From chris.ellam at dfs.de Wed Sep 1 11:09:00 2010 From: chris.ellam at dfs.de (chris.ellam at dfs.de) Date: Wed, 1 Sep 2010 13:09:00 +0200 Subject: Kickstart Installation usage of key --skip when installing Workstation Packages. Message-ID: Hi All, I am having problems with the Kickstart installaion process with RH Workstation 5.3. The dvd or its NFS mounted iso image has both the Client and Workstation packages and I want to install some packages from the Workstation repository. This is in principle Ok when an installation key is provided, but when the installation key is skipped the workstation package repository is ignored and missing packages (abort or continue) questions arise, which means that manual intervention is required. I could specify the ignore missing packages, but then the installed system does not have the configuration that is required. What I need is a solution to install a new/replacement workstation without inititially having a key (which gets provided later). When an installation key is provided, i.e nfs --server=10.200.7.14 --dir=/os/RHWS53 key 5d44a61f5d08xxxx in /var/log/anaconda.log I see that it picks up both the Workstation and the client repositiories 12:11:29 INFO : moving (1) to step regkey 12:11:29 INFO : Adding Workstation repo 12:11:29 INFO : Adding Base repo 12:11:29 INFO : repopaths is {'workstation': 'Workstation', 'base': 'Client'} 12:11:30 INFO : moving (1) to step reposetup 12:11:30 INFO : package paths is {'workstation': ['file:///mnt/source/./Workstation'], 'base': ['file:///mnt/source/./Client']} However, when the installation key is skipped (which is a valid option), i.e. nfs --server=10.200.7.14 --dir=/os/RHWS53 key --skip in /var/log/anaconda.log I see that it picks only the client repository 14:37:54 INFO : skipping install key 14:37:55 INFO : package paths is {'base': ['file:///mnt/source/./Client']} 14:37:57 DEBUG : Checking for virtual provide or file-provide for imake 14:37:57 DEBUG : no package matching imake 14:40:31 DEBUG : Checking for virtual provide or file-provide for arpwatch 14:40:31 DEBUG : no package matching arpwatch To overcome this I tried to add the Workstation repository to kickstart nfs --server=10.200.7.14 --dir=/os/RHWS53 key --skip repo --name=Workstation --baseurl=nfs://10.200.7.14:/os/RHWS53/Workstation or repo --name=Workstation --baseurl=nfs://10.200.7.14:/os/RHWS53/Workstation/repodata or repo --name=Workstation --baseurl=file://mnt/source/./Workstation in the anaconda log I see that it didn't except this 14:37:54 INFO : skipping install key 14:37:55 INFO : moving (1) to step reposetup 14:37:55 INFO : package paths is {'base': ['file:///mnt/source/./Client']} 14:37:55 INFO : added repository Workstation with source URL file://mnt/source/./Workstation 14:37:55 WARNING : /usr/lib64/python2.4/site-packages/snack.py:250: DeprecationWarning: integer argument expected, got float self.w = _snack.scale(width, total) 14:37:55 WARNING : /usr/lib64/python2.4/site-packages/snack.py:247: DeprecationWarning: integer argument expected, got float self.w.scaleSet(amount) 14:37:55 DEBUG : repo time: 0.001 14:37:55 DEBUG : Setting up Package Sacks 14:37:56 DEBUG : pkgsack time: 0.836 14:37:56 DEBUG : Reading Local RPMDB 14:37:56 DEBUG : rpmdb time: 0.000 14:37:56 DEBUG : repo time: 0.001 14:37:56 DEBUG : Setting up Package Sacks 14:37:56 ERROR : reading package metadata: Cannot retrieve repository metadata (repomd.xml) for repository: Workstation. Please verify its path and try again 14:37:56 ERROR : disabling non-base repo Workstation: Cannot retrieve reposito ry metadata (repomd.xml) for repository: Workstation. Please verify its path and try again 14:37:56 DEBUG : Getting group metadata 14:37:57 DEBUG : Checking for virtual provide or file-provide for imake 14:37:57 DEBUG : no package matching imake 14:40:31 DEBUG : Checking for virtual provide or file-provide for arpwatch 14:40:31 DEBUG : no package matching arpwatch The later appears to be a known problem and as yet I have not found a solution to this either. Thanks, Chris -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- DFS Deutsche Flugsicherung GmbH Am DFS-Campus 10 D - 63225 Langen Tel.: +49-(0)6103-707-0 Sitz der Gesellschaft: Langen/Hessen Zust?ndiges Registergericht: AG Offenbach am Main, HRB 34977 Vorsitzender des Aufsichtsrates: Prof. Klaus-Dieter Scheurle Gesch?ftsf?hrer: Dieter Kaden (Vors.), Ralph Riedle, Jens Bergmann Internet: http://www.dfs.de Public-Key der DFS: http://www.dfs.de/dfs/public_key.asc From Floydsmith at aol.com Thu Sep 2 09:14:00 2010 From: Floydsmith at aol.com (Floydsmith at aol.com) Date: Thu, 2 Sep 2010 05:14:00 EDT Subject: revisor fc 13 isobuild error after removing anaconda updates from repository Message-ID: <16a5d1.7426bf4f.39b0c4d8@aol.com> In order to resolve issue "need to know what repositories are needed for revisor in fc 13" in accordance with _kanarip at kanarip.com_ (mailto:kanarip at kanarip.com) instructions I have disabled anaconda repositorys in the /etc/revisor/conf.d/revisor-f13-i386.conf file and I now no longer get the: Fatal Error: Unable to retrieve software information. This could be caused by one of the following: - not having a network connection available, - Server refusing connections, - Using a mirror that isn't fully synchronized, - Misconfigured repositories. But, I now get (near the end of a fc 13 revisor run) the following error: "Build isolinux and installer": Got an error from /usr/lib/anaconda-runtime/buildinstall (return code 1). Linking in Release - never starts (just hangs saying pending). Could the isolinux build error be a result of not being able to get any anaconda updates and if so, when will that server which crashed be back up or is there another mirror for anaconda repository? Floyd, -------------- next part -------------- An HTML attachment was scrubbed... URL: From loudredz71 at yahoo.com Thu Sep 2 15:09:34 2010 From: loudredz71 at yahoo.com (Jonathan Horne) Date: Thu, 2 Sep 2010 08:09:34 -0700 (PDT) Subject: editing log files at post Message-ID: <152404.40397.qm@web45303.mail.sp1.yahoo.com> unfortunately i dont know anything about awk and sed (which is pitiful considering my other experience), but i think this is likely the answer to my issues. problem: servers are not compressing log files in /var/log, despite the fact that 'compress' exists in the logrotate.conf. at the end of my kickstart .cfg file, i have a line (its actually the final line) echo "compress" >> /etc/logrotate.conf EOF which causes my /etc/logrotate.conf to look like this: [root at psl-sipas04 log]# cat /etc/logrotate.conf # see "man logrotate" for details # rotate log files weekly weekly # keep 4 weeks worth of backlogs rotate 4 # create new (empty) log files after rotating old ones create # uncomment this if you want your log files compressed #compress # RPM packages drop log rotation information into this directory include /etc/logrotate.d # no packages own wtmp -- we'll rotate them here /var/log/wtmp { monthly minsize 1M create 0664 root utmp rotate 1 } # system-specific logs may be also be configured here. compress [root at psl-sipas04 log]# as you can see, the 'compress' is the last line of the .conf file. but despite the fact its there, the server(s) is not compressing the log riles upon rotation. ive been under the impression that the exitance of a 'compress' in the uncommented state would cause the behavior i want, but this appears to not be so (not sure why here, other research required). im wondering if i use something other than echoing the compress to the end of the file, or if i use awk or sed to locate the... #compress and change it to... compress if that would help? would anyone be able to help me with an example that would do this at the post of my kickstart, instead of just appending the 'compress' at the bottom/end of file? -------------- next part -------------- An HTML attachment was scrubbed... URL: From hbrown at divms.uiowa.edu Thu Sep 2 15:47:25 2010 From: hbrown at divms.uiowa.edu (Hugh Brown) Date: Thu, 02 Sep 2010 10:47:25 -0500 Subject: editing log files at post In-Reply-To: <152404.40397.qm@web45303.mail.sp1.yahoo.com> References: <152404.40397.qm@web45303.mail.sp1.yahoo.com> Message-ID: <4C7FC70D.6030609@divms.uiowa.edu> Jonathan Horne wrote: > unfortunately i dont know anything about awk and sed (which is pitiful > considering my other experience), but i think this is likely the answer to my > issues. > > problem: servers are not compressing log files in /var/log, despite the fact > that 'compress' exists in the logrotate.conf. > > > at the end of my kickstart .cfg file, i have a line (its actually the final > line) > > echo "compress" >> /etc/logrotate.conf > EOF > > which causes my /etc/logrotate.conf to look like this: > > [root at psl-sipas04 log]# cat /etc/logrotate.conf > # see "man logrotate" for details > # rotate log files weekly > weekly > > # keep 4 weeks worth of backlogs > rotate 4 > > # create new (empty) log files after rotating old ones > create > > # uncomment this if you want your log files compressed > #compress > > # RPM packages drop log rotation information into this directory > include /etc/logrotate.d > > # no packages own wtmp -- we'll rotate them here > /var/log/wtmp { > monthly > minsize 1M > create 0664 root utmp > rotate 1 > } > > # system-specific logs may be also be configured here. > compress > [root at psl-sipas04 log]# > > > as you can see, the 'compress' is the last line of the .conf file. but despite > the fact its there, the server(s) is not compressing the log riles upon > rotation. ive been under the impression that the exitance of a 'compress' in > the uncommented state would cause the behavior i want, but this appears to not > be so (not sure why here, other research required). > > im wondering if i use something other than echoing the compress to the end of > the file, or if i use awk or sed to locate the... > > #compress > > and change it to... > > compress > > if that would help? > > would anyone be able to help me with an example that would do this at the post > of my kickstart, instead of just appending the 'compress' at the bottom/end of > file? > > According to the man page for logrotate: Each configuration file can set global options (local definitions override global ones, and later definitions override earlier ones) So, your compress is set globally and is probably being overridden in one of the /etc/logrotate.d/* files. You'll need to verify that compress is set appropriately there for the log files you care about. Generally, I do cp /etc/logrotate.conf /etc/logrotate.conf.dist sed -e 's/^#compress/compress/' /etc/logrotate.conf.dist >/etc/logrotate.conf chmod 644 /etc/logrotate.conf the sed command should be all on one line in your script. Be warned that if you leave *.dist files in /etc/logrotate.d/ they will get processed as well. I generally copy them to /root/ before doing the sed command. Hugh From email at jasonkohles.com Thu Sep 2 16:33:34 2010 From: email at jasonkohles.com (Jason Kohles) Date: Thu, 2 Sep 2010 12:33:34 -0400 Subject: editing log files at post In-Reply-To: <152404.40397.qm@web45303.mail.sp1.yahoo.com> References: <152404.40397.qm@web45303.mail.sp1.yahoo.com> Message-ID: <264CFD81-0B9A-4988-A6AC-C42DA5B15ADF@jasonkohles.com> Yes, order is important, as the options will only affect logfiles that come after them, and since most of your logfiles are defined in logrotate.d, the compress option comes too late. To uncomment the existing compress option with sed, just use: sed -i 's/^#compress/compress/' /etc/logrotate.conf An easier (and more reliable) technique, however, is to add your option overrides to a file that gets included before any of the other included files, like so: echo "compress" >> /etc/logrotate.d/000-local-overrides That way you don't have to worry about figuring out whether the commented out option is in the config file or what to do if someone has added a space after the comment character. On Sep 2, 2010, at 11:09 AM, Jonathan Horne wrote: > unfortunately i dont know anything about awk and sed (which is pitiful considering my other experience), but i think this is likely the answer to my issues. > > problem: servers are not compressing log files in /var/log, despite the fact that 'compress' exists in the logrotate.conf. > > at the end of my kickstart .cfg file, i have a line (its actually the final line) > > echo "compress" >> /etc/logrotate.conf > EOF > > which causes my /etc/logrotate.conf to look like this: > > [root at psl-sipas04 log]# cat /etc/logrotate.conf > # see "man logrotate" for details > # rotate log files weekly > weekly > > # keep 4 weeks worth of backlogs > rotate 4 > > # create new (empty) log files after rotating old ones > create > > # uncomment this if you want your log files compressed > #compress > > # RPM packages drop log rotation information into this directory > include /etc/logrotate.d > > # no packages own wtmp -- we'll rotate them here > /var/log/wtmp { > monthly > minsize 1M > create 0664 root utmp > rotate 1 > } > > # system-specific logs may be also be configured here. > compress > [root at psl-sipas04 log]# > > > as you can see, the 'compress' is the last line of the .conf file. but despite the fact its there, the server(s) is not compressing the log riles upon rotation. ive been under the impression that the exitance of a 'compress' in the uncommented state would cause the behavior i want, but this appears to not be so (not sure why here, other research required). > > im wondering if i use something other than echoing the compress to the end of the file, or if i use awk or sed to locate the... > > #compress > > and change it to... > > compress > > if that would help? > > would anyone be able to help me with an example that would do this at the post of my kickstart, instead of just appending the 'compress' at the bottom/end of file? > > _______________________________________________ > 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 kickstart at hastek.net Thu Sep 2 17:26:04 2010 From: kickstart at hastek.net (kickstart) Date: Thu, 2 Sep 2010 13:26:04 -0400 Subject: editing log files at post In-Reply-To: <152404.40397.qm@web45303.mail.sp1.yahoo.com> References: <152404.40397.qm@web45303.mail.sp1.yahoo.com> Message-ID: <20100902172603.GA2622@florian> Jonathan Horne: > unfortunately i dont know anything about awk and sed (which is pitiful > ... > would anyone be able to help me with an example that would do this at > the post of my kickstart, instead of just appending the 'compress' at > the bottom/end of file? Jonathan, At "post" time, perl should be available, you could try a one-liner. Have a look at /etc/logrotate.d/syslog -- if it's the files mentioned in there, it might be an option to just put your "compress" directive in there. As an example, with perl you could: perl -pi -e 's#endscript#endscript\n\tcompress#;' /etc/logrotate.d/syslog or something similar, which might do the trick. Hope it helps! -Harold --Harold From kanarip at kanarip.com Fri Sep 3 13:47:57 2010 From: kanarip at kanarip.com (Jeroen van Meeuwen) Date: Fri, 3 Sep 2010 15:47:57 +0200 Subject: revisor fc 13 isobuild error after removing anaconda updates from repository In-Reply-To: <16a5d1.7426bf4f.39b0c4d8@aol.com> References: <16a5d1.7426bf4f.39b0c4d8@aol.com> Message-ID: <201009031547.58290.kanarip@kanarip.com> Floydsmith at aol.com wrote: > In order to resolve issue "need to know what repositories are needed for > revisor in fc 13" > in accordance with _kanarip at kanarip.com_ (mailto:kanarip at kanarip.com) > instructions I have disabled anaconda repositorys > in the /etc/revisor/conf.d/revisor-f13-i386.conf file and I now no longer > get the: > Fatal Error: Unable to retrieve software information. > This could be caused by one of the following: > - not having a network connection available, > - Server refusing connections, > - Using a mirror that isn't fully synchronized, > - Misconfigured repositories. > But, I now get (near the end of a fc 13 revisor run) the following error: > "Build isolinux and installer": > Got an error from /usr/lib/anaconda-runtime/buildinstall (return code 1). > Linking in Release - never starts (just hangs saying pending). > Could the isolinux build error be a result of not being able to get any > anaconda updates > and if so, when will that server which crashed be back up or is there > another mirror for > anaconda repository? > Could you run this again with --debug 9 and send me the output off-list? -- Jeroen From bishop at platypus.bc.ca Sun Sep 5 05:13:05 2010 From: bishop at platypus.bc.ca (bishop) Date: Sat, 04 Sep 2010 22:13:05 -0700 Subject: editing log files at post In-Reply-To: <264CFD81-0B9A-4988-A6AC-C42DA5B15ADF@jasonkohles.com> References: <152404.40397.qm@web45303.mail.sp1.yahoo.com> <264CFD81-0B9A-4988-A6AC-C42DA5B15ADF@jasonkohles.com> Message-ID: <4C8326E1.6000807@platypus.bc.ca> At the list of adding more trash to the fire.. Jason's comment below, I think is excellent -- that while sed-i will edit the file effectively, adding the snippet into the .d tree should accomplish the same thing. With an added bonus: That second option allows the .conf file to remain unchanged, which has value in updates where the package being replaced has a %config(noreplace) and a bug fix which *addresses* the config. I run into people on both ends of the packaging process who haven't grokked that yet, so I thought to draw attention. - bish Jason Kohles wrote: > Yes, order is important, as the options will only affect logfiles that > come after them, and since most of your logfiles are defined in > logrotate.d, the compress option comes too late. > > To uncomment the existing compress option with sed, just use: > > sed -i 's/^#compress/compress/' /etc/logrotate.conf > > An easier (and more reliable) technique, however, is to add your option > overrides to a file that gets included before any of the other included > files, like so: > > echo "compress" >> /etc/logrotate.d/000-local-overrides > > That way you don't have to worry about figuring out whether the > commented out option is in the config file or what to do if someone has > added a space after the comment character. From ghaase at syntheticgenomics.com Tue Sep 7 20:07:51 2010 From: ghaase at syntheticgenomics.com (Greg Haase) Date: Tue, 07 Sep 2010 13:07:51 -0700 Subject: Unexpected results with Kickstart Message-ID: I have a kickstart setup that I scratched together. I created a boot disk with a reference to my kickstart file served up on a web server. The lines in the isolinux.cfg look like this: label linux menu label ^Install a new system or upgrade an existing system kernel vmlinuz append initrd=initrd.img stage2=hd:LABEL="Fedora" ks=http:///fedora/ks.cfg It all works fine on the web server I have all of my versions of kickstart files and I simply move my link to the appropriate kickstart file