Discuss

discuss@lists.centos.org

34753 discussions

GFS & Vsftpd? maybe even Iscsi?
by Karl R. Balsmeier 17 May '07

17 May '07

Looking to cluster an FTP server. Right now we do a mix of stuff where we use a few separate servers linked through various hacks, that utilizes VSFTPd heavily for hundreds of television station news websites, where, of course, the current weather maps get uploaded from everywhere, all day, from NOAA and the like. I'd like to move from the current setup to something that's more HA oriented, -is GFS and option in Centos and is it viable to make an FTP server setup like this? (please no responses about SFTP, I use that too on other servers and yes, prefer it even if it slows things down). Hardware is not an issue, just curious about what some of you sage folk might have experimented with, and possibly have put into production. A google search turned up relatively empty, so hopefully this isn't one of those questions that gets a collective "huh?". Essentially am just trying to get away from the many separate machines I use now where I have to clone entries from /etc/passwd and /etc/shadow all over the place. Bonus: I have an ISCSI NAS device/promiseraid, so perhaps there's an option there instead of doing GFS I could grab a slice of space that way and just have one server with a huge amount of space? It doesn't answer point of failure totally, but... -krb

2 1

Disk accounting with LVM Snapshots
by Mike Fedyk 17 May '07

17 May '07

Hi, I was reading a lot of data from one LVM logical volume and wanted to watch the disk statistics with iostat today. To my surprise, I was seeing the same activity on two logical volumes (LV) and the physical disk containing the two. A little investigation showed that one LV was a snapshot of the other, and I was reading from one of the snapshots. While I understand that the two LVs share some common Logical Extents (LE), I believe it is a bug that the disk statistics show the activity on both logical volumes. Should I report this in the centos bugzilla or upstream's bts? Thanks, Mike Fedyk

1 1

Using Xen under CentOS 5.0
by Ioannis Vranos 17 May '07

17 May '07

OS: CentOS 5.0 x86. I have an internal modem with a driver that works with CentOS 4.x but not with 5.0. So I am thinking of using Xen in conjunction to SL 4.x. My question is, does this require separate physical partitions for SL 4.x, or will these be part of a VM file or something with CentOS 5.0 as a host?

4 9

Squirrelmail Problem on CentOS 4.4
by John Hinton 17 May '07

17 May '07

Recently, some of my accounts cannot read any email in the INBOX within Squirrelmail. The error.... The server couldn't find the message you requested. Most probably your message list was out of date and the message has been moved away or deleted (perhaps by another program accessing the same mailbox). The messages are not being moved out. In fact, I can delete one of those messages just fine. I can also move messages to another folder. Oddly, once I move a message I can open and read it.. and I can open and read any message in the Trash folder. So, it seems that something has changed during an update to who knows what, which has borked up reading mail in Squirrelmail. It does seem that this is occurring so far on what would amount to root users of an account. In other words, the user created during a hosting installation. So, files are stored in /var/www/rootusername. When I add additional email users to an account, I put their user files in /var/www/homes/rootuser/additionalusername. Has anybody else experienced this? I'm a bit lost as to where to look. I am using Sendmail/Dovecot and have been on this server since CentOS 4 came out. This could be any of so many things from PHP to IMAP, but mostly it seems to be some pathing problem.. just not sure where to begin. The only thing I am finding in apache error logs is this: PHP Notice: Undefined variable: charset_converted in /usr/share/squirrelmail/functions/mime.php on line 316, referer: http://domain_name_replaced_here_by_john.com/webmail/src/right_main.php?PG_… Which doesn't seem to have any affect on other users and looks like it refers to the folder list and not the email reader. Thanks for any hints/help on this. Best, John Hinton

2 3

samba + mysql
by David Hrbáč 17 May '07

17 May '07

Hi, what is the preferred way for Samba password mysql backend? samba-bdb[-mysql] or pam_mysql? Thanks, David

1 0

RE: [CentOS] Getting CentOS Server CD w/kickstart config
by Plant, Dean 17 May '07

