Discuss June 2008

discuss@lists.centos.org

353 participants
415 discussions

Re: [CentOS] sendmail not sending after reboot UPDATE
by Ray Leventhal 09 Jun '08

09 Jun '08

<snip> > > where does the server point to for its dns? if it's (supposed to be) > running a nameserver (e.g., you're pointing to 127.0.0.1) but the dns > server isn't running you'll get a timeout. if you have a second > nameserver configured in /etc/resolv.conf you'll likely ultimately > get a response, but only after you time out on the first. > > by the way, you may want to look at the server's "client" sendmail > queue ("mailq -Ac"). the remote clients dump to the regular queue > (/var/spool/mqueue), but localhost mail goes into the clientmqueue > (/var/spool/clientmqueue). if mail is hanging there it might give > hints that are easier to debug than are remote connections. [you may > also want to try originating mail from localhost and see what > happens.] > > > > - Rick > > The server points to another server on the network for dns, which is running and accepting queries on 53, as expected. resolv.conf points to the correct nameserver IP. I'll check the client queue..hadn't thought of that. Thanks, -Ray

1 0

Re: [CentOS] sendmail not sending after reboot UPDATE
by Ray Leventhal 09 Jun '08

09 Jun '08

lists-centos wrote: > are you configured for spamassassin or clamav as milters on the > server (in the sendmail.cf) and either/both of these didn't start up > on the reboot? sendmail will still accept mail with sa/clamav not > running, but it takes a bit longer. > there's also the possibility of an inverse-map lookup on the > ipnumbers of the inbound client machines. if the server is trying to > do the lookups but is having trouble reaching the appropriate > in-addr.arpa server you'll get long hangs. > > - Rick > Hi Rick, no milters are configured...this is a really small intranet with absolutely no connectivity to the outside world. I'm baffled. Regarding your mention of a dns issue...it may be, but i'm still baffled as this was all working before the weekend. I rebooted the box on Sunday and am in this <pun intended> bind today :) Thanks for your input...much appreciated. -R

1 0

new install of 5.1 with KVM-over-IP - can't install with GUI - need assistance
by Alon 09 Jun '08

09 Jun '08

Hi All, I have several dozens of CentOS and WhiteBox servers. Most of them are CentOS 4.6. Our installation service is done in the datacenter where the servers are located. When we install a fresh clean install, we use the GUI menus, while using the KVMoverIP. That was working great with CentOS 4.x In CentOS 5.x, the installation process 'annonces" that "Hey,. I know you are using a KVM and don't have a monitor attached" (who cares??). But, "since you are using a KVM and no monitor is attached,. you CANNOT use the GUI installation". Why??? 4.x didn't care for this. Why make my life difficult? I have to 'cheat' by calling the datacenter to plug a monitor for the first minute of the install and then plug back the KVM cable. Isn't this stupid? of course it is! I am looking for a solution such as a parameter that I can pass to the boot sequence (vga=nommconf or something like that) so that it won't do the probing for the VGA and just let me go about my business. Mind you that this is a problem both with DELL PowerEdge 1950iii with DRAC5 remote consule as well as with just plain PCs that use ATEN KVM 9116. Once the probe understand that we don't use a monitor it prompts the 'can't use the GUI'. Any pointers about this? Thanks, -Sup.

5 7

RE: [CentOS] Copying files from specific date.
by Rajeev R. Veedu 09 Jun '08

09 Jun '08

-----Original Message----- From: centos-bounces(a)centos.org [mailto:centos-bounces@centos.org] On Behalf Of Steve Huff Sent: Monday, June 09, 2008 7:34 PM To: centos(a)centos.org Subject: Re: [CentOS] Copying files from specific date. On Jun 9, 2008, at 11:16 AM, Rajeev R. Veedu wrote: > Does anyone aware of any utility to copy files which are created > or modify form a specific date ?. > >>to copy all files in /dir1 modified within the last 5 days to /dir2: >>$ find /dir1 -mtime -5 | xargs -I {} cp {} /dir2 >>if the filenames have whitespace in them, you can use this trick: >>$ find /dir1 -mtime -5 -print0 | xargs -0 -I {} cp {} /dir2 >>for more details on selecting by time: >>$ man find >>pay particular attention to the options -atime, -amin, -ctime, -cmin, >>-mtime, -mmin, and -daystart. >>-steve -- Actually I need to copy this on to another server with same folder structure. I think I need to explain bit of history. I had a server crash last week, and we have restored the files from the tape. However during this period of making the server up, the users having adding or changed files from our backup Server (Samba server which rsync to production server every night.) now I need to copy the files which user added/ modify last 7 days. Ideally if I can get this option in rsync it would be better. Otherwise I need to have a method so that all changed files to go on the relevant folder on the production server. I cannot take the full files in the backup files since they are historical backup and there are some unwanted files. Can I use scp instead of cp in your statement?. But how does it take the same directory name as the original location? Eg:from ServerA/FLDR2/FLDR3/Filename should go to ServerB/FLDR2/FLDR3/FILENAME Only change is the server name all other values will remain same. Any help would be really appreciated. Thanks Rajeev

