Discuss

discuss@lists.centos.org

3 participants
34748 discussions

Re: [CentOS] Entries in /var/log/messages
by Jeffrey Means 22 Aug '05

22 Aug '05

CentOS mailing list <centos(a)centos.org> wrote: > On Sun, 2005-08-21 at 21:09, Jeffrey Means wrote: > > One other method I have sucessfully used / am using is to change the port number > > of the service being attacked. If we are talking about ssh this can be done in > > the /etc/ssh/sshd_config file by changing / adding a Port xxxx line to the file. > > > > I hope this helps you it has drastically decreased the number of people trying > > to break down my front door. > > --Jeff Means > > MeansPC - Custom Web Development for your needs. > > > > CentOS mailing list <centos(a)centos.org> wrote: > > > On Sun, 2005-08-21 at 17:03 -0500, Jerry Geis wrote: > > > > I have quite a few entries in /var/log/messages for connection attempts. > > > > Is there anything other > > > > than ignoring them I can do? Example is below. > > > > > > > > > > There are a number of scripts (some Perl, some Python) out there to > > > monitor the log and add an entry in hosts.deny to block any further > > > attempts from the offending IP when too many failed password attempts > > > are noted. You can find them with some "googling". > > > > > > I am using a modified one to stop these breakin attempts on my servers. > > > > > > > Aug 21 15:48:19 machine sshd(pam_unix)[17903]: check pass; user unknown > > > > Aug 21 15:48:19 machine sshd(pam_unix)[17903]: authentication failure; > > > > logname= uid=0 euid=0 tty=ssh ruser= > > > > rhost=wsip-24-234-149-156.lv.lv.cox.net > > > > > It is good to know that this type of attack against ssh is generally > automated. Most likely run by script kiddies looking for a system with > poor passwords or default passwords on that service. > > If you take the actions others have already posted you should be in good > shape. Just make sure you use non-trivial passwords, limit which users > are allowed to login into ssh, and if you want to eliminate this type of > traffic in your log files use a different port. It is important to > realize that changing the port number is not a security measure. Any > good hacker will scan your system and find it. But it does prevent > these automated scripts from finding your system for the most part. > > > _______________________________________________ > CentOS mailing list > CentOS(a)centos.org > http://lists.centos.org/mailman/listinfo/centos > This is a very good point as a rule I only allow PKCS (public key) logins from offsite so I have little to worry about this particular issue, again this is a line item chage you can make in the sshd_config file. I am running two sshd servers one for internal access wich does run on port 22 but is firewalled from the outside and one on port xxxx wich only allows connections with PKCS auth. If you want a copy of my sshd startup script wich starts daemons for all the files ending in .config contained in the /etc/ssh/ directory let me know and I'll gladly make that available. Using PKCS it makes it really hard to break a password that is over 256 characters, has builtin validity tests and travels nicely on a usb key (even old small ones which makes this very inexpensive to implement even for a medium to large business.) Note to the wise: Be sure to password protect the keys issued for if they fall into the wrong hands you just gave away the keys to your server / network. -- -- Jeffrey D. Means meaje(a)meanspc.com Owner / CIO for MeansPC http://www.meanspc.com/ Custom Web Development For Your Needs. (970)308-1298 - The stupidity of a stupid person is exercised in a restricted field; the stupidity of an intelligent individual has a much broader diffusion, and far greater effect, aided as it is by the element of surprise.

1 0

nfs4 on CentSO4.1
by Tony 22 Aug '05

22 Aug '05

I'm struggling to get nfs v4 going at the moment- following instructions from http://www.brennan.id.au/19-Network_File_System.html#nfs4 Can anyone else confirm that they already have nfs4 working ok, just to give me heart to continue fighting? -- Cheers, Tony

