Re: [CentOS] squid stops working several times a day
Alexander Farber wrote: Does anybody know of good Squid rpm's? I have sourced our 2.6 packages from http://people.redhat.com/mnagy/squid/ for the last few years without problems. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Re: [CentOS] Postfix header check help - SOLVED
mouss wrote: Plant, Dean a écrit : Hi list, I have a requirement for a mail server that only allows email to pass with a particular word in the subject line. Reading the header checks docs for Postfix I thought I may be able to add this rule: !/^Subject: .*dingdong/ REJECT Need to add dingdong to subject line to send Which should reject all mail unless the subject line contains dingdong. Unfortunately this seems to block all mail from passing. if /^Subject:/ !/dingdong/ REJECT blah blah endif Removing the ! works as expected and the server allows all mail to pass, unless dingdong is in the subject line which is rejected. Is this rule valid? Or is there a better way to make this work? the rule is (syntactically) valid. but it doesn't do what you want. take the following header: Date: Tue, 13 Jan 2009 15:51:40 - does it contain dingdong? does it start with Subject? so it's a REJECT. Thank you. Your example works as expected. It seems I was focusing only on the subject line but as you pointed out header checks apply to all headers. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Postfix header check help
Hi list, I have a requirement for a mail server that only allows email to pass with a particular word in the subject line. Reading the header checks docs for Postfix I thought I may be able to add this rule: !/^Subject: .*dingdong/ REJECT Need to add dingdong to subject line to send Which should reject all mail unless the subject line contains dingdong. Unfortunately this seems to block all mail from passing. Removing the ! works as expected and the server allows all mail to pass, unless dingdong is in the subject line which is rejected. Is this rule valid? Or is there a better way to make this work? Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] ks
[EMAIL PROTECTED] wrote: I made a file partition-ks.cfg and put it in our machine 192.168.5.17 under /var/www/html. When I install a new machine I just do a linux ks=http://192.168.5.17/partition-ks.cfg I have a couple questions though: 1) I have the following in it: clearpart --all part /boot --fstype ext3 --size=100 part /var --fstype ext3 --size=1 part / --fstype ext3 --size=1 part swap --size=2048 part /home --fstype ext3 --size=100 --grow which is OK if I am installing over a previous installation. But for new machines it does not seem to recognize the clearpart or it is not enough because I am still prompted whether I want to remove partitions. Should I be using something else? Take a look at system-config-kickstart to help build your kickstart files. Anyway I think what you are looking for is, # Clear the Master Boot Record zerombr # Partition clearing information clearpart --all --initlabel 2) I would like to duplicate the CentOS installation disk and place partition-ks.cfg on the disk too so that during boot it automatically is used (or barring that I can still say ks=dvd drive/partition-ks.cfg. Automatically would be better. I tried some ways I thought it said to do that but it did not work. Any tips? Copy the DVD put the ks.cfg file in the root of the DVD files vi /dvd-root/isolinux/isolinux.cfg and edit the append line or add a new entry like this label yourlabel kernel vmlinuz append initrd=initrd.img ks=cdrom:/ks.cfg Make iso and burn as described here: http://www.redhat.com/docs/manuals/enterprise/RHEL-4-Manual/x8664-multi- install-guide/s1-steps-install-cdrom.html Boot DVD and type yourlabel to activate your kickstart file. HTH Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Command line partition manipulation
I want to write a script to automatically modify the partition table I am starting with 3 partitions like this # parted /dev/sda print Model: VMware, VMware Virtual S (scsi) Disk /dev/sda: 10.7GB Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start End SizeType File system Flags 1 32.3kB 107MB 107MB primary ext3 boot 2 107MB 4828MB 4721MB primary ext3 3 4828MB 9023MB 4195MB primary lvm And I want to remove partitions 2 3 and recreate 2 from the end of 1 to the end of the disk i.e. all the free space. Removing 2 and 3 is easy. # parted /dev/sda rm 3 # parted /dev/sda rm 2 # parted /dev/sda print Model: VMware, VMware Virtual S (scsi) Disk /dev/sda: 10.7GB Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start EndSize Type File system Flags 1 32.3kB 107MB 107MB primary ext3 boot Is there a way to use a non-interactive command to create the new 2nd partition in the free space? Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Command line partition manipulation
Karanbir Singh wrote:
Plant, Dean wrote:
Is there a way to use a non-interactive command to create the new 2nd
partition in the free space?
parted works well for such situations, and if python is your thing,
pyparted is already included in the distro.
Sorry, I should have worded my first email better.
I know there are command line tools like fdisk, sfdisk and parted.
I am trying to script the creation of a partition in the free space of a
disk without knowing the exact structure of the disk. Parted seems to
want to know the exact layout you want to create
From the man page
mkpart part-type [fs-type] start end
But the script will not know the start or the end point.
Something similar to what I am trying to achieve is described here
http://ubuntuforums.org/showthread.php?p=5435429
They seem to be creating a single partition the size of the disk without
specifying the exact sizes.
# parted /dev/${i} --script -- mkpart primary 0 -1
Similar to this, I want to create a partition without knowing the exact
start or end point but im not interested in that, I just want to use the
free space.
Hope this is clearer.
Thanks
Dean.
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Whole disk encryption - SOLVED
Timothy Selivanow wrote: On Mon, 2008-08-04 at 16:51 +0100, Plant, Dean wrote: Has there been any updates to support encrypting the whole disk in 5.2? There hasn't been any built-in support until Fedora 9, so perhaps at the earliest it would be 5.3 if at all. There are however, ways you can implement it yourself. The biggest things you have to keep in mind are that you need to make a change to the mkinitrd script and then generate a new initrd image to be able to encrypt /, otherwise you could just modify init. There are a number of websites that have some docs on how to do it, here is just one that I've seen in the past: http://www.tummy.com/Community/Articles/cryptoroot-f8/ Thank you to everyone who replied. Tim, Thanks for the above link. Those instructions work fine on v5.2. You have to manually edit /sbin/mkinitrd but the changes are obvious enough from their patch file. Not quite whole disk encryption but a good compromise. Now just need to put this in a kickstart file to try and automate the setup. Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Whole disk encryption
Has there been any updates to support encrypting the whole disk in 5.2? If not, Is anyone doing this and can point me to some good documentation? Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Wheel and YUM!!
Harry Sukumar wrote: Hi James, Yes I tried to do this on another machine unfortunately this didn't work for me, good you raised this topic On the other machine I edited the visudo User_Alias PROJECT = emma, paddy, sb PROJECT ALL = !/usr/su, !/bin/su, !/usr/bin/passwd I see your trying to protect your users from becoming root. You do realise that with that sudo configuration a user can still run sudo /bin/bash or any of the other shells to gain root access. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Wheel and YUM!!
Ian Blackwell wrote: Plant, Dean wrote: I see your trying to protect your users from becoming root. You do realise that with that sudo configuration a user can still run sudo /bin/bash or any of the other shells to gain root access. Hi Dean, I don't think that's correct. One of the purposes of the sudoers file is you can limit users to specific programs, excluding them from others. This entry, which I just tested in my CentOS5.2 box, allows the user tldap to use yum, but not /bin/bash:- tldap ALL=/usr/bin/yum Here's what happens when I try sudo as user tldap:- [EMAIL PROTECTED] ~]$ sudo yum update Loading fastestmirror plugin Loading priorities plugin Loading mirror speeds from cached hostfile * base: rsync.atworks.co.jp * updates: rsync.atworks.co.jp * centosplus: mirror.exetel.com.au * addons: mirror.exetel.com.au * extras: mirror.exetel.com.au 0 packages excluded due to repository priority protections Setting up Update Process No Packages marked for Update [EMAIL PROTECTED] ~]$ sudo /bin/bash Sorry, user tldap is not allowed to execute '/bin/bash' as root on www [EMAIL PROTECTED] ~]$ Yes you are correct in your example only yum would run and no exclude is required. We can not see from his posting exactly what commands he is allowing as he is calling groups. What I was trying to point out is that if is he is disabling commands like su then they must be enabled somewhere in the groups he is calling and it is good practice to disable all the shells as well. All my sudoers lines that call groups like he was trying to do always have a !SU, !SHELLS to specifically deny root access. Anyway I will shut up now as none of this will help fix his problem. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Hardened ver of CentOS?
