Discuss

discuss@lists.centos.org

2 participants
34758 discussions

Occasional faults - CentOS server running subversion
by David Fletcher 07 Feb '06

07 Feb '06

Dear All, I've got a small server at work, built from a Via Epia SP 8000E mini-ITX motherboard and a Seagate SATA hard drive. The application of this server is to provide versioned file storage, using subversion. I'm running CentOS, with Apache 2 and SSL to provide controlled access to the subversion repositories. There have been two problems with this machine:- Problem 1 Every once in a while one of the repositories gets corrupted, which produces an error when attempting to use it with tortoisesvn or svn from another Linux box. I then have to go in as root on ssh and do svnadmin recover followed by chown -R apache:apache just to be sure, then it works again. Problem 2 The server had been running continually for maybe 70 days without problems apart from problem 1. A couple of weeks ago there was a kernel update. I rebooted the machine to run the new kernel in case it included security fixes, and at 23:00 on Thursday night it died. When I came in on Friday morning it was sat there with the hard drive permanently lit and not responding. I had to push the reset button to get it going again. /var/log/messages entries from this incident follow:- Feb 2 21:01:01 localhost crond(pam_unix)[1399]: session opened for user root by (uid=0) Feb 2 21:01:01 localhost crond(pam_unix)[1399]: session closed for user root Feb 2 22:01:01 localhost crond(pam_unix)[1403]: session opened for user root by (uid=0) Feb 2 22:01:01 localhost crond(pam_unix)[1403]: session closed for user root Feb 2 23:00:01 localhost crond(pam_unix)[1407]: session opened for user root by (uid=0) Feb 3 09:10:02 localhost syslogd 1.4.1: restart. Feb 3 09:10:02 localhost syslog: syslogd startup succeeded At 23:00 on Thursday crond would have been running a bash script to do a daily dump of the repositories. BTW the server is supplied by an APC UPS so there should be no issues with mains power supply problems. As I'm not running on Intel hardware, does anybody know of any issues with running CentOS on Via motherboards with Eden processors? Thanks. Dave Fletcher

1 0

works best with centos 4+? CPanel or Plesk?
by Dustin Krysak 07 Feb '06

07 Feb '06

Hi there i was wondering if anyone on this list has any comments on CPanel vs Plesk on Centos 4+? Looking for the one that best work with the "out of the box" RPMs for Centos. Thanks in advance! Dustin

8 10

[ANN] Footprint - kickstart management tool
by Dag Wieers 07 Feb '06

07 Feb '06

Hi, I'm planning to create a new tool to simplify creation and management of kickstart-files (as well as automate some of the tasks that come with automated installation). If you know of another mailinglist that fits this subject, or you know of similar projects (because I can imagine there are countless implementations inside companies) please let me know. I'm just in the design phase and i'm interested in what already exists and how it was implemented (the structure/interface/features). Let me first explain what I aim for (and how it is currently in my head). First of all, it should be straight-forward to manage a list of systems (or groups of systems) that have similar configurations. A config-file could be something like: /etc/footprint/system/proxy-servers.conf [main] boot options = panic=10 acpi=off dist = rhel4as #modules = e100 profile = proxy-servers [proxy01] ip = 10.0.0.1 netmask = 255.255.255.0 mac address = 00:01:02:03:004:05 ... (or even more simple, only hold 1 system) There would be a collection of profiles, each specifying a set of variables that influence the kickstart file. This could be hardware-related or functional. And a directory-structure for different methods, like LVM setup for rhel3as, rhel4as for different disk configurations, or a debug-method (to ease development), or a method that sets up callback and log functionality (to follow progress from remote), lots of things are possible here to extend anaconda. The tool would then be able to construct a working kickstart file, depending on the combination of profiles/variables it could generate a custom initial ramdisk, use another kernel, create the PXE and DHCP configuration, or make a floppy or ISO file used for custom hardware booting. Also, it would be nice to have a script in Apache that is able to (based on the IP address or mac address) to generate the correct kickstart file (and automatically save/timestamp it for debug/management purposes). In the end, the command-line tool (based on the config files) would automatically control DHCP/PXE, all the files that make remote booting possible and the management (and revisions) of kickstart files. And that's how I got the name, what is left after a good kickstart ? A footprint :) The aim is to integrate it to work with Yam, so that setting this up becomes very easy. And custom packages and updates are merged with the default packagelist so you don't have to do customisations afterwards. Additionally, I'd like to have methods for doing extra configuration tweaks (like populating hosts files, dropping SSH keys, configuring user management and much much more). For all these things I can use help and ideas, so let me know what additional features you have in mind. Discussions go onto the rpmforge tools mailinglist: http://lists.rpmforge.net/mailman/listinfo/tools And I have a non-functional website at: http://dag.wieers.com/home-made/footprint/ (required for registering the freshmeat project) Kind regards, -- dag wieers, dag(a)wieers.com, http://dag.wieers.com/ -- [all I want is a warm bed and a kind word and unlimited power]

