenham & Gloucester Savings is a division of Lloyds TSB Bank plc. HBOS plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 218813. Telephone: 0870 600 5000 Lloyds Banking Group plc. Registered Office: The Mound, Edinburgh EH1 1YZ. Registered in Scotland, number 95000. Telephone: 0131 225 4555 This e-mail (including any attachments) is private and confidential and may contain privileged material. If you have received this e-mail in error, please notify the sender and delete it (including any attachments) immediately. You must not copy, distribute, disclose or use any of the information in it or any attachments. Telephone calls may be monitored or recorded. ============================================================================== From streamscalelance at gmail.com Fri Aug 27 15:43:05 2010 From: streamscalelance at gmail.com (Lance Gropper) Date: Fri, 27 Aug 2010 08:43:05 -0700 Subject: harddrive in %pre In-Reply-To: References: Message-ID: Hello Ya'all: Thank you all for the kind responses so far. Below is my kickstart (so far) - it is intended to be an automated install. Basically, the %pre finds the USB key, and sets it as the "install from" drive, and based on this sets the "install to" drive. The "install to" drive and network configuration currently works, however the "install from" harddrive statement does not, even though the USB key is properly detected. It would be great if the USB device was the same Linux short device name on every system, but it's not. Today, I am going to modify the script so it attempts other devices, other than hda, sda, and sdb. One of our customers' systems going to IBC the week after next has 73 drives, and the USB key shows up last. I'd like to keep the kickstart as light and fluffy as possible... Lance # Kickstart file %include /tmp/inst-include install key --skip lang en_US.UTF-8 keyboard us skipx rootpw --iscrypted $1$kt/7H6d3$g96NCQCYIJL.bIF7/FEJh/ firewall --disabled authconfig --useshadow --enablemd5 selinux --disabled timezone America/Tijuana %packages @core %pre #!/bin/bash INSTALLFROM='sda5' INSTALLTO='hda' touch /tmp/inst-include if test -e '/sys/block/hda' ; then ls -l /sys/block/hda/device | grep /usb if [ $? -neq 0 ] ; then INSTALLFROM='hda5' INSTALLTO='sda' fi fi if test -e '/sys/block/sda' ; then ls -l /sys/block/sda/device | grep /usb if [ $? -neq 0 ] ; then INSTALLFROM='sda5' if test -e '/sys/block/hda' ; then INSTALLTO='hda' else INSTALLTO='sdb' fi fi fi if test -e '/sys/block/sdb' ; then ls -l /sys/block/sdb/device | grep /usb if [ $? -neq 0 ] ; then INSTALLFROM='sdb5' if test -e '/sys/block/hda' ; then INSTALLTO='hda' else INSTALLTO='sda' fi fi fi NICPORTS=0 if test -e '/sys/class/net/eth0' ; then NICPORTS=1 fi if test -e '/sys/class/net/eth1' ; then NICPORTS=2 fi echo 'harddrive --partition='$INSTALLFROM' --dir=/' >>/tmp/inst-include if [ $NICPORTS -gt 0 ] ; then if [ $NICPORTS -eq 1 ] ; then echo 'network --device eth0 --bootproto static --ip 192.168.1.129 --netmask 255.255.255.0 --gateway 192.168.1.1 --nameserver None --hostname mybox' >>/tmp/inst-include else echo 'network --device eth0 --bootproto dhcp --hostname mybox' >>/tmp/inst-include echo 'network --device eth1 --bootproto static --ip 192.168.1.129 --netmask 255.255.255.0 --gateway 192.168.1.1 --nameserver None --hostname mybox' >>/tmp/inst-include fi fi echo 'bootloader --location=mbr --driveorder='$INSTALLTO' --append="rhgb quiet crashkernel=128M at 16M"' >>/tmp/inst-include echo 'clearpart --all --drives='$INSTALLTO >>/tmp/inst-include echo 'part /boot --fstype ext3 --size=100 --ondisk='$INSTALLTO >>/tmp/inst-include echo 'part pv.100000 --size=0 --grow --ondisk='$INSTALLTO >>/tmp/inst-include echo 'volgroup VolGroup00 --pesize=32768 pv.100000' >>/tmp/inst-include echo 'logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=384 --grow --maxsize=768' >>/tmp/inst-include echo 'logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow' >>/tmp/inst-include %end -------------- next part -------------- An HTML attachment was scrubbed... URL: From dansmood at gmail.com Fri Aug 27 15:50:08 2010 From: dansmood at gmail.com (Dan Davis) Date: Fri, 27 Aug 2010 11:50:08 -0400 Subject: harddrive in %pre In-Reply-To: <201008271511.o7RFBiAe020028@mx1.redhat.com> References: <201008271511.o7RFBiAe020028@mx1.redhat.com> Message-ID: What we do is