1 0

Copying files from specific date.
by Rajeev R. Veedu 09 Jun '08

09 Jun '08

Dear all, Does anyone aware of any utility to copy files which are created or modify form a specific date ?. Thanks Rajeev R. Veedu

3 2

Re: [CentOS] Dell 1950 Perc 6/i with centos 4.5
by Steve Moccio 09 Jun '08

09 Jun '08

Jeff, I cant thank you enough. I not only did this for the kernel but I did this for the pxeboot's jumpstart installation of 4.5. Thanks again for taking the time to respond. ---- On Sat, 7 Jun 2008, Jeff Collins (jeff(a)chpscchome.org) wrote: I had a somewhat similar problem with a Dell 2600 running Centos 4.4. I found that 4.6 no longer had the megaraid driver for my card, so I:Here is a detailed description of what I did, and perhaps this may be of help if you adapt it to your situation.============================================================================================== This is a brief description of how to set up the 4.6 kernel so that it will boot on list.The reason that the stock CentOS 4.6 kernel will not boot on list is that LSI and Dell madethe decision to remove support for the PERC/LSI controller adapter shipped with the PowerEdge2600 from the 4.6 kernel. The new megaraid subsystem supports the PERC/LSIcontrollers shipped with the newer PowerEdge 2900 models. The stock 4.5 and 4.6 kernelswill not boot on an older 2600.The 4.4 kernel is called 2.6.9-42.EL or 2.6.9-42.ELsmp in the symmetric multiprocessor version.The 4.6 kernel that first shipped is called 2.6.9-67. The current topof tree is 2.6.9-67.0.15 as of May 13 2008). When we compare the megaraid_sas.cdrivers (for example), we see that a lot of generic support has been taken out of the newerversion, compared with the driver from 4.4 (2.6.9-42.EL).For a full scan of the differences between all source files in the megaraid directory, seenewdifftxt in this directory.So what we have to do is take out a top of tree 4.6 kernel, wipe out all the code in themegaraid directory, put back all the old megaraid code, then build a kernel. To get theold source kernel, use www.rpmfind.org to search for the proper kernel:kernel-2.6.9-42.EL.src.rpm. The current top of tree source magically appears in/usr/src/redhat/BUILD/kernel-2.6.9/linux-2.6.9 when we performyum install kernel* and follow the instructions to build a custom kernel.This is not trivial, but all the instructions are athttp://wiki.centos.org/HowTos/Custom_KernelThe basic steps I took were slightly different:-----------------------------------------------1) Download the stock 4.4 kernel source using rpmfind2) Build the stock 4.4 kernel, install, and test boot it to validate the process.3) Save off the megaraid directory code which is probably in/usr/src/redhat/BUILD/kernel-2.6.9/linux-2.6.9/drivers/scsi/megaraid4. Change the name of the prior kernel to something like/usr/src/redhat/BUILD.old, etcSo you get [root@list scsi]# ls /usr/src/redhat/BUILD BUILD.old RPMS SOURCES SOURCES.old SPECS SPECS.old SRPMS5. Download the top of tree kernel sources for the version we are on (Centos 4)6. Blast the contents of/usr/src/redhat/BUILD/kernel-2.6.9/linux-2.6.9/drivers/scsi/megaraidand replace with the old megaraid code you saved.7. Build and test the kernel (The instructions tell us to use rpmbuild to create a full rpmpackage, but I took the shortcut of just making a stock kernel from the top of theBUILD/kernel-2.6.9/linux-2.6.9 directory. The steps area. make oldconfigb. make menuconfig (you can take out extra modules if you wish, but make sureyou know what you are doing. It is easy to create a kernel that does not boot)c. make -j 4 bzImaged. cp ./arch/i386/boot/bzImage /boot/vmlinuz-2.6.9-your-versione. cp System.map /boot/System.map-2.6.9-your-versionf. make -j 4 modulesg. make modules_installh. mkinitrd /boot/mkinitrd-2.6.9-your-version.img 2.6.9-your-versioni. edit /etc/grub.conf to add your kerneli. sync; rebootj. Choose your kernel (do not make it the default until it has been tested!)k. Test your kernelSo on list we end up with four Linux kernels:The first is the old smp kernel first shipped as CentOS 4.4.The second is a test kernel for use in creating a custom install CD set.The third is the patched 4.6 kernel we are running.The fourth is the uniprocessor stock kernel shipped as CentOS 4.4.Since grub counting is 0 based, we have set the default to 2 so we boot thethird kernel by default (2.6.9-jgc).Here is the grub.conf file in /etc===========================================================================# grub.conf generated by anaconda## Note that you do not have to rerun grub after making changes to this file# NOTICE: You have a /boot partition. This means that# all kernel and initrd paths are relative to /boot/, eg.# root (hd0,1)# kernel /vmlinuz-version ro root=/dev/sda6# initrd /initrd-version.img#boot=/dev/sda2default=2timeout=5splashimage=(hd0,1)/grub/splash.xpm.gzhiddenmenutitle CentOS-4 i386 (2.6.9-42.ELsmp) root (hd0,1) kernel /vmlinuz-2.6.9-42.ELsmp ro root=LABEL=/ rhgb quiet initrd /initrd-2.6.9-42.ELsmp.imgtitle CentOS-4 i386 (2.6.9-prep) root (hd0,1) kernel /vmlinuz-2.6.9-prep ro root=LABEL=/ rhgb quiet initrd /initrd-2.6.9-prepimgtitle CentOS-4 i386 (2.6.9-jgc) root (hd0,1) kernel /vmlinuz-2.6.9-jgc ro root=LABEL=/ rhgb quiet initrd /initrd-2.6.9-jgc.imgtitle CentOS-4 i386-up (2.6.9-42.EL) root (hd0,1) kernel /vmlinuz-2.6.9-42.EL ro root=LABEL=/ rhgb quiet initrd /initrd-2.6.9-42.EL.imgtitle dos rootnoverify (hd0,0) chainloader +1 ============================================================================Perhaps this will be of some help to you, Jeff CollinsCommunity Health Partnership of Santa Clara County