1 0

Kernel or kernel-smp ?
by Robert 21 Aug '05

21 Aug '05

Since this is National Kernel Day, I have a question. No, 2 questions: I'm running an Asus A7N8X Deluxe ver 2 m/b with an AMD processor: [root@mavis ~]# cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 6 model : 10 model name : AMD Athlon(tm) XP 2600+ stepping : 0 cpu MHz : 1912.933 cache size : 512 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov patpse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow bogomips : 3776.51 For some reason, Anaconda thinks I need an smp kernel. This is a wrong-headed notion that showed up in FC1 and continued to FC2, FC3 and CentOS 4. It always installs both kernels, making -smp the default which I have to change to non-smp for ntpd to work right (Gives off-the-chart jitter, never syncs, etc). I have read in one place or another that: 1. It's O.K. to run an smp kernel on a single-processor machine 2. The installer picks the smp kernel if the cpu flag "ht" is set -- which mine isn't. So, can anyone explain why, on a fresh bare-metal install I'm blessed with an smp kernel? Also, is the statement about an smp kernel running O.K. on a single processor machine pure hogwash or is there something goofy about my m/b and/or processor? Of course, once there is an -smp kernel installed "yum update kernel*" keeps the string of luck going. -- It is Sun Aug 21 09:48:33 CDT 2005 ratnow. The next Next-Step will begin in 2526087 seconds at 3:30 P.M. on Monday 9/19/2005 at the Golden Gristle ratcheer: http://snipurl.com/ag20

2 2

The scariest set up of all: Printer Support
by Dave Gutteridge 21 Aug '05

21 Aug '05

My experience with printers is that of all the peripherals, regardless of OS, they are the trickiest to install, configure, and use. So I'm expecting a battle here. What I have is a Canon Pixus iP3100. It's a Japanese model, but it's the same model as the US Canon Pixma iP3000, just a different name. First, I plugged in the USB cable. CentOS seemed to detect it, and gave me a list of options. Neither the US model number nor Japanese model number was there, so I selected "Canon (unknown)". It spent a while installing, and then finished without any message to say for sure if it had installed successfully. I tried to print from a number of application, but nothing happened. Just nothing. No error messages, no printer noises, nothing. So, I figured the driver didn't take hold. I looked around the net, and first I cam across "Turboprint for Linux". At first it seemed promising. But my first snag was that after installing the RPM, I didn't seem to have any printer configuration dialogues or anything. I went on the net to see about a manual, and *then* I discovered that in the free, consumer use version, they stamp a logo onto whatever you print. I was a little annoyed that they didn't make that really clear up front, they just tell you that commercial users must pay and consumers can use for free. Anyway, I decided to look for different alternatives. I came across a web site that said that my printer model could work under a different Canon driver: http://www.linuxprinting.org/pipermail/canon-list/2004q4/001797.html But, and here's the heart of my question, I can't find any place to make changes to my printer configuration. According to some web sites I looked at, I should have a printer configuration application in my menu, under "Preferences". But all I have is a "printer settings" application which, despite the promising name, doesn't have any menu options at all for configuring the printer or it's drivers. It looks more like a print spool manager for when you have print jobs lined up. Where/how do I configure my printer and try alternate drivers? Dave

11 49

Yum still won't behave.
by Dave Gutteridge 21 Aug '05

21 Aug '05

Before I reported that YUM often crashed. I was given help here on this list which stopped YUM from freezing my computer. However, YUM is still not behaving. Recently, I went to install a program called transcode. Yum found 14 packages that needed to be installed. It installed one of them and then stopped. It gave me this error: /sbin/ldconfig: File /usr/lib/libk3bdevice.so.2.0.0.#prelink#.X8kEMh is empty, not checked. So, I went to /etc/sysconfig/prelink and set PRELINKING=no, because I was told that this would stop that error message from coming up. I went back to YUM to isntall the application, it found the remaining 13 packages to install, got as far as installing one, then the same error message came up and it stopped again. I rechecked that I had properly edted the prelink file, and tried again. This time, it found the remaining 12 files to install, gave no error message, but nonetheless it stopped after installing one package. And so on. Eventually to install the application I had to run YUM 14 times to get all 14 packages. What could be wrong with my isntallation of YUM? Dave

1 0

Are these steps for re-installing Gnome correct/safe?
by Dave Gutteridge 21 Aug '05

21 Aug '05

My desktop environment has become a little chaotic, and, long story short, I would like to uninstall KDE and Gnome, and then reinstall Gnome. (I want to remove KDE for the time being because I just don't think I want it taking up space on my hard drive). The thing is, I don't want to throw the baby out with the bathwater. For instance, I want to make sure I don't lose settings in X, like my xorg.conf file. I just want to reset my Gnome interface back to how it was when I first installed it. I'm thinking the steps I need to take are roughtly these, but it would be great if someone could verify that these will be a non-destructive way of getting Gnome back to it's original state: 1. Stop X (By running telinit 3 at a command prompt?) 2. Use YUM to remove Gnome and KDE (#yum remove gnome kde?) 3. Use YUM to install Gnome (#yum install gnome?) 4. restart X (#telinit 5?) Could someone clarify if I have the commands right, and if this process is safe to undertake? Dave