17 May '07

Devin Atencio wrote: > I would like to how how I can take the CentOS 4.4 Server Cd for > example and modify the > ISO to put on my anaconda-ks.cfg file and then modify the kernel > parameters to automatically > load the anaconda-ks.cfg file so that I can create a CD that I just > stick into my machine it will > format the drive, install the basic configuration I like without > asking any questions. I need to > install like 20 machines and I wanted to basically just stick a CD in > the drive and let it do it's thing. > > Any help on how to accomplish this would be appreciated. > Here's how I did it for RHEL CD's -------------------------------------------------- Construct a single DVD from the four RHEL CDRom's Insert & mount CD1 # mkdir /DVD # mount /dev/cdrom /media/cdrecorder Copy cd1 to /DVD # cd /media/cdrecorder # tar cf - . | (cd /DVD ; tar xvf - .) # cd / # eject cdrom Carry out the above action for CD 2-4 After all CD's have been copied. Correct the .discinfo file to reference all four disks. # cd /DVD # vi .discinfo Ensure line 4 has 1,2,3,4 A working .discinfo should be as follows. # cat .discinfo 1141417589.425945 Red Hat Enterprise Linux 4 i386 1,2,3,4 RedHat/base RedHat/RPMS RedHat/pixmaps Create the bootable ISO file. (all on one line) # mkisofs -A "RHEL/4.3 i386 3AS" -o DVD.iso -b isolinux/isolinux.bin -c isolinux/boot.cat -no-emul-boot -boot-load-size 4 -boot-info-table -R -J -v -T /DVD Burn the DVD.iso as an image not a file. References http://www.redhat.com/docs/manuals/enterprise/RHEL-4-Manual/x8664-multi- install-guide/s1-steps-install-cdrom.html Modify the DVD to include custom kickstart scripts. Create the /DVD tree as per the previous steps. The default files read on boot up of an RH install CDRom are located in /isolinux/ To run custom kickstart scripts we need to modify the /isolinux/isolinux.cfg and add the kickstart file to the boot process. See here for kickstart script generation. For each kickstart install we need to add 3 lines to the isolinux.cfg file. A label line - This is the identifier you type at the linux boot prompt A kernel line - This does not change A append line - This is where we tell the boot process to use a kickstart file. Firstly backup the original file. # cp /DVD/isolinux/isolinux.cfg /DVD/isolinux/isolinux.cfg.orig Now edit and add our custom kickstart file. # vi /DVD/isolinux/isolinux.cfg Add label myboot1 kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks.cfg This will enable you to type myboot1 at the boot prompt which will call the ks.cfg located in the / of the DVD disk. Multiple kickstart files can be added by using different label and append lines. For Example: label myboot1 kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks1.cfg label myboot2 kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks2.cfg -------------------------------------------------- Have fun. Dean

1 0

Re: [CentOS] Make Raid1 2nd disk bootable?
by David G. Miller 17 May '07

17 May '07

Les Mikesell <lesmikesell(a)gmail.com> wrote: > David G. Miller wrote: >> > >> >>> >> Grub isn't so much the issue here as the difference in the rescue mode >>> >> boot. I'm used to being able to boot the CD, chroot into the existing >>> >> system and have pretty much normal access regardless of what was >>> >> broken. Now that the system /dev directory is basically empty, things >>> >> don't work when you have to mount the partitions manually. Is there >>> >> a step to set up devices so the chroot will work? >>> >> > >> > I haven't tried this but were you looking at /dev *after* you did the >> > chroot? It sort of makes sense to me that the running kernel would only >> > populate it's /dev, not the chrooted /dev. Rescue mounts at least the >> > root partition under /mnt/sys (or something like that). Before you >> > chroot, you should probably also mount any other partitions you want >> > under /mnt/sys and then chroot. >> > > It makes sense for the boot code, but not for me afterwards... The real > problem is that the rescue mode startup doesn't mount RAID1 partitions > at all (this seems like a bug). The startup code might populate the > chroot'ed /dev if it had done the mount - I just haven't found how to do > it myself after doing the mount by hand. Thus there's no /dev/sda or > /dev/sdb as you'd expect when you want to install grub. Let me try explaining this a different way. After you let the rescue software mount the partition, *don't* chroot. It's been a long time since I had to mess with booting in rescue mode but lets assume your old root partition gets mounted under /mnt/sysimage and you told the rescue software to go ahead and mount it. If your /boot is a separate partition, you'll need to mount it manually. Assuming you keep everything where it belongs, mount it under /mnt/sysimage/boot. Then run grub by doing: /mnt/sysimage/sbin/grub --config-file=/mnt/sysimage/boot/grub/grub.conf This way grub "sees" /dev for the running system but uses your real grub config file. You may want to confirm that grub.conf has what you want in it. Cheers, Dave -- Politics, n. Strife of interests masquerading as a contest of principles. -- Ambrose Bierce

2 1

Getting CentOS Server CD w/kickstart config
by Devin Atencio 17 May '07

17 May '07

I would like to how how I can take the CentOS 4.4 Server Cd for example and modify the ISO to put on my anaconda-ks.cfg file and then modify the kernel parameters to automatically load the anaconda-ks.cfg file so that I can create a CD that I just stick into my machine it will format the drive, install the basic configuration I like without asking any questions. I need to install like 20 machines and I wanted to basically just stick a CD in the drive and let it do it's thing. Any help on how to accomplish this would be appreciated. Devin Atencio

2 1

Cuba and tzdata-2007d-1.el5
by Jesús García González 17 May '07

17 May '07

Hello all, It seems, tzdata-2007d-1.el5.noarch.rpm contains information about Cuba/Havana time zone that activates DST at a time other than the one that is going on in the country... Is there an updated tzdata somewhere correcting this info? or how can I correct this on my own? Thanks, J.

1 0

centos rsync paths: cobbler
by Karl R. Balsmeier 17 May '07

17 May '07

am using cobbler to import via rsync -and having trouble munging paths properly to download the 64-bit dvd or others, -what's a known good rsync path to use, or a command line that would allow me to determine this myself? When I use the web browsers of course and go to centos.org, you'll get the usual "not a suppoorted protocol" when clicking around trying to isolate a good path/server. i'm doing: cobbler import --mirror=rsync://need.servername.domain/path/to/centos86_64 --name=centos64 -any suggestions? (sorry, new to this, but definitely interested in rolling out cobbler). -krb

2 1

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