2 1

ftp access directory
by Andrew Rice 07 Feb '06

07 Feb '06

using vsftpd, I know its possible to specify what directory you want a user to be dropped into when he ftps in...but I dont remember where to go to change that. any suggestions? -- Andrew Rice Jr

5 7

How to add System V Filesystem to Centos?
by Mace Eliason 07 Feb '06

07 Feb '06

Hi, I am working on a project where I have to move some parts files from an old Unix server. I have tared the files to floppy disc, but I can't mount them. After doing some reasearch I have found that the SCO Unix ver 3.2 uses a System V filesystem I think. From the article that I have read I need to add this filesystem to my kernel to be able to mount and read the floppy discs. Can anyone help me with this I have never recomipled my kernel before. Then I should be able to use # mount -t sysv /dev/fd0 /mnt/floppy I have found in the .config file under misc filesystems that I can turn on sysv filesytem. I am not sure what to do after that. Is it better to make it a module or to compileit into the kernel? kernel CONFIG_SYSV_FS=y as a modual CONFIG_SYSV_FS=m If I go the modual way where do I find the modual? I would use modprobe to install it right? What commands do you run to recompile the kernel? Do I run make menuconfig? I have been reading alot but not finding what I need. I have a fresh install off Centos 4.2 just for this purpose so if I mess it up that s okay I will start again. I am sure this is so simple once you have done it once. Thanks Scandog

7 23

Install from USB Pendrive
by Chris Hammond 07 Feb '06

07 Feb '06

I would like to take a USB pen drive and make it CentOS 4.2 install cdrom 1. I have found the boot image for USB drives or floppies but I want to make my pen drive Install CD 1. I have looked around and the closest thing I have found is a how to for Debian but I can't seem to figure out how to adapt it to what I want to do. Any help or links would be much appreciated. Thanks Chris

3 4

32bit 3d X apps under CentOS 4 x86_64?
by Steve Bergman 07 Feb '06

07 Feb '06

Is it possible to run 32 bit X applications under CentOS using DRI? I'm finding that the app looks for: /usr/X11R6/lib/modules/dri/r200_dri.so and does not find it. Although /usr/X11R6/lib64/modules/dri/r200_dri.so does exist. And so I end up with software gl rendering. Thanks, Steve

3 3

Oracle and kernel 2.6 vs 2.4 performance question
by Benjamin J. Weiss 07 Feb '06

07 Feb '06

All, We're currently running Oracle 9i on a RHEL AS 2.1 installation. We need to upgrade to Oracle 10g, and I've half-way convinced my boss to upgrade to RHEL 4 (we're still discussing AS vs ES, since we're running 2 HT processors...). Anyway, I want to upgrade because of EOL issues whenever RHEL 5 comes out, but before he okays me wiping the server and installing 4, he wants to know if we're going to get any specific benefits out of upgrading the OS. I recall that the 2.6 kernel did a bunch of optimization stuff, but is there anything in there that would affect the performance of Oracle 10g? And if so, are there any web sites to which I can direct him? So far my googling hasn't been very productive. Maybe I'm using the wrong search terms. Thanks! Ben Weiss