Rogelio wrote: John R Pierce wrote: CentOS /is/ a distro, there is only one centos 'distribution'. centos configured with selinux enabled, appropriate firewall rules, and the minimum number of services required for your application should be fairly 'hardened' as-is. Understood. I meant CentOS-based, but I suppose the best way is to just roll something myself. I was hoping to find a nice list HOWTO or script that someone else was using so I wouldn't have to think through everything from scratch. Have you looked at Bastille Linux? http://bastille-linux.sourceforge.net/ ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] centos 5.1 i386 on a 1GIG flash
Jerry Geis wrote: I have found this link http://owlriver.com/tips/tiny-centos for installing centos on a minimal system. I am looking at putting centos on a 1 GIG flash drive. The above page talks about removing packages after install to attain the small size. however, I am getting blocked at the install page about not enough room to install. All I have selected is the base package. nothing else. Is there a way to install less thank base... I tried to option off items in the base package but the last message I got was about 1044 was my packages size... Anyone have suggestions or have tried to install on a small device. THanks, Jerry Create a kickstart file with the packages removed as shown in the appendix on their web page. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Re: Kickstart network settings problem -SOLVED
Tony Mountifield wrote: In article [EMAIL PROTECTED], Plant, Dean [EMAIL PROTECTED] wrote: If like me you have never had to mess with the stage2.img file to fix installations issues, below shows what is required. cd ~ wget http://yourwebserver/centos5-dvd/images/stage2.img mkdir tmp mount -t squashfs -o loop stage2.img ./tmp cd tmp vi ./usr/lib/anaconda/kickstart.py # comment out dispatch.skipStep(network) at around line 876 or use Ray's patch. hmm, isn't this editing the file on a read-only file system? Oops..yes, not going to do much good editing the ro file system! Sorry. mkdir ../new-stage2 tar cf - . | ( cd ../new-stage2/ ; tar xf - . ) I suspect here is where you need to cd ../new-stage2 and edit kickstart.py cd .. yum -y install squashfs-tools mksquashfs ./new-stage2/ stage2.img.new Now replace the stage2.img file on the web server with the stage2.img.new file, renamed to stage2.img of course. Cheers Tony ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] Kickstart network settings problem
I would like one of my kickstart scripts to prompt for disk and networking only. Commenting out the disk and network settings makes the installer prompt for disk settings but not for network settings, they default to DHCP. Any idea's how to fix this? Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Kickstart network settings problem
James Pearson wrote: Plant, Dean wrote: I would like one of my kickstart scripts to prompt for disk and networking only. Commenting out the disk and network settings makes the installer prompt for disk settings but not for network settings, they default to DHCP. Any idea's how to fix this? I think you need to have the network settings defined - something like: network --bootproto static If you don't give it an IP address (with the --ip option), then it will prompt for it. This is what I do for NFS installs - as my netmask, gateway and nameserver addresses are static, I also give these on the network line. Does not work, still defaults to DHCP with no prompt. Also if it makes any difference im using v5.1. I have tried with the 2 network lines shown below in the kickstart file, with and without interactive variable. Still no network configuration screen. :-( Below is my test kickstart. #platform=x86, AMD64, or Intel EM64T # System authorization information authconfig --enableshadow --enablemd5 # System bootloader configuration bootloader --location=mbr --md5pass=apssword # Reboot after installation reboot # Interactive interactive # Use graphical install graphical # Firewall configuration firewall --disabled # Run the Setup Agent on first boot firstboot --disable # System keyboard keyboard uk # Network settings #network --bootproto=static --device=eth0 --gateway=192.168.10.1 --nameserver=192.168.10.1 --netmask=255.255.255.0 --onboot=on network --bootproto=static # System language lang en_GB # Installation logging level # Use this to debug logging --host=alogmachine.domain.co.uk --level=info # Use network installation url --url=http://amachine.domain.co.uk/5/en/os/i386/ # Root password rootpw --iscrypted apassword # SELinux configuration selinux --disabled # System timezone timezone --isUtc Europe/London # Install OS instead of upgrade install # XWindows configuration information xconfig --depth=16 --resolution=1280x1024 --defaultdesktop=GNOME --startxonboot ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Kickstart network settings problem