8 23

trying to recompile
by Andrew Zahn 21 Aug '05

21 Aug '05

Hi All, I am trying to recompile the stock kernel to include XFS. At install time of the OS I select to install the kernel source code however I do not have a /usr/src/linux-2.6.9-5.0.3.EL-smp directory. I then learned you must run "yum install kernel-sourcecode" to properly install the source. When I run the command this is what happens ( see below) . The error I get is: warning: rpmts_HdrFromFdno: V3 DSA signature: NOKEY, key ID 443e1821 public key not available for qt-devel-3.3.3-9.3.i386.rpm does anyone know what to do at this point? Thanks, Andrew [root@se1 kernels]# yum install kernel-sourcecode Setting up Install Process Setting up Repos update 100% |=========================| 951 B 00:00 base 100% |=========================| 1.1 kB 00:00 addons 100% |=========================| 951 B 00:00 extras 100% |=========================| 1.1 kB 00:00 Reading repository metadata in from local files update : ################################################## 109/109 base : ################################################## 1406/1406 extras : ################################################## 32/32 Parsing package install arguments Resolving Dependencies --> Populating transaction set with selected packages. Please wait. ---> Package kernel-sourcecode.noarch 0:2.6.9-11.EL set to be updated --> Running transaction check --> Processing Dependency: qt-devel for package: kernel-sourcecode --> Processing Dependency: gtk2-devel for package: kernel-sourcecode --> Restarting Dependency Resolution with new changes. --> Populating transaction set with selected packages. Please wait. ---> Package qt-devel.i386 1:3.3.3-9.3 set to be updated ---> Package gtk2-devel.i386 0:2.4.13-14 set to be updated --> Running transaction check --> Processing Dependency: atk-devel >= 1.6.0-1 for package: gtk2-devel --> Processing Dependency: pango-devel >= 1.4.0-1 for package: gtk2-devel --> Restarting Dependency Resolution with new changes. --> Populating transaction set with selected packages. Please wait. ---> Package pango-devel.i386 0:1.6.0-9 set to be updated ---> Package atk-devel.i386 0:1.8.0-2 set to be updated --> Running transaction check Dependencies Resolved Transaction Listing: Install: kernel-sourcecode.noarch 0:2.6.9-11.EL - base Performing the following to resolve dependencies: Install: atk-devel.i386 0:1.8.0-2 - base Install: gtk2-devel.i386 0:2.4.13-14 - base Install: pango-devel.i386 0:1.6.0-9 - base Install: qt-devel.i386 1:3.3.3-9.3 - base Total download size: 60 M Is this ok [y/N]: y Downloading Packages: warning: rpmts_HdrFromFdno: V3 DSA signature: NOKEY, key ID 443e1821 public key not available for qt-devel-3.3.3-9.3.i386.rpm

3 5

Kicker Bug in KDE 3.4.2
by duffmckagan 21 Aug '05

21 Aug '05

I know that this is not a Cent OS centric question, but still. I recently upgraded to KDE 3.4.2 on Cent OS 4. Does anybody knows the solution for the Kicker Bug? Whenever I Log-Off, I get an Error saying that Kicker has crashed. -- "No-one dies a virgin. Life screws everyone."

1 0

YUM and RPM often freezes my computer
by Dave Gutteridge 21 Aug '05

21 Aug '05

Sometimes when I attempt a YUM install, it will get to the point where it is looking for packages to install, and my hard drive will make a lot of noise, but there is no progress. My mouse cursor freezes, and I can not switch between programs or kill the process. Eventually I have to reset the computer in order to get control back. The same thing will also happen when I install an RPM. When it checks for dependancies, it will freeze up my computer in the same way. This happens about 50% of the time. However, after resetting and rebooting the computer, if I attempt the same install again, it will work fine. Also, this may or may not be related, but in the case of YUM, when it's installing multiple RPMS, sometimes it will install only some of them, and then stop, without any explanation. If I repeat the install, then it will install the remaining packages, and complete the process. Any possible explanations for this behaviour? Dave

3 6

Poprelayd parser released for Dovecot on RHEL 4
by dan1 20 Aug '05

20 Aug '05

Hi all. For those of you who are interested by using Poprelayd, I just made a fix today for using Dovecot and sendmail with poprelayd. Poprelayd allows you to log the IP addresses of people logging in through POP and allowing them to send e-mails using your smtp server (allowing relaying) for 15 minutes. It is very handy. You can download the parser here: http://sourceforge.net/forum/forum.php?forum_id=21494 It is just a little info that might be useful for some people. Kind regards, Daniel

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss