Page 1 of 2
New 1and1 images
Posted: Fri Apr 25, 2008 8:34 am
by Highland
I'm kinda slack jawed right now. 1and1 just did a beautiful thing and made CentOS 5 their default image. In addition they've made pretty much any image you could want available in a minimal install
The images in my list as of this morning are
- CentOS 5 with Plesk 8.3 (32Bit)
- Fedora Core 6 with Plesk 8.2
- SUSE 10.2 with Plesk 8.2
- CentOS 5 minimal system
- SuSE-10.3 minimal system (32Bit)
- SUSE 10.2 minimal system
- Fedora Core 8 minimal system
- Fedora Core 6 minimal system
- Debian 4.0 (etch) minimal system
- Ubuntu 6.06 minimal system
Posted: Fri Apr 25, 2008 9:25 am
by scott
I'd hold that thought until we see what kind of damage they did to it. I suspect at the very least that they trashed the kernel, and screwed up the partitioning.
Posted: Wed Apr 30, 2008 4:22 pm
by cmaxwell
They've still got their stupid custom kernels on all the new images which really sucks.
I'm trying to compile a new stock kernel on a brand new CentOS 5 (64 bit) box and everything goes well until the mkinitrd stage where I get this error:
Code: Select all
mkinitrd /boot/initrd-2.6.25-1.img 2.6.25
No module uhci-hcd found for kernel 2.6.25, aborting.
Has anyone got any suggestions on how to get around this? I'm tearing my hair out and really need to get a standard kernel on the system so I can install some other stuff.
Thanks!
Posted: Wed Apr 30, 2008 5:34 pm
by scott
Thats probably easy to get around, take the reference to uhci-hcd out of your modprobe.conf.
Worst case, AOOI the box.
Posted: Wed Apr 30, 2008 6:23 pm
by cmaxwell
Hi Scott,
Thanks for your reply. I don't see an entry for uhci-hcd in /etc/modprobe.conf:
Code: Select all
alias eth0 e1000
alias scsi_hostadapter1 sata_nv
alias scsi_hostadapter mptbase
alias scsi_hostadapter1 mptspi
alias snd-card-0 snd-ens1371