James Pearson wrote: Plant, Dean wrote: James Pearson wrote: Plant, Dean wrote: I would like one of my kickstart scripts to prompt for disk and networking only. Commenting out the disk and network settings makes the installer prompt for disk settings but not for network settings, they default to DHCP. Any idea's how to fix this? I think you need to have the network settings defined - something like: network --bootproto static If you don't give it an IP address (with the --ip option), then it will prompt for it. This is what I do for NFS installs - as my netmask, gateway and nameserver addresses are static, I also give these on the network line. Does not work, still defaults to DHCP with no prompt. Also if it makes any difference im using v5.1. I have tried with the 2 network lines shown below in the kickstart file, with and without interactive variable. Still no network configuration screen. :-( Hmmm, works OK for me with CentOS 4.X - however, I used a hacked version of anaconda - one of my hacks is to set 'netDev.isDynamic' to 0 in loader2/loader.c i.e. make static IP the default ... however, I thought this was for non-kickstart installs only ... but may be it isn't. Looks like it's a CentOS 5 feature/bug as I have just tested every incarnation of the network line and they all get ignored unless I put in the complete line as per Marc-Andre's post. This DID work in CentOS 4 as long as the interactive line was specified. Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Kickstart network settings problem
Ray Van Dolson wrote: On Tue, Apr 22, 2008 at 12:47:39PM +0100, Plant, Dean wrote: I would like one of my kickstart scripts to prompt for disk and networking only. Commenting out the disk and network settings makes the installer prompt for disk settings but not for network settings, they default to DHCP. Any idea's how to fix this? I'm assuming you are using CentOS 5 here. I'm not sure why upstream disabled this feature -- it was very useful to be able to use kickstart with a DHCP IP address initially, but still be prompted for network information during the installer portion later (as it did in RHEL4). See: https://bugzilla.redhat.com/show_bug.cgi?id=401531 In addition, I have an RFE in with Upstream to re-add this feature and I believe it's slated for inclusion in 5.2 with the --bootproto=query option. The attachment there however: https://bugzilla.redhat.com/attachment.cgi?id=291881 patches Anaconda to not skip the network step during installation. I've been using this successfully here to do what I need while waiting for the official fix. You can include this in an updates.img file and have it loaded automatically on installs. Ray Ah. Im not going daft then. Ill give the patches a go. Thanks everyone for your replies. Dean. ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] CentOS 5.1 on Dell R200 with 6iR?
Hello list, Can anyone confirm if CentOS 5.1 works happily on a Dell R200 with a SAS 6iR RAID controller. Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
[CentOS] ISCSI help
I am fairly new to ISCSI and SAN technology but having recently invested in the technology I am trying to find out exactly what can and can not be manipulated, filesystem wise, without requiring a reboot. I am using the inbuilt software ISCSI initiator and multipathing in CentOS 5.1. My steps so far. Create 10GB volume on SAN # iscsiadm -m session -R # fdisk /dev/mapper/mpath0 # kpartx -a /dev/mapper/mpath0 # mke2fs -j /dev/mapper/mpath0p1 # mount /dev/mapper/mpath0p1 /test-mount --works fine to here-- Now I want to extend the volume on the SAN to 15GB run fdisk and use resize2fs to extend the filesystem, is this possible without a reboot? Currently, I don't seem to be able to get fdisk to see the new disk size after extending the volume. I know this can be done using LVM if I created 2 volumes rather than extending but I am curious to know if it can be done without LVM. Any other tips about what can be done with ISCSI would be welcome. Thanks Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Centos 5.1 distro: How to upgrade 5.0 to 5.1 with yum??
Robinson Tiemuqinke wrote:
--- Jancio Wodnik [EMAIL PROTECTED] wrote:
Robinson Tiemuqinke pisze:
Currently I download the Centos 5.0 upgrade packages from mirror
sites on Internet (.../centos/5/updates/{SRPMS,i386,x86_64}/...),
with the arrival of Centos 5.1, are the existing Centos 5.0
update packages will be removed in honor of Centos 5.1
updates? or it will stay? And how about the Centos
5.1's default packages in distro?
Give an example here:
2.6.18-8.1.15 version kernel is the most recently
updated kernel for Centos 5.0 distro, if the Centos
5.1 distro comes with 2.6.18-8.1.1 kernel and in
the first few weeks there are no kernel updates for
5.1, then how can I upgrade my kernel to
2.6.18-8.1.1 naturally -- will the
2.6.18-8.1.1 shows in the same update sources
directories(.../centos/5/updates/{SRPMS,i386,x86_64/...)?
A similar question is: are the update diretories
contains only updates for 5.1 distro, or both 5.0 and
5.1?
Any clarifications are greatly appreciated.
It's simply. All things from 5.0 to 5.1 will be done automatically
via yum (as standard update) or in rare situation can be possibly
depedency problem (when mixing different repo ?).