that I want to use and run the install. Yet in my kickstart file I specify only the core package like: %packages @Core %end Yet after the install I look at the anaconda-ks.cfg on the system created with the kickstart and it has: %packages @Base @Core @core %end Or some variation with the caps / lowercase. The thing is that I want as simple an install as possible. Is there something I am missing? Thanks Greg -------------- next part -------------- An HTML attachment was scrubbed... URL: From clumens at redhat.com Tue Sep 7 20:34:23 2010 From: clumens at redhat.com (Chris Lumens) Date: Tue, 7 Sep 2010 16:34:23 -0400 Subject: Unexpected results with Kickstart In-Reply-To: References: Message-ID: <20100907203422.GK30596@exeter.usersys.redhat.com> > Yet after the install I look at the anaconda-ks.cfg on the system created > with the kickstart and it has: > > %packages > @Base > @Core > @core > %end > > Or some variation with the caps / lowercase. The thing is that I want as > simple an install as possible. Is there something I am missing? The Base group is always selected unless you use "%packages --nobase" (see https://fedoraproject.org/wiki/Anaconda/Kickstart#Chapter_3._Package_Selection). The case-sensitive thing there is a little strange, but nothing to be concerned about. It should not affect the number of packages installed. - Chris From ghaase at syntheticgenomics.com Tue Sep 7 21:46:25 2010 From: ghaase at syntheticgenomics.com (Greg Haase) Date: Tue, 07 Sep 2010 14:46:25 -0700 Subject: Unexpected results with Kickstart In-Reply-To: <20100907203422.GK30596@exeter.usersys.redhat.com> Message-ID: Chris, That did it! In my efforts I even tried to specify that with a - at base on a line. Yet that didn't work. Thanks Greg On 9/7/10 1:34 PM, "Chris Lumens" wrote: >> Yet after the install I look at the anaconda-ks.cfg on the system created >> with the kickstart and it has: >> >> %packages >> @Base >> @Core >> @core >> %end >> >> Or some variation with the caps / lowercase. The thing is that I want as >> simple an install as possible. Is there something I am missing? > > The Base group is always selected unless you use "%packages --nobase" > (see > https://fedoraproject.org/wiki/Anaconda/Kickstart#Chapter_3._Package_Selection > ). > > The case-sensitive thing there is a little strange, but nothing to be > concerned about. It should not affect the number of packages installed. > > - Chris > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From email at jasonkohles.com Wed Sep 8 01:05:34 2010 From: email at jasonkohles.com (Jason Kohles) Date: Tue, 7 Sep 2010 21:05:34 -0400 Subject: Unexpected results with Kickstart In-Reply-To: <20100907203422.GK30596@exeter.usersys.redhat.com> References: <20100907203422.GK30596@exeter.usersys.redhat.com> Message-ID: On Sep 7, 2010, at 4:34 PM, Chris Lumens wrote: >> Yet after the install I look at the anaconda-ks.cfg on the system created >> with the kickstart and it has: >> >> %packages >> @Base >> @Core >> @core >> %end >> >> Or some variation with the caps / lowercase. The thing is that I want as >> simple an install as possible. Is there something I am missing? > > The Base group is always selected unless you use "%packages --nobase" > (see > https://fedoraproject.org/wiki/Anaconda/Kickstart#Chapter_3._Package_Selection). > > The case-sensitive thing there is a little strange, but nothing to be > concerned about. It should not affect the number of packages installed. > The core/Core thing is just because it looks at all the packages, and if every package that specifies a certain group is installed, then it includes that group in the %packages list. In this case some packages list their group as 'Core' and some as 'core', so it ends up in the list twice... From Marc.Herbert at gmail.com Wed Sep 8 09:00:37 2010 From: Marc.Herbert at gmail.com (Marc Herbert) Date: Wed, 08 Sep 2010 10:00:37 +0100 Subject: if PYTHONUNBUFFERED=x ksvalidator my.ks | grep '.'; then false; fi Message-ID: Hi, I was glad I just discovered "ksvalidator", implementing a useful feature. I was less happy about how painful it was to script it (see workarounds in email subject). Shouldn't these exit status and buffering problems be at least flagged in the usage message? Cheers, Marc From jwildebo at redhat.com Wed Sep 8 15:12:59 2010 From: jwildebo at redhat.com (Jan Wildeboer) Date: Wed, 08 Sep 2010 17:12:59 +0200 Subject: Unexpected results with Kickstart In-Reply-To: References: <20100907203422.GK30596@exeter.usersys.redhat.com> Message-ID: <4C87A7FB.2010309@redhat.com> On 09/08/2010 03:05 AM, Jason Kohles wrote: > The core/Core thing is just because it looks at all the packages, and if every package that specifies a certain group is installed, then it includes that group in the %packages list. In this case some packages list their group as 'Core' and some as 'core', so it ends up in the list twice... Is it worth finding all packages with "core" and changing them to "Core"? Or should the grouping become case-insensitive? Jan -- Jan H Wildeboer | EMEA Open Source Affairs | Office: +49 (0)89 205071-207 Red Hat GmbH | Mobile: +49 (0)174 33 23 249 Technopark II, Haus C | Fax: +49 (0)89 205071-111 Werner-von-Siemens-Ring 11 -15 | 85630 Grasbrunn | _____________________________________________________________________ Reg. Adresse: Red Hat GmbH, Technopark II, Haus C, Werner-von-Siemens-Ring 11 -15 85630 Grasbrunn, Handelsregister: Amtsgericht Muenchen HRB 153243 Geschaeftsfuehrer: Brendan Lane, Charlie Peters, Michael Cunningham, Charles Cachera _____________________________________________________________________ GPG Key: 3AC3C8AB Fingerprint: 3D1E C4E0 DD67 E16D E47A 9564 A72F 5C39 3AC3 C8AB From darko.listsniffer at gmail.com Wed Sep 8 18:00:59 2010 From: darko.listsniffer at gmail.com (darko gavrilovic) Date: Wed, 8 Sep 2010 14:00:59 -0400 Subject: Kickstart/TFTP not removing profiles In-Reply-To: References: Message-ID: On Wed, Sep 8, 2010 at 1:40 PM, darko gavrilovic wrote: > Hi, we are getting an issue where "pxeos -d" is not properly removing > TFTP profiles. Anyone able to fix it other than manually editing the > files in /tftp? > The OS version is RHEL 5.5 dg From ezyang at mit.edu Mon Sep 20 06:17:30 2010 From: ezyang at mit.edu (Edward Z. Yang) Date: Mon, 20 Sep 2010 02:17:30 -0400 Subject: Anaconda DBUS error on kickstart Message-ID: <1284963252-sup-8030@ezyang> Hello all, I am receiving the following Python exception when running Fedora 13 Kickstart installation: anaconda 13.42 exception report Traceback (most recent call first): File "/usr/lib/python2.6/site-packages/dbus/bus.py", line 125, in __new__ bus = cls._new_for_bus(address_or_type, mainloop=mainloop) File "/usr/lib/python2.6/site-packages/dbus/_dbus.py", line 108, in __new__ bus = BusConnection.