search for: linuxweb

...d prosper with his great knowledge .... His work deserves to be listed at the top of those lists of top Free Open Source Software success stories. Happy New Year Tom! PS- Thanks also to everyone else on the shorewall project. May the FOSS be with you! -- David R. Hartley Linux Web Project penguin@linuxweb.org

I installed a CentOS-5 core OS (using --nobase in my kickstart). For some reason, it included mysql-5.0.22. When I do "yum remove mysql", it says it will also remove exim and mdadm for dependencies. I don't care that exim will be removed, but I need mdadm as I'm doing software RAID. But why are these even related? When I do: rpm -q --requires mysql neither exim or mdadm is

I'm interested in doing centralized logging & analysis of logs from my CentOS boxes. I messed around with syslog-ng and it seems like it's a better syslog. But I noticed most of the "usual suspects" of third-party repo maintainers (Dag, Axel, etc.) don't include it. Karanbir has an el4 version, but all my boxes are CentOS 5. I was going to go ahead and roll my own

Using audit2allow, I was able to create a policy module for selinux: audit2allow -i /var/log/audit/audit.log -M mysqld (creates mysqld.pp and mysqld.te) I want to distribute this to all my puppet clients. I can easily put this file in /etc/selinux/targeted/modules/active/modules But even after reboot, although I can see the module listed: semodule -l ... it doesn''t seem to actually

I''m almost afraid to ask because this probably sticks out like a sore thumb on the wiki somewhere, and I''m just not seeing it. But I have a bunch of files with the same owner/mode. Is there a way to somehow collapse these, so that all I repeat is the file name? Example: file { "/opt/test/1": owner => "jtan", group =>

I had a 5.0 kickstart server which did "core" installs of CentOS (i.e.: "%packages --nobase"). I recently setup a 5.2 kickstart server. Using the same kickstart configs, I notice a few packages are always missing, notably: yum (!!), selinux-policy-targeted (even though I have "selinux --enforcing" in the kickstart, it always ends up with that package missing and

I saw this in Logwatch today for one of my servers: --------------------- yum Begin ------------------------ Packages Installed: samba-common.i386 3.0.23c-2.el5.2.0.2 samba.i386 3.0.23c-2.el5.2.0.2 Packages Erased: samba-common samba ---------------------- yum End ------------------------- No one, including myself, has even logged into this box in the past few

As I sit here staring at my manifests (and pondering a module to control Glassfish), I''m pondering the best way to do a first-run of puppet to get the initial repositories and puppet configuration downloaded. My theory is that I''ll modify rc.local (or replace firstboot-tui, we don''t use it with automated builds) to run puppetd twice - once to request the certificate,

Today I installed a core CentOS-5 (--nobase). After that, I "yum update", rebooted, then "yum install xen kernel-xen". When I boot into the xen kernel, I can't get networking on eth0. It can ping itself, but not anything else. There are no routes listed, when I try to add a default gateway, it just says there's no network reachable. When I change

We've noticed our fully-virtualized KVM guests' time keeps getting faster, despite running ntpd. A quick google reveals this to be a known problem with virtual guests. It seems there are a variety of solutions, some apparently vendor-specific. What is the best practice as of now for KVM guests? * Setting independent_wallclock=1 (or is that a xen-only thing)? * Passing kernel

I have a NFS mount that I want apache to be able to serve files from. According to this doc: http://www.centos.org/docs/5/html/5.1/Deployment_Guide/rhlcommon-section-0097.html I should be able to mount it with a context that will allow apache to access it. But when I try the command they suggest: [root at vm-37:~] mount -t nfs -o \ context=system_u:object_r:httpd_sys_content_t \

Does Red Hat make available spec files for future releases? I took the existing mysql 5.0 spec file and, with a few mods here and there, was able to build the 5.1 rpm. But there are some new things which I am curious how they will deal with (ndb stuff, primarily), and was wondering if they have available the "beta" SRPMs or spec files for future versions of software. johnn

I believe I used to pull sox for el5 from dag's site, but it doesn't appear to be there anymore. Nor on rpmforge, atrpms, epel. Does any (trusted/reliable) repo out there have sox built for el5? I started a rebuild of fedora-12 version, but it's a pretty major dependency hell. johnny

The FileServingConfiguration doesn''t even mention directories. And the reference page doesn''t give an example. Really, all I want is to fileserve a directory, but with everything in it retaining its attributes. When I do this: file { "/var/www/html": ensure => directory, recurse => true, owner => "apache",

Is there a way to list the packages that are in a "base" install, but that are not in a "nobase" (core) install? I did a nobase install, then ran "yum groupinstall Base", but this just lists everything in base, including the core packages. Mainly, I'm just looking to audit the packages, and add only necessary ones back to a nobase install. So far, I've

I get the same error as this person: http://www.centos.org/modules/newbb/viewtopic.php?topic_id=9349 Like the poster, I get the error with tomcat 6.0.13 on x86_64, kernel 2.6.18-8.1.8.el5xen and both jre 1.6.0_01 and 1.6.0_02 When I revert to jre 5u12 on the same kernel, I don't get the problem. If I stay with jre 6, but use kernel 2.6.18-8.el5, I also don't get the error. So

Does anyone know of a CentOS-5 (el5) repo for rsyslog, that's being maintained on a relatively regular basis? I checked all the usual suspects (dag, karan, epel, etc.), but they either don't have rsyslog at all, or they have an old version. Currently, I'm doing a rebuild of the Fedora-8 srpm, but it's at version 2.0.2. I wanted to move to 3.11.x for the disk-assisted

How similar should my rpm build machine be to my target deploy machines? Like, do you have to build on a multi-core machine if you plan to run on a multi-core machine? Or as long as the arch is the same, nothing else matters? johnn

Searching through the archives, it seems this topic has come up several times in the past 2.5 years. Did anyone ever take on the tasks of creating an official CentOS AMI? We're currently using the RightScale ones, but I think it'd be ideal to have a clean CentOS AMI with the current kernels. Karanbir, you seem to have been the one on CentOS side who has had the most conversations with

Like many of you, I like to avoid modifying /etc/httpd/conf/httpd.conf as much as possible. Instead, modifying or adding files in /etc/httpd/conf.d However, because /var/www/html is declared within the main httpd.conf and certain directives are set along with "AllowOverride None" (thus forbidding use of .htaccess), it seems like the only way to override any of these options is to