So don't worry. if you are really affraid - update
only one box and test
... test ... then update the rest.
I must say. In Centos 4 world all updates from 1 to 2 to 3 ... to 5
were go in smooth way.
Regards,
Irens
I have had my local 5.0 update repository
(.../centos/5.0/updates/{SRPMS,i386,x86_64}/...) setup
and used it for my 900+ boxes's daily upgrade already.
The repository is synchronized with official Internet
mirrors daily to keep it current.
My major concerns is: After the 5.1 is released, the
update channel/directory
(.../centos/5.1/updates/{SRPMS,i386,x86_64}/...) may
change to contain only updates for 5.1
snapshot/release, not updates since 5.0. If so, then
all my Centos 5.0 boxes will suffer.
I have the serious concern because most Centos Mirror
sites on Internet ONLY keep the updates for latest
release/snapshot, not holds updates since the
base(3.0, 4.0, 5.0 etc) release. This seems like a big
problem if we would like to install from base|initial
release (3.0, 4.0, 5.0) continuously and then use a
single up-to-date update/ repository to upgrade
machines to current level.
For example, at Stanford's Centos 4 mirror site, only
4.5 is mirrored while all the other
4.0/4.1/4.2/4.3/4.4 are not. and in the updates/
directory only updates for 4.5 are kept there. If the
same is true for all other sites honoring 5.0 series,
then I think I will definitely get screwed If I tried
to keep on using base 5.0 and daily synced updates/
(exactly the same) for upgrade.
Any mirror sites hold updates since base release? Or I
have to keep on adding more repositories to yum's
configuration? 5.0 distro, 5.0 updates, 5.1 distro,
5.1 updates, 5.2 distro, 5.2 updates. etc. If so, then
it is too low-performanced and erro-prone.
Any one have experience on upgrade Centos 4 releases
from 4.0 to 4.5 can shed a light on this?
Thanks a lot.
--Robinson
If you remove the point release numbers from your update paths you do
not need the old versions. If you really need the old updates use
http://vault.centos.org/
From http://mirror.centos.org/centos-4/4.4/readme
This directory (and version of CentOS) is depreciated. For normal
users,
you should use /4/ and not /4.4/ in your path. Please see this FAQ
concerning the CentOS release scheme:
http://www.centos.org/modules/smartfaq/faq.php?faqid=34
If you know what you are doing, and absolutely want to remain at the 4.4
level, go to http://vault.centos.org/ for packages.
To simplfy our yum configuration I disable all the repo files in
/etc/yum.repos.d and call the yum configuration from a local web server
in yum.conf by using the include statement. One change on the web server
changes the yum configuration on all CentOS machines.
In the yum.conf I add
# Site yum files
include=http://ourserver.ourdomain.com/centos/yum/centos.repo
include=http://ourserver.ourdomain.com/centos/yum/centos-updates.repo
The centos.repo looks like this,
[base]
name=CentOS-$releasever - Base
baseurl=http://ourserver.ourdomain.com/centos/$releasever/os/$basearch/
enabled=1
gpgcheck=1
And the centos-updates.repo looks like this,
[updates-released]
name=CentOS-$releasever - Updates
baseurl=http://ourserver.ourdomain.com/centos/$releasever/updates/$basea
rch/
enabled=1
gpgcheck=1
[custom-repo]
name=CentOS-$releasever - Custom-rpms
baseurl=http://ourserver.ourdomain.com/centos/$releasever/customrepo/
enabled=1
gpgcheck=1
Hope this helps.