__new__(subclass, bus_type, mainloop=mainloop) File "/usr/lib/python2.6/site-packages/dbus/_dbus.py" line 202, in __new__ private=private) File "/usr/lib/anaconda/isys.py", line 329, in getDeviceProperties bus = dbus.SystemBus() File "/usr/lib/anaconda/network.py", line 277, in __init__ devhash = isys.getDeviceProperties(dev=None) File "/usr/bin/anaconda", line 546, in network self._network = network.Network() File "/usr/lib/anaconda/kickstart.py", line 550, in execute devices = anaconda.network.netdevices File "/usr/lib/anaconda/kickstart.py", line 1124, in execute obj.execute(self.anaconda) File "/usr/bin/anaconda", line 1211, in anaconda.ksdata.execute() DBusException: org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. Here is the Kickstart file we are using: http://scripts.mit.edu/old-faithful.txt It appears to get past the disk partioning stage before getting stuck. We are running Anaconda inside Xen with a somewhat special network setup (eth0 for public network and eth1 for private network), I can provide more configuration details if you want. Googling did not indicate anyone had seen this particular error before. Installing the server manually is fine. Cheers, Edward From ezyang at mit.edu Mon Sep 20 06:37:44 2010 From: ezyang at mit.edu (Edward Z. Yang) Date: Mon, 20 Sep 2010 02:37:44 -0400 Subject: Anaconda DBUS error on kickstart In-Reply-To: <1284963252-sup-8030@ezyang> References: <1284963252-sup-8030@ezyang> Message-ID: <1284964323-sup-701@ezyang> Here are some more logs: http://web.mit.edu/~ezyang/Public/anaconda-log.txt Cheers, Edward Excerpts from Edward Z. Yang's message of Mon Sep 20 02:17:30 -0400 2010: > Hello all, > > I am receiving the following Python exception when running > Fedora 13 Kickstart installation: > > anaconda 13.42 exception report > Traceback (most recent call first): > File "/usr/lib/python2.6/site-packages/dbus/bus.py", line 125, in __new__ > bus = cls._new_for_bus(address_or_type, mainloop=mainloop) > File "/usr/lib/python2.6/site-packages/dbus/_dbus.py", line 108, in __new__ > bus = BusConnection.__new__(subclass, bus_type, mainloop=mainloop) > File "/usr/lib/python2.6/site-packages/dbus/_dbus.py" line 202, in __new__ > private=private) > File "/usr/lib/anaconda/isys.py", line 329, in getDeviceProperties > bus = dbus.SystemBus() > File "/usr/lib/anaconda/network.py", line 277, in __init__ > devhash = isys.getDeviceProperties(dev=None) > File "/usr/bin/anaconda", line 546, in network > self._network = network.Network() > File "/usr/lib/anaconda/kickstart.py", line 550, in execute > devices = anaconda.network.netdevices > File "/usr/lib/anaconda/kickstart.py", line 1124, in execute > obj.execute(self.anaconda) > File "/usr/bin/anaconda", line 1211, in > anaconda.ksdata.execute() > DBusException: org.freedesktop.DBus.Error.NoReply: > Did not receive a reply. Possible causes include: > the remote application did not send a reply, the > message bus security policy blocked the reply, the > reply timeout expired, or the network connection > was broken. > > Here is the Kickstart file we are using: > > http://scripts.mit.edu/old-faithful.txt > > It appears to get past the disk partioning stage before getting stuck. > We are running Anaconda inside Xen with a somewhat special network > setup (eth0 for public network and eth1 for private network), I can provide > more configuration details if you want. > > Googling did not indicate anyone had seen this particular error before. > Installing the server manually is fine. > > Cheers, > Edward From ezyang at mit.edu Mon Sep 20 06:50:03 2010 From: ezyang at mit.edu (Edward Z. Yang) Date: Mon, 20 Sep 2010 02:50:03 -0400 Subject: Anaconda DBUS error on kickstart In-Reply-To: <1284963252-sup-8030@ezyang> References: <1284963252-sup-8030@ezyang> Message-ID: <1284965266-sup-6972@ezyang> More information: this error only occurs when we craft a custom initrd with our kickstart file in /root/scripts.ks and then invoke the kernel with file:/root/scripts.ks. If we retrieve the kickstart file over the network the install proceeds normally. Cheers, Edward From tom at ng23.net Mon Sep 20 10:05:02 2010 From: tom at ng23.net (Tom Brown) Date: Mon, 20 Sep 2010 11:05:02 +0100 Subject: Encrypt disk at install time Message-ID: Hi Apart from /boot is it possible to encrypt a disk at install time? It would be a huge win if we can do this but i cant see anything about how i may achieve it. This would be on RHEL 5.x thanks From hhoffman at ip-solutions.net Mon Sep 20 10:59:26 2010 From: hhoffman at ip-solutions.net (Harry Hoffman) Date: Mon, 20 Sep 2010 06:59:26 -0400 Subject: Encrypt disk at install time In-Reply-To: References: Message-ID: <1284980366.10755.5.camel@n1-20-152.dhcp.drexel.edu> Here's the output from a anaconda generated kickstart file after install: # The following is the partition information you requested # Note that any partitions you deleted are not expressed # here so unless you clear all partitions first, this is # not guaranteed to work #clearpart --all --initlabel #part /boot --fstype ext3 --size=250 --ondisk=sda --asprimary #part swap --encrypted --size=1024 --ondisk=sda --asprimary #part pv.100000 --size=1 --grow --ondisk=sda #volgroup internal_hd --pesize=32768 pv.100000 #logvol / --fstype ext3 --encrypted --name=slash --vgname=internal_hd --size=75008 I guess you just need to specify how/where to set the encryption key. HTH, Harry On Mon, 2010-09-20 at 11:05 +0100, Tom Brown wrote: > Hi > > Apart from /boot is it possible to encrypt a disk at install time? It > would be a huge win if we can do this but i cant see anything about > how i may achieve it. > > This would be on RHEL 5.x > > thanks > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Moray.Henderson at ict-software.org Mon Sep 20 16:20:27 2010 From: Moray.Henderson at ict-software.org (Moray Henderson) Date: Mon, 20 Sep 2010 17:20:27 +0100 Subject: Encrypt disk at install time In-Reply-To: References: Message-ID: <005201cb58df$bd631650$382942f0$@Henderson@ict-software.org> Tom Brown wrote: >Apart from /boot is it possible to encrypt a disk at install time? It >would be a huge win if we can do this but i cant see anything about >how i may achieve it. > >This would be on RHEL 5.x Yes. Wherever you have a "part" command in your kickstart, add --encrypted and --password options. In your case, part pv.100000 --size=1 --grow --ondisk=sda would become (still on the one line) part pv.100000 --size=1 --grow --ondisk=sda --encrypted --password HelloWorld See http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Instal lation_Guide/ch-kickstart2.html DO NOT forget the password! Moray. "To err is human.? To purr, feline" From csd_b at daudt.org Thu Sep 23 18:24:30 2010 From: csd_b at daudt.org (Christian Daudt) Date: Thu, 23 Sep 2010 11:24:30 -0700 Subject: conflicting packages in kickstart Message-ID: [I tried #kickstart channel on IRC but it seems awfully quiet, so reposting to the list] hi, I have a kickstart question (well Meego's image-creator, which uses kickstart): If I have 2 pkgs that provide the same thing, how can I get kickstart to be happy about using the one I picked? There is "grep.rpm" already in one of the repos I'm using, and I added a busybox rpm with grep in it, which I want to use in my image instead of the grep already in the repo so I listed my busybox in the ks file, but it still tries to grab the grep rpm through other indirect dependencies and complains: Error: failed to create image : Unable to install: [('file /bin/grep conflicts between attempted installs of busybox-mxc-1.16.0-24.1.armv5tel and grep-2.5.3-6.1.armv5tel' I have the busybox rpm listed in my ks file but not the grep rpm (its being brought in indirectly). I even tried "-grep" in the rpm list to see if that would help but it didn't. Any pointers would be appreciated. Here's are the relevant parts of the ks file I tried: ---------------------------------------------------------------- ### repos start ### repo --name=csddev --baseurl=http://amigo:82/home:/csd:/branches:/MeeGo:/1.0:/Core/ARM_Cross/ --save --debuginfo --source repo --name=core --baseurl=http://amigo:82/MeeGo:/1.0:/Core/ARM_Cross --save --debuginfo --source ### repos end ### %packages --excludedocs busybox-mxc <<<<< Comes from csddev repo above, and has a "provides: grep" defined in it, along with /bin/grep -grep <<< tried both with this line and without it openssh-server <<< this one is in repo "core" and pulls in "grep" rpm indirectly --------------------------------------------------------------------------- thanks, csd From ghaase at syntheticgenomics.com Thu Sep 23 19:22:28 2010 From: ghaase at syntheticgenomics.com (Greg Haase) Date: Thu, 23 Sep 2010 12:22:28 -0700 Subject: conflicting packages in kickstart In-Reply-To: Message-ID: Hi Christian, I had a similar issue when using rpms I created. To get them to install properly I needed to set the priority with yum priorities. Here is an article that discusses how to do so for RPM Forge. Look under Point 2. Where it says "2.1. Priorities" on the following link. http://wiki.centos.org/AdditionalResources/Repositories/RPMForge I would do this in the %post section ################ %post yum install yum-plugin-priorities cat <> /etc/yum/pluginconf.d/priorities.conf [main] enabled=1 EOF cat <> /etc/my-custom-yum-repo.repo repo --name=csddev --baseurl=http://amigo:82/home:/csd:/branches:/MeeGo:/1.0:/Core/ARM_Cross/ --save --debuginfo --source repo --name=core --baseurl=http://amigo:82/MeeGo:/1.0:/Core/ARM_Cross --save --debuginfo --source priority=1 EOF ## Then install the package Yum erase the-package-I-don't-want Yum install my-special-package-with-a-duplicate-name ############# I hope that helps. Greg On 9/23/10 11:24 AM, "Christian Daudt" wrote: > [I tried #kickstart channel on IRC but it seems awfully quiet, so > reposting to the list] > hi, > I have a kickstart question (well Meego's image-creator, which uses > kickstart): If I have 2 pkgs that provide the same thing, how can I > get kickstart to be happy about using the one I picked? There is > "grep.rpm" already in one of the repos I'm using, and I added a > busybox rpm with grep in it, which I want to use in my image instead > of the grep already in the repo so I listed my busybox in the ks file, > but it still tries to grab the grep rpm through other indirect > dependencies and complains: > Error: failed to create image : Unable to install: [('file /bin/grep > conflicts between attempted installs of > busybox-mxc-1.16.0-24.1.armv5tel and grep-2.5.3-6.1.armv5tel' > I have the busybox rpm listed in my ks file but not the grep rpm (its > being brought in indirectly). I even tried "-grep" in the rpm list to > see if that would help but it didn't. Any pointers would be > appreciated. > > Here's are the relevant parts of the ks file I tried: > ---------------------------------------------------------------- > ### repos start ### > repo --name=csddev > --baseurl=http://amigo:82/home:/csd:/branches:/MeeGo:/1.0:/Core/ARM_Cross/ > --save --debuginfo --source > repo --name=core --baseurl=http://amigo:82/MeeGo:/1.0:/Core/ARM_Cross > --save --debuginfo --source > > > ### repos end ### > > %packages --excludedocs > > busybox-mxc <<<<< Comes from csddev repo above, and has a "provides: > grep" defined in it, along with /bin/grep > > -grep <<< tried both with this line and without it > > openssh-server <<< this one is in repo "core" and pulls in "grep" rpm > indirectly > --------------------------------------------------------------------------- > > thanks, > csd > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From csd_b at daudt.org Thu Sep 23 20:24:17 2010 From: csd_b at daudt.org (Christian Daudt) Date: Thu, 23 Sep 2010 13:24:17 -0700 Subject: conflicting packages in kickstart In-Reply-To: References: Message-ID: On Thu, Sep 23, 2010 at 12:22 PM, Greg Haase wrote: > Hi Christian, I had a similar issue when using rpms I created. To get them > to install properly I needed to set the priority with yum priorities. > > Here is an article that discusses how to do so for RPM Forge. ?Look under > Point 2. ?Where it says "2.1. Priorities" on the following link. > > > http://wiki.centos.org/AdditionalResources/Repositories/RPMForge > > I would do this in the %post section > ################ > %post > yum install yum-plugin-priorities > > cat <> /etc/yum/pluginconf.d/priorities.conf > [main] > enabled=1 > EOF > > cat <> /etc/my-custom-yum-repo.repo > ?repo --name=csddev > --baseurl=http://amigo:82/home:/csd:/branches:/MeeGo:/1.0:/Core/ARM_Cross/ > --save --debuginfo --source > repo --name=core ?