set a label with mkfs.vfat and change the syslinux.cfg on the USB to make reference to the USB label. In %pre and %post scripts, you can use udevadm info on the devices and figure out which is USB, but that only works if your "real" hard disks are not USB. -------------- next part -------------- An HTML attachment was scrubbed... URL: From dansmood at gmail.com Fri Aug 27 15:57:37 2010 From: dansmood at gmail.com (Dan Davis) Date: Fri, 27 Aug 2010 11:57:37 -0400 Subject: pungi question Message-ID: Is there anyway I can instruct pungi to use one set of repositories to build the isolinux sub-directory, and another set of repositories for the Packages to install? I've wanted to be able to do that for a long time, just so that nothing important to the DVD/CD changes, but we can change what we install (including the kernel). I got hope the other day of this because I did a pungi build that turned out odd, in that these files didn't match: x86_64/os/Packages/kernel* x86_64/os/vmlinuz All I did was list fedora-updates with a repository cost of 1000, trying out some options for listing my internal mirrors. Is this a Feature or a Bug? I'd love to use it... but I'm not sure why the build was different. -------------- next part -------------- An HTML attachment was scrubbed... URL: From streamscalelance at gmail.com Fri Aug 27 16:09:41 2010 From: streamscalelance at gmail.com (Lance Gropper) Date: Fri, 27 Aug 2010 09:09:41 -0700 Subject: Kickstart-list Digest, Vol 78, Issue 10 In-Reply-To: References: Message-ID: Hello Dan: My script already figures out the device - the problem is using that device in the harddrive statement, by putting the harddrive statement in the %pre isn't working... Lance On Fri, Aug 27, 2010 at 9:00 AM, wrote: > Send Kickstart-list mailing list submissions to > kickstart-list at redhat.com > > To subscribe or unsubscribe via the World Wide Web, visit > https://www.redhat.com/mailman/listinfo/kickstart-list > or, via email, send a message with subject or body 'help' to > kickstart-list-request at redhat.com > > You can reach the person managing the list at > kickstart-list-owner at redhat.com > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Kickstart-list digest..." > > Today's Topics: > > 1. Re: harddrive in %pre (Lance Gropper) > 2. Re: harddrive in %pre (Dan Davis) > 3. pungi question (Dan Davis) > > > ---------- Forwarded message ---------- > From: Lance Gropper > To: kickstart-list at redhat.com > Date: Fri, 27 Aug 2010 08:43:05 -0700 > Subject: Re: harddrive in %pre > Hello Ya'all: > > Thank you all for the kind responses so far. Below is my kickstart (so far) > - it is intended to be an automated install. Basically, the %pre finds the > USB key, and sets it as the "install from" drive, and based on this sets the > "install to" drive. The "install to" drive and network configuration > currently works, however the "install from" harddrive statement does not, > even though the USB key is properly detected. It would be great if the USB > device was the same Linux short device name on every system, but it's not. > Today, I am going to modify the script so it attempts other devices, other > than hda, sda, and sdb. One of our customers' systems going to IBC the week > after next has 73 drives, and the USB key shows up last. I'd like to keep > the kickstart as light and fluffy as possible... > > Lance > > # Kickstart file > > %include /tmp/inst-include > install > key --skip > lang en_US.UTF-8 > keyboard us > skipx > > rootpw --iscrypted $1$kt/7H6d3$g96NCQCYIJL.bIF7/FEJh/ > firewall --disabled > authconfig --useshadow --enablemd5 > selinux --disabled > timezone America/Tijuana > > %packages > @core > > %pre > #!/bin/bash > > INSTALLFROM='sda5' > INSTALLTO='hda' > > touch /tmp/inst-include > if test -e '/sys/block/hda' ; then > ls -l /sys/block/hda/device | grep /usb > if [ $? -neq 0 ] ; then > INSTALLFROM='hda5' > INSTALLTO='sda' > fi > fi > if test -e '/sys/block/sda' ; then > ls -l /sys/block/sda/device | grep /usb > if [ $? -neq 0 ] ; then > INSTALLFROM='sda5' > if test -e '/sys/block/hda' ; then > INSTALLTO='hda' > else > INSTALLTO='sdb' > fi > fi > fi > if test -e '/sys/block/sdb' ; then > ls -l /sys/block/sdb/device | grep /usb > if [ $? -neq 0 ] ; then > INSTALLFROM='sdb5' > if test -e '/sys/block/hda' ; then > INSTALLTO='hda' > else > INSTALLTO='sda' > fi > fi > fi > NICPORTS=0 > if test -e '/sys/class/net/eth0' ; then > NICPORTS=1 > fi > if test -e '/sys/class/net/eth1' ; then > NICPORTS=2 > fi > > echo 'harddrive --partition='$INSTALLFROM' --dir=/' >>/tmp/inst-include > > if [ $NICPORTS -gt 0 ] ; then > if [ $NICPORTS -eq 1 ] ; then > echo 'network --device eth0 --bootproto static --ip 192.168.1.129 > --netmask 255.255.255.0 --gateway 192.168.1.1 --nameserver None --hostname > mybox' >>/tmp/inst-include > else > echo 'network --device eth0 --bootproto dhcp --hostname mybox' > >>/tmp/inst-include > echo 'network --device eth1 --bootproto static --ip 192.168.1.129 > --netmask 255.255.255.0 --gateway 192.168.1.1 --nameserver None --hostname > mybox' >>/tmp/inst-include > fi > fi > > echo 'bootloader --location=mbr --driveorder='$INSTALLTO' --append="rhgb > quiet crashkernel=128M at 16M"' >>/tmp/inst-include > echo 'clearpart --all --drives='$INSTALLTO >>/tmp/inst-include > echo 'part /boot --fstype ext3 --size=100 --ondisk='$INSTALLTO > >>/tmp/inst-include > echo 'part pv.100000 --size=0 --grow --ondisk='$INSTALLTO > >>/tmp/inst-include > echo 'volgroup VolGroup00 --pesize=32768 pv.100000' >>/tmp/inst-include > echo 'logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 > --size=384 --grow --maxsize=768' >>/tmp/inst-include > echo 'logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 > --size=1024 --grow' >>/tmp/inst-include > %end > > > > ---------- Forwarded message ---------- > From: Dan Davis > To: Discussion list about Kickstart > Date: Fri, 27 Aug 2010 11:50:08 -0400 > Subject: Re: harddrive in %pre > What we do is set a label with mkfs.vfat and change the syslinux.cfg on the > USB to make reference to the USB label. > > In %pre and %post scripts, you can use udevadm info on the devices and > figure out which is USB, but that only works if your "real" hard disks are > not USB. > > > ---------- Forwarded message ---------- > From: Dan Davis > To: kickstart-list at redhat.com > Date: Fri, 27 Aug 2010 11:57:37 -0400 > Subject: pungi question > Is there anyway I can instruct pungi to use one set of repositories to > build the isolinux sub-directory, and another set of repositories for the > Packages to install? > > I've wanted to be able to do that for a long time, just so that nothing > important to the DVD/CD changes, but we can change what we install > (including the kernel). > > I got hope the other day of this because I did a pungi build that turned > out odd, in that these files didn't match: > > x86_64/os/Packages/kernel* > x86_64/os/vmlinuz > > All I did was list fedora-updates with a repository cost of 1000, trying > out some options for listing my internal mirrors. > > Is this a Feature or a Bug? I'd love to use it... but I'm not sure why the > build was different. > > _______________________________________________ > 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 bjs at redhat.com Fri Aug 27 16:56:06 2010 From: bjs at redhat.com (Bryan J Smith) Date: Fri, 27 Aug 2010 12:56:06 -0400 Subject: harddrive in %pre In-Reply-To: References: <201008271511.o7RFBiAe020028@mx1.redhat.com> Message-ID: <1282928166.9672.74.camel@localhost.localdomain> On Fri, 2010-08-27 at 11:50 -0400, Dan Davis wrote: > What we do is set a label with mkfs.vfat and change the syslinux.cfg > on the USB to make reference to the USB label. > In %pre and %post scripts, you can use udevadm info on the devices and > figure out which is USB, but that only works if your "real" hard disks > are not USB. But one can capture unique IDs of the USB device. It would require a slight, extra-effort on each USB device, but it would be worth it. -- Bryan J Smith Senior Consultant Red Hat, Inc Professional Consulting http://www.redhat.com/consulting mailto:bjs at redhat.com +1 (407) 489-7013 (Mobile) mailto:b.j.smith at ieee.org (Blackberry/Red Hat-External) -------------------------------------------------------- You already know Red Hat as the entity dedicated to 100% no-IP-strings-attached, community software development. But do you know where CIOs rate Red Hat versus other software and services firms for their own, direct needs, year after year? http://www.redhat.com/promo/vendor/ From dansmood at gmail.com Fri Aug 27 19:10:57 2010 From: dansmood at gmail.com (Dan Davis) Date: Fri, 27 Aug 2010 15:10:57 -0400 Subject: how best to resolve dependencies Message-ID: I use this pure rpm command-line way to check for full package dependency resolution before burning a DVD: mkdir /tmp/rpmdb rpm --initdb --dbpath /tmp/rpmdb pushd trunk-f13/x86_64/os/Packages rpm --dbpath /tmp/rpmdb --test -Uvh *.rpm ... I do the pungi build install steps separately from split media, and after the build and install: pungi -c pungi-f13.cfg -G -B -C --name=IE --ver=trunk-f13 --nosource --force I see that I have some duplicative packages: [root at FC13BUILD installer]# ls trunk-f13/x86_64/os/Packages/gcc-* trunk-f13/x86_64/os/Packages/gcc-4.4.4-10.fc13.x86_64.rpm trunk-f13/x86_64/os/Packages/gcc-4.4.4-2.fc13.x86_64.rpm trunk-f13/x86_64/os/Packages/gcc-c++-4.4.4-10.fc13.x86_64.rpm trunk-f13/x86_64/os/Packages/gcc-gfortran-4.4.4-10.fc13.x86_64.rpm trunk-f13/x86_64/os/Packages/gcc-gfortran-4.4.4-2.fc13.i686.rpm Is there anyway to prevent that? Is there a better way to check that all dependencies are resolved based on the manifest? -------------- next part -------------- An HTML attachment was scrubbed... URL: From dansmood at gmail.com Fri Aug 27 19:27:15 2010 From: dansmood at gmail.com (Dan Davis) Date: Fri, 27 Aug 2010 15:27:15 -0400 Subject: Kickstart-list Digest, Vol 78, Issue 10 In-Reply-To: References: Message-ID: On Fri, Aug 27, 2010 at 12:09 PM, Lance Gropper wrote: > Hello Dan: > > My script already figures out the device - the problem is using that device > in the harddrive statement, by putting the harddrive statement in the %pre > isn't working... > > Lance > We use the same temporary file mechanism that you do, where the kickstart file includes a temporary file written by %pre, but we don't use the harddrive command to tell anaconda where to find the media. I guess that anaconda resolves where to get too early for the temporary file mechanism to work. What I was saying before should work ... but more detail is needed. We modify the ks= command to make it dynamically depend on the name of the USB volume. When we build our DVD, we have an isolinux.cfg that includes: kernel vmlinuz append selinux=0 initrd=initrd.img ks=cdrom:/upgrade.cfg Once we have the ISO, we make our USB disk bootable with parted, and then build a labeled vfat partition: mkfs.vfat -n MINE /dev/sde1 After we've transferred the ISO onto a USB with livecd-iso-to-disk, we modify syslinux.cfg to change the ks= line: kernel vmlinuz append initrd=initrd.img repo=hd:UUID=6D6C-9D8A:/ ks=hd:LABEL=MINE:/upgrade.cfg We never have to tell anaconda where to find the media. I hadn't noticed the UUID in there before, it must be exactly what is suggested by Bryan. I suspect that using the UUID= would work just as well as the LABEL=, and vice versa. Hopefully, this helps, but I inherited this build from someone else ... I just got the USB install to work yesterday. -------------- next part -------------- An HTML attachment was scrubbed... URL: From kanarip at kanarip.com Fri Aug 27 22:19:52 2010 From: kanarip at kanarip.com (Jeroen van Meeuwen) Date: Fri, 27 Aug 2010 23:19:52 +0100 Subject: need to know what repositories are needed for revisor in fc 13 In-Reply-To: References: Message-ID: <201008272319.52797.kanarip@kanarip.com> Floydsmith at aol.com wrote: > I am trying to run revisor with the following cmdline: > [u at floydsmi ~]$ sudo revisor --cli --model=f13-i386 > --kickstart=/dosc/anaconda-ks.cfg --install-dvd > > And I get an error message saying it can't communicate with the > repositories. > Hi Floyd, I'm travelling for a while now, and my anaconda repositories back home are on a system that went belly up. Please disable the anaconda repository in /etc/revisor/conf.d/revisor-f13- i386.conf, and try again. -- Jeroen From kanarip at kanarip.com Fri Aug 27 22:43:48 2010 From: kanarip at kanarip.com (Jeroen van Meeuwen) Date: Fri, 27 Aug 2010 23:43:48 +0100 Subject: pungi question In-Reply-To: References: Message-ID: <201008272343.49575.kanarip@kanarip.com> Dan Davis wrote: > Is there anyway I can instruct pungi to use one set of repositories to build > the isolinux sub-directory, and another set of repositories for the Packages > to install? > I suppose what you want comes close to re-using the original installer images, and so only modify the payload of the media and not rebuilt install.img / isolinux/*? -- Jeroen