Dean
___
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] NIS binding probs w/Firewall and SELinux
Scott Ehrlich wrote: Is there a file I can edit to ensure SELinux is disabled? The system was initially installed with SELinux Enabled, then disabled later by me. Are you aware of the commands getenforce setenforce? Dean ___ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
RE: [CentOS] Adaptec 39320A woes
Simple fix, Adaptec CentOS = :-( LSI CentOS = :-) We spent a few days trying to get Adaptec cards working with LTO-3 drives. It was suggested to me from this list to try an LSI, we bought one, it worked, problem solved. Dean David Groep wrote: [sorry for breaking the threading, only joined after reading this on the web] Michael St. Laurent wrote: I'm having speed problems with the SCSI card we're using to do tape backup. It seems to be functioning in 16 bit mode and the current thinking is that perhaps it's using a legacy driver instead of the correct one. The Adaptec site has a 'driver' for RHEL5 which I've downloaded and tried to install but it seems to have a problem installing on a CentOS-5 system. Hit this same issue with a slightly different configuration, and although not able to offer a solution, I can at least give some more boundary conditions. As I see it now, it starts to look like a driver issue. - I found the issue with a 39320 controller on CentOS5/i386, talking to an Infortrend IFT-7250F enclosure. It should be doing U160 (Wide, 80 MHz synchonous), but it decays to 3.300 MB/s transfers. The driver indeed even says the Goal is 3.300 MB/s ... It is confirmed by the infortrend: it shows asynchonous/narrow as transfer settings for the SCSI channels after loading the aic79xx driver. - Exchanging the 39320 for a 29320, still on the same machine with CentOS5/i386, does not change it: still 3.300 MB/s (as confirmed by the infortrend itself) - putting this same 29320, with the same IFT7250F and with the same cabling, in another machine running CentOS3.7: presto, it synchonizes again at 160 MB/s (Wide, 80 MHz)! - similar Adaptec 29320 and 39320's in a CentOS4/i386 box, connected to similar Infortrends: work correctly at 160 and 320 MB/s - Booting into kernel 2.6.18-8.el5 or 2.6.18-8.1.8.el5 makes no difference For CentOS5, the aic79xx driver is Version: 3.0. For CentOS3 it's using driver 1.3.10-RH1, and for the CentOS4 systems, it's 1.3.11. What changed between 1.3.11 and 3.0? At least you're not alone ... Regards, DavidG. PS: and indeed, the adaptec driver RPM does not install correctly, with the same errors you reported below. [root at hcnas modules]# cd /proc/scsi [root at hcnas scsi]# cat scsi Attached devices: Host: scsi0 Channel: 00 Id: 08 Lun: 00 Vendor: DP Model: BACKPLANERev: 1.05 Type: EnclosureANSI SCSI revision: 05 Host: scsi0 Channel: 02 Id: 00 Lun: 00 Vendor: DELL Model: PERC 5/i Rev: 1.03 Type: Direct-AccessANSI SCSI revision: 05 Host: scsi1 Channel: 00 Id: 06 Lun: 00 Vendor: CERTANCE Model: ULTRIUM 2Rev: 1914 Type: Sequential-AccessANSI SCSI revision: 03 Host: scsi1 Channel: 00 Id: 06 Lun: 01 Vendor: DELL Model: PV-124T Rev: 0043 Type: Medium Changer ANSI SCSI revision: 02 Host: scsi3 Channel: 00 Id: 00 Lun: 00 Vendor: Dell Model: Virtual CDROM Rev: 123 Type: CD-ROM ANSI SCSI revision: 02 Host: scsi4 Channel: 00 Id: 00 Lun: 00 Vendor: Dell Model: Virtual Floppy Rev: 123 Type: Direct-AccessANSI SCSI revision: 02 [root at hcnas scsi]# ll total 0 dr-xr-xr-x 2 root root 0 Aug 3 13:21 aic79xx -r--r--r-- 1 root root 0 Aug 3 13:21 device_info -r--r--r-- 1 root root 0 Aug 3 13:21 scsi dr-xr-xr-x 2 root root 0 Aug 3 13:21 sg dr-xr-xr-x 2 root root 0 Aug 3 13:21 usb-storage [root at hcnas scsi]# cd aic79xx/ [root at hcnas aic79xx]# ll total 0 -rw-r--r-- 1 root root 0 Aug 3 13:09 1 -rw-r--r-- 1 root root 0 Aug 3 13:09 2 [root at hcnas aic79xx]# cat 1 Adaptec AIC79xx driver version: 3.0 Adaptec 39320A Ultra320 SCSI adapter aic7902: Ultra320 Wide Channel A, SCSI Id=7, PCI-X 101-133Mhz, 512 SCBs Allocated SCBs: 4, SG List Length: 128 Serial EEPROM: 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x17c8 0x09f4 0x0142 0x2807 0x0010 0x 0x 0x 0x 0x 0x 0x 0x 0x 0x 0x0430 0xb3f3 Target 0 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 1 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 2 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 3 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 4 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 5 Negotiation Settings User: 320.000MB/s transfers (160.000MHz RDSTRM|DT|IU|RTI|QAS, 16bit) Target 6 Negotiation Settings User: 320.000MB/s transfers (160.000MHz