--baseurl=http://amigo:82/MeeGo:/1.0:/Core/ARM_Cross > ?--save --debuginfo --source > priority=1 > ?EOF > > ## Then install the package > Yum erase ? ?the-package-I-don't-want > Yum install my-special-package-with-a-duplicate-name > ############# > > I hope that helps. Hi Greg, Unfortunately this won't work in my case - my objective is to build a very small target (which is why I'm replacing some utilities which busybox counterparts) so I am not installing yum (or rpm for that matter) and can't do this post-install. If nothing else works, I'm planning on munging the 2 repos into a third one, but I don't think this will scale well if I have > 1 busybox options later (or other replacement tools). So I'm really hoping to find a solution within the kickstart configuration. Your reply doesn't give me much hope that such a solution exists though... Thanks, csd > > > Greg > > > > On 9/23/10 11:24 AM, "Christian Daudt" wrote: > >> [I tried #kickstart channel on IRC but it seems awfully quiet, so >> reposting to the list] >> hi, >> ?I have a kickstart question (well Meego's image-creator, which uses >> kickstart): If I have 2 pkgs that provide the same thing, how can I >> get kickstart to be happy about using the one I picked? There is >> "grep.rpm" ?already in one of the repos I'm using, and I added a >> busybox rpm with grep in it, which I want to use in my image instead >> of the grep already in the repo so I listed my busybox in the ks file, >> but it still tries to grab the grep rpm through other indirect >> dependencies and complains: >> Error: failed to create image : Unable to install: [('file /bin/grep >> conflicts between attempted installs of >> busybox-mxc-1.16.0-24.1.armv5tel and grep-2.5.3-6.1.armv5tel' >> I have the busybox rpm listed in my ks file but not the grep rpm (its >> being brought in indirectly). I even tried "-grep" in the rpm list to >> see if that would help but it didn't. Any pointers would be >> appreciated. >> >> Here's are the relevant parts of the ks file I tried: >> ---------------------------------------------------------------- >> ### repos start ### >> repo --name=csddev >> --baseurl=http://amigo:82/home:/csd:/branches:/MeeGo:/1.0:/Core/ARM_Cross/ >> --save --debuginfo --source >> repo --name=core ?--baseurl=http://amigo:82/MeeGo:/1.0:/Core/ARM_Cross >> ?--save --debuginfo --source >> >> >> ### repos end ### >> >> %packages --excludedocs >> >> busybox-mxc <<<<< Comes from csddev repo above, and has a "provides: >> grep" defined in it, along with /bin/grep >> >> -grep <<< tried both with this line and without it >> >> openssh-server <<< this one is in repo "core" and pulls in "grep" rpm >> indirectly >> --------------------------------------------------------------------------- >> >> thanks, >> ? ?csd >> >> _______________________________________________ >> 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 ezyang at mit.edu Sat Sep 25 22:06:28 2010 From: ezyang at mit.edu (Edward Z. Yang) Date: Sat, 25 Sep 2010 18:06:28 -0400 Subject: Error when printing error Message-ID: <1285452373-sup-2483@ezyang> Looks like the error code is buggy. 15:03:00,290 CRITICAL anaconda: anaconda 13.42 exception report Traceback (most recent call first): File "/usr/lib64/python2.6/site-packages/snack.py", line 225, in __init__ self.w = _snack.textbox(width, height, text, scroll, wrap) File "/usr/lib/anaconda/text.py", line 309, in detailedMessageWindow lt = Textbox(60, 6, longText, scroll=1, wrap=1) File "/usr/lib/anaconda/kickstart.py", line 123, in run intf.detailedMessageWindow(_("Scriptlet Failure"), msg, err) File "/usr/lib/anaconda/kickstart.py", line 1282, in map (lambda s: s.run(anaconda.rootPath, flags.serial, anaconda.intf), postScripts) File "/usr/lib/anaconda/kickstart.py", line 1282, in runPostScripts map (lambda s: s.run(anaconda.rootPath, flags.serial, anaconda.intf), postScripts) File "/usr/lib/anaconda/dispatch.py", line 205, in moveStep rc = stepFunc(self.anaconda) File "/usr/lib/anaconda/dispatch.py", line 126, in gotoNext self.moveStep() File "/usr/lib/anaconda/text.py", line 659, in run anaconda.dispatch.gotoNext() File "/usr/bin/anaconda", line 1223, in anaconda.intf.run(anaconda) TypeError: argument 3 must be string, not list Cheers, Edward From jlaska at redhat.com Wed Sep 29 19:15:37 2010 From: jlaska at redhat.com (James Laska) Date: Wed, 29 Sep 2010 15:15:37 -0400 Subject: Error when printing error In-Reply-To: <1285452373-sup-2483@ezyang> References: <1285452373-sup-2483@ezyang> Message-ID: <1285787737.14936.136.camel@flatline> On Sat, 2010-09-25 at 18:06 -0400, Edward Z. Yang wrote: > Looks like the error code is buggy. > > 15:03:00,290 CRITICAL anaconda: anaconda 13.42 exception report > Traceback (most recent call first): > File "/usr/lib64/python2.6/site-packages/snack.py", line 225, in __init__ > self.w = _snack.textbox(width, height, text, scroll, wrap) > File "/usr/lib/anaconda/text.py", line 309, in detailedMessageWindow > lt = Textbox(60, 6, longText, scroll=1, wrap=1) > File "/usr/lib/anaconda/kickstart.py", line 123, in run > intf.detailedMessageWindow(_("Scriptlet Failure"), msg, err) > File "/usr/lib/anaconda/kickstart.py", line 1282, in > map (lambda s: s.run(anaconda.rootPath, flags.serial, anaconda.intf), postScripts) > File "/usr/lib/anaconda/kickstart.py", line 1282, in runPostScripts > map (lambda s: s.run(anaconda.rootPath, flags.serial, anaconda.intf), postScripts) > File "/usr/lib/anaconda/dispatch.py", line 205, in moveStep > rc = stepFunc(self.anaconda) > File "/usr/lib/anaconda/dispatch.py", line 126, in gotoNext > self.moveStep() > File "/usr/lib/anaconda/text.py", line 659, in run > anaconda.dispatch.gotoNext() > File "/usr/bin/anaconda", line 1223, in > anaconda.intf.run(anaconda) > TypeError: argument 3 must be string, not list > > Cheers, > Edward Can you reproduce this failure