4 3

FreeNX problem
by Kanwar Ranbir Sandhu 06 Feb '06

06 Feb '06

Hello All, I have FreeNX deployed on a CentOS 4 server at a client site. When I log into my Gnome desktop on the client's server and start an app, the window borders aren't visible. The applications themselves act strangely too - when I try to click on a menu item, either nothing happens or the app loses focus. Overall, the window manager looks to be acting up. At first I thought something was wrong with my Gnome profile, so I moved the .gnome* and .gconf* directories out of the way and logged in again. This didn't fix the problem. So, I then backed up my home folder, deleted my account, recreated it, setup the nx keys, and logged in again - same problem. The next course of action was to remove and reinstall the NX client on my laptop. I did this, but the problem remains. Logging into my account on the local network (i.e. at the client site) does not exhibit the same problem. None of my client's employees have reported issues, so it's completely isolated to my account. Everything seems to point to FreeNX/NX as the culprit. There must be some temporary file sitting around that's causing the window manager problems, but I can't find it. Anyone have any idea where I should look next? I'm out of ideas at the moment. Regards, Ranbir -- Kanwar Ranbir Sandhu Linux 2.6.15-1.1830_FC4 i686 GNU/Linux 14:49:33 up 7:49, 3 users, load average: 0.68, 0.52, 0.32

1 0

Relaying of spam
by Marcel 06 Feb '06

06 Feb '06

Hi, sorry if this isn't the right place to post, but I'm having some trouble figuring out a spamming issue. If anyone here can help, that'd be amazing. I'm running Brian's CentOS/BlueQuartz CD, version 3.5 from Nuonce.net. Everything seemed to be running fine for several days until this morning, when I received a zillion "returned mail" notices from the mailer daemon. Within it, it said it was unable to complete sending to the following users for various reasons and blah blah blah. That's fine, but I never initiated the email. In my logs, entries like the following shows up ('portal' is the name of the box obviously): Feb 5 12:11:45 portal sendmail[17135]: k15EXFZf015093: SMTP outgoing connect on portal.xxxxxxx.com Feb 5 12:12:51 portal sendmail[17135]: k15EXFZf015093: makeconnection (mobilemail.caii-dc.com. [209.135.227.253]) failed: Connection timed out with mobilemail.caii-dc.com. Feb 5 12:12:51 portal sendmail[17135]: k15EXFZf015093: to=<aldara(a)caii-dc.com>, ctladdr=<username(a)portal.xxxxxxxxxxxxxxxxxxxx.com> (502/100), delay=03:39:35, xdelay=00:01:06, mailer=esmtp, pri=3188891, relay=mobilemail.caii-dc.com. [209.135.227.253], dsn=4.0.0, stat=Deferred: Connection timed out with mobilemail.caii-dc.com. Irregardless of the errors, I can't figure out why/where the outbound email is being generated. There are many entries in the log like this, and I assume alot of it, is going through. The user never initiated it. It has to be the server itself? Plus, it's using the full name of the server which is portal.domainname.com in the email address. It seems to only use ONE user's name though. AND it's ONLY using 1 user's name from a list of several. The user account gets some spam every now and then with the following header info, then these returned emails. These emails are from the local server using an account that doesn't exist: =============================== Subject: The hottest issue we've seen this year From: ThePickOfTheYear2696(a)domainname.com Date: Sun, 5 Feb 2006 08:52:47 -0600 To: ThePickOfTheYear2696(a)portal.domainname.com =============================== Since the "pickoftheyear" account doesn't exist.... Is there any suggestions from the group? I'm a newb at running a mail server, just trying to figure out what's going on. The site in question did have a couple formmail scripts that I deleted. I am interested in running chkrootkit but is there a specific package required for CentOS/BQ? Or just download and compile? Thanks. M

6 8

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss