devel May 2005

devel@lists.centos.org

7 participants
6 discussions

Trivial mod to httpd-suexec-2.0.52-9.ent.centos4.1
by Ed Clarke 02 Jun '05

02 Jun '05

The httpd-suexec package is part of the httpd source RPM. As part of the security model of suexec, a directory is hard coded into /usr/sbin/suexec (/var/www in Centos 4) that must be the root of all cgi-bin directories on the system. As an alternate, the UserDir (/home/*/public_html) may be enabled for CGI execution - but this is not done by default. As a web-hosting company, we prefer to move the default cgi-bin directory to /home/cgi-bin (and subdirectories) rather than /var/www. This permits us to keep all customer files on one filesystem (/home) and still use Webmin/Usermin/Virtualmin. This also makes it easier to enforce quota restrictions. This is the way we add virtual systems (using cilia as an example): mkdir /home/cgi-bin/cilia chmod 755 /home/cgi-bin/cilia chown cilia.cilia /home/cgi-bin/cilia ln -s /home/cilia/cgi-bin /home/cgi-bin/cilia This follows the security model described in http://httpd.apache.org/docs-2.0/suexec.html although I'm not sure why this restriction is necessary. You get some obscure error messages about "premature end of script headers" if you don't do this correctly. The real error is written to /var/log/httpd/suexec.log but takes a while to find. The change to make this is simple - two lines in the httpd.spec file ( could be one ): %define cgidir /home/cgi-bin <--- added line --with-suexec-docroot=%{cgidir} \ <--- changed line Is this worth doing in CentosPlus? It looks like you have to recompile all of httpd (Apache 2.x) even though you're only changing the one file in the sub package.

3 2

Mirror breakage
by John Newbigin 24 May '05

24 May '05

The tool I use for managing my local CentOS-3 mirror has started to choke on some changes to the centos mirrors. The URL I am using is: http://mirror.centos.org/centos-3/3/updates/i386/ And I get the error: /tmp/rpmfindjLOxxX:271: error: htmlParseEntityRef: expecting ';' to donate a server: <a href="mailto:johnny@centos.org?subject=Server Donation&c The offending html is: <a href="mailto:johnny@centos.org?subject=Server Donation&cc=lance(a)centos.org ">Contact us</a> I am not using rpmfind, but reusing code I wrote for rpmfind (the http/html downloader). The problem appears to be libxml (libxml2-2.4.19-6.ent) barfing. I wonder though if the page being generated by the server is valid? As a workaround I pointed to: http://master.centos.org/centos-3/3/updates/i386/ which works but kind of defeats the purpose. ... I checked the page with the w3c html valiadtor. Not good http://validator.w3.org/check?uri=http%3A%2F%2Fmirror.centos.org%2Fcentos-3… Line 57, column 206: cannot generate system identifier for general entity "cc" ...y(a)centos.org?subject=Server Donation&cc=lance(a)centos.org ">Contact us</a></b> An entity reference was found in the document, but there is no reference by that name defined. Often this is caused by misspelling the reference name, unencoded ampersands, or by leaving off the trailing semicolon (;). The most common cause of this error is unencoded ampersands in URLs as described by the WDG in "Ampersands in URLs". Entity references start with an ampersand (&) and end with a semicolon (;). If you want to use a literal ampersand in your document you must encode it as "&" (even inside URLs!). Be careful to end entity references with a semicolon or your entity reference may get interpreted in connection with the following text. Also keep in mind that named entity references are case-sensitive; &Aelig; and æ are different characters. Note that in most documents, errors related to entity references will trigger up to 5 separate messages from the Validator. Usually these will all disappear when the original problem is fixed. John. -- John Newbigin Computer Systems Officer Faculty of Information and Communication Technologies Swinburne University of Technology Melbourne, Australia http://www.ict.swin.edu.au/staff/jnewbigin

1 0

Re: [CentOS] centosplus and contrib rpms
by John Newbigin 16 May '05

16 May '05

Johnny Hughes wrote: > > Now there is a Readme.txt in this directory: With a capital R? README or readme.txt would be more common spellings under Unix/Linux. The other problem with a single readme file is that it can't be contributed by more than one person, so if a new RPM comes along, the person with the master readme has to edit and upload it. If the readme was per RPM or set of RPMs then that would not be a problem. You could also include info like who built the RPM and how to contact them etc. README.php README.kernel etc. could be one way of doing this (in a docs directory?) John. > > http://mirror.centos.org/centos/4/centosplus/ > > There will be one for contrib and extras later today. > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > CentOS mailing list > CentOS(a)centos.org > http://lists.centos.org/mailman/listinfo/centos -- John Newbigin Computer Systems Officer Faculty of Information and Communication Technologies Swinburne University of Technology Melbourne, Australia http://www.ict.swin.edu.au/staff/jnewbigin

3 2

Anaconda errors when building custom CentOS-based distro
by Sam Hart 06 May '05

06 May '05

I am building a custom CentOS 4.0 based distribution (basically, trimmed down size and a few custom packages). I have successfully built other RPM-based distributions in the past without much trouble. However, this CentOS 4.0-based one is giving me headaches. The document I follow for remastering CentOS 4.0 is this one: http://fedoranews.org/contributors/gene_czarcinski/update_distro/ I know it is written for FC3, but I also was under the impression it contained information more general. When I run buildinstall with the CentOS 4.0 RPMs (right now, it contains no custom RPMs, these are all stock CentOS 4.0) it creates boot images which do not contain the anaconda script (/usr/bin/anaconda). Here is the relevant output: prickly-pear:/work# buildinstall --pkgorder $BASE/custom/pkgfile \ > --version 1 --product "Custom01" --release "Custom01" --prodpath CentOS $BASE/custom/i386 warning: /work/custom/i386/CentOS/RPMS/anaconda-runtime-10.1.1.13-1.centos4.i386.rpm: V3 DSA signature: NOKEY, key ID 443e1821 Running buildinstall... /work/custom/i386/buildinstall.tree.21011 /work /work Going to run buildinstall again warning: /work/custom/i386/CentOS/RPMS/anaconda-runtime-10.1.1.13-1.centos4.i386.rpm: V3 DSA signature: NOKEY, key ID 443e1821 Building images... Assembling package list... Expanding text packages... Expanding graphical packages... retrieving timezones Creating nsswitch.conf Fixing up /etc/man.config to point into /mnt/sysimage Running mkfontdir... /work/custom/i386/buildinstall.tree.21011/upd-instroot: line 929: /usr/X11R6/bin/mkfontdir: No such file or directory Getting pango modules /usr/sbin/chroot: cannot run command `/usr/bin/pango-querymodules': No such file or directory Getting gtk2 input method modules /usr/sbin/chroot: cannot run command `/usr/bin/gtk-query-immodules-2.0': No such file or directory Getting gtk2 gdk-pixbuf loaders /usr/sbin/chroot: cannot run command `/usr/bin/gdk-pixbuf-query-loaders': No such file or directory Scrubbing trees... /tmp/treedir.21024/image-template /usr/sbin/chroot: cannot run command `usr/sbin/ldconfig': No such file or directory Scrubbing trees... /tmp/treedir.21024/instimage cp: cannot stat `/tmp/treedir.21024/instimage/boot/memtest*': No such file or directory /usr/sbin/chroot: cannot run command `usr/sbin/ldconfig': No such file or directory Compressing .mo files in stage2 images... Creating fontconfig cache /usr/sbin/chroot: cannot run command `/usr/bin/fc-cache': No such file or directory Creating debug dir Patching python library... Removing unused python files in hdimage... /usr/sbin/chroot: cannot run command `/usr/bin/anaconda': No such file or directory I have googled for this error, and found many mentions of it, but no real solutions. The anaconda-runtime, anaconda, anaconda-help and busybox-anaconda packages are all installed and present as RPMs in the RPM asset directory (/work/custom/i386/CentOS/RPMS). Similarly, the other files that report errors (`usr/sbin/ldconfig', etc) are installed and present with their parent RPMs in the RPM asset directory. I am reasonably certain that I have dependency closure on these RPMs (I have made in-house rpmstrap scripts for it and can install via bootstrap), and my comps.xml file seems correct. Has anyone ever seen this "/usr/sbin/chroot: cannot run command `/usr/bin/anaconda'" error ever before? If so, any suggestions for what I can do to fix this? I'm not certain if this is the right place to ask this, but since the only RPM-base that is giving me this problem is CentOS 4.0 I figured it was the best place to go. -- ''''''''''''''''''''''''' .O. Sam Hart, sam(a)progeny.com ..O Progeny Linux Systems, Inc OOO <http://www.progeny.com/>

2 5

Hello~centos can support this?!
by ryo wong 06 May '05

06 May '05

centos can support promise sata150 tx2 plus?! can support or not plz reply me~thx!!

1 0

kernel-2.6.9-5.0.5.EL missing /lib/modules/2.6.9-5.0.5.EL/build directory
by Charlie Brady 03 May '05

03 May '05

Does anyone know whether this is an upstream change or a build problem? --- Charlie

2 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

devel May 2005