using Fedora 14 Beta [1]? A quick inspection of anaconda in Fedora 14 shows that the failure is likely present there as well. Either way, please file a bug in bugzilla.redhat.com [2] so that anaconda developers can prioritize this failure. Thanks, James [1] http://fedoraproject.org/get-prerelease [2] https://bugzilla.redhat.com/enter_bug.cgi?product=Fedora&version=14&component=anaconda -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 198 bytes Desc: This is a digitally signed message part URL: From co55-sy1t at dea.spamcon.org Thu Sep 30 05:35:21 2010 From: co55-sy1t at dea.spamcon.org (whitivery) Date: Wed, 29 Sep 2010 22:35:21 -0700 Subject: The %end tag causes errors Message-ID: <9s78a6pubeuipo8tf4a676g1g1260u6ttk@4ax.com> In a CentOS 5.5 kickstart file, adding %end to mark the end of the %packages, %pre, or %post sections causes an error. In the first two cases the kickstart fails and stops. For the %post it gives an error message but it has already run all of my post steps so it doesn't have a functional impact. Without the %end lines, the kickstart file works fine. Details: - %end on %packages: traceback starts with: File "/usr/bin/anaconda", line 733, in ? and near end says KickstartParseError and identifies a kickstart line (which doesn't correspond to a %end) - %end on %pre: traceback starts with: File "/usr/bin/anaconda", line 900, in ? - %end on %post: my steps work but at end: {temporary ks filename}: line 490: fg: no job control I don't see anaconda identify its version, but the RPM included with CentOS 5.5 is anaconda-11.1.2.209-1.el5. From srobson at cadence.com Thu Sep 30 16:14:47 2010 From: srobson at cadence.com (Steve Robson) Date: Thu, 30 Sep 2010 17:14:47 +0100 Subject: Kickstart-list Digest, Vol 79, Issue 11 In-Reply-To: References: Message-ID: <4CA4B777.6020403@cadence.com> kickstart-list-request at redhat.com wrote: > Subject: The %end tag causes errors > From: whitivery > Date: Thu, 30 Sep 2010 07:35:21 +0200 > > To: "Kickstart-list at redhat.com" > > In a CentOS 5.5 kickstart file, adding %end to mark the end of the > %packages, %pre, or %post sections causes an error. In the first two > cases the kickstart fails and stops. For the %post it gives an error > message but it has already run all of my post steps so it doesn't have a > functional impact. > > Without the %end lines, the kickstart file works fine. Hi there, Not sure where you got "%end" from but I've never heard of it or used it. It's not mentioned in the RHEL5 kickstart documentation starting at http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Installation_Guide/s1-kickstart2-file.html -- Regards, Steve From ezyang at mit.edu Thu Sep 30 16:21:01 2010 From: ezyang at mit.edu (Edward Z. Yang) Date: Thu, 30 Sep 2010 12:21:01 -0400 Subject: Kickstart-list Digest, Vol 79, Issue 11 In-Reply-To: <4CA4B777.6020403@cadence.com> References: <4CA4B777.6020403@cadence.com> Message-ID: <1285863630-sup-8284@ezyang> Excerpts from Steve Robson's message of Thu Sep 30 12:14:47 -0400 2010: > Not sure where you got "%end" from but I've never heard of it or used > it. It's not mentioned in the RHEL5 kickstart documentation starting at > http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Installation_Guide/s1-kickstart2-file.html Hello Steve, %end is recommended practice by Fedora's wiki page on Kickstart: http://fedoraproject.org/wiki/Anaconda/Kickstart Cheers, Edward From srobson at cadence.com Thu Sep 30 16:32:09 2010 From: srobson at cadence.com (Steve Robson) Date: Thu, 30 Sep 2010 17:32:09 +0100 Subject: Kickstart-list Digest, Vol 79, Issue 11 In-Reply-To: <1285863630-sup-8284@ezyang> References: <4CA4B777.6020403@cadence.com> <1285863630-sup-8284@ezyang> Message-ID: <4CA4BB89.5010006@cadence.com> Edward Z. Yang wrote: > Excerpts from Steve Robson's message of Thu Sep 30 12:14:47 -0400 2010: >> Not sure where you got "%end" from but I've never heard of it or used >> it. It's not mentioned in the RHEL5 kickstart documentation starting at >> http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Installation_Guide/s1-kickstart2-file.html > > Hello Steve, > > %end is recommended practice by Fedora's wiki page on Kickstart: > > http://fedoraproject.org/wiki/Anaconda/Kickstart Hi Edward, Current Fedora is years ahead of RHEL5. For instance, I see "fc6" in some of the package names on the RHEL5.3 I'm currently using. The "%end" directive must be new. Well, newer than RHEL5.3. -- Regards, Steve From ghaase at syntheticgenomics.com Thu Sep 30 16:40:51 2010 From: ghaase at syntheticgenomics.com (Greg Haase) Date: Thu, 30 Sep 2010 09:40:51 -0700 Subject: Kickstart-list Digest, Vol 79, Issue 11 In-Reply-To: <4CA4B777.6020403@cadence.com> Message-ID: Looking at the documentation listed on the Fedora Project, I see a reference to the %end tag. Under the title heading " Creating the Kickstart File" you will find this: "Old-style kickstart syntax does not specify for any end token after the %packages section or scripts. In this case, the following section order is required. New-style kickstart syntax requires using %end following these sections, which means that ordering is not required" http://fedoraproject.org/wiki/Anaconda/Kickstart I think this works in Fedora and not CentOS where I actually had to remove them. So I guess the %end is bleeding edge while the rest of the body has yet to accept it into common practice. Greg Greg Haase ghaase at syntheticgenomics.com On 9/30/10 9:14 AM, "Steve Robson" wrote: > kickstart-list-request at redhat.com wrote: >> Subject: The %end tag causes errors >> From: whitivery >> Date: Thu, 30 Sep 2010 07:35:21 +0200 >> >> To: "Kickstart-list at redhat.com" >> >> In a CentOS 5.5 kickstart file, adding %end to mark the end of the >> %packages, %pre, or %post sections causes an error. In the first two >> cases the kickstart fails and stops. For the %post it gives an error >> message but it has already run all of my post steps so it doesn't have a >> functional impact. >> >> Without the %end lines, the kickstart file works fine. > > Hi there, > > Not sure where you got "%end" from but I've never heard of it or used > it. It's not mentioned in the RHEL5 kickstart documentation starting at > http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Installation > _Guide/s1-kickstart2-file.html