1 0

Re: RE-export nfs mounted share
by David G. Miller 09 Jun '08

09 Jun '08

"whoami i" <hicheerup(a)gmail.com> wrote: > Hi Barry, > > First of all thanx for your reply but i already used the options > crossmnt,fsid=0 in my exports file still i am not able to re-export it. > > > Regards > lingu > > On Fri, May 30, 2008 at 5:30 PM, <centos-request(a)centos.org> wrote: > > >> > Send CentOS mailing list submissions to >> > centos(a)centos.org >> > >> > To subscribe or unsubscribe via the World Wide Web, visit >> > http://lists.centos.org/mailman/listinfo/centos >> > or, via email, send a message with subject or body 'help' to >> > centos-request(a)centos.org >> > >> > You can reach the person managing the list at >> > centos-owner(a)centos.org >> > >> > When replying, please edit your Subject line so it is more specific >> > than "Re: Contents of CentOS digest..." >> > >> > >> > Today's Topics: >> > >> > 1. CentOS-announce Digest, Vol 39, Issue 13 >> > (centos-announce-request(a)centos.org) >> > 2. Re: Low-memory Centos5? (Anne Wilson) >> > 3. Re: Low-memory Centos5? (Wojtek Pilorz) I think we just found another argument as to why top-posting is a bad idea... You can easily miss the fact that you included the whole digest and not just the message you're actually replying to. Cheers, Dave -- Politics, n. Strife of interests masquerading as a contest of principles. -- Ambrose Bierce

1 0

OT: really dumb question about APC UPS
by Jamie Lists 09 Jun '08

09 Jun '08

We have about 4 servers connected to an APC UPS. The UPS only has one UPS connection. So my question is, in a power outage how do i get all my servers to safely shutdown? Do i run a USB hub? Do i write a script on the one machine connected to turn off all the other machines? I'm confused. Jamie

5 4

Chroot'ed SSH
by Filipe Brandenburger 09 Jun '08

09 Jun '08

Hi, Is anyone chrooting users that connect through SSH? I looked for it on Google and I basically saw several methods: - OpenSSH 5 supports ChrootDirectory (FC9 apparently has RPMs that probably could be rebuilt under CentOS 5) - There seem to be several patches for OpenSSH 4.x to do the chroot, the most popular seems to be http://chrootssh.sf.net/ - There appears to be a pam_chroot - There are solutions based on setting the user's shell to a script/binary that does the chroot By quickly looking at yum list, it doesn't seem like neither RHEL nor CentOS directly support any of those, at least I didn't find any RPMs for any of those. If anyone is doing it, I would like to know what were your experiences and if you would recommend doing it or not. I'm specially interested in anything that doesn't involve replacing the OpenSSH that comes with CentOS, after all, that's what CentOS is all about, if you start replacing the pieces, what's the point... Thanks a lot! Filipe

3 3

Hardening CentOS by removing "hacker" tools
by Filipe Brandenburger 09 Jun '08

09 Jun '08

Hi, My boss asked me to harden a CentOS box by removing "hacker" tools, such as nmap, tcpdump, nc (netcat), telnet, etc. I would like to know which list of packages would you remove from a base install. I would appreciate if someone could point me to a "standard" way of doing this. I know there are procedures for hardening a machine (I remember reading about Bastille Linux) but I don't know how effective they are and if they include the removal of such tools in their procedures. Any advice would be very appreciated! Thanks, Filipe

13 17

← Newer
1
...
30
31
32
33
34
35
36
...
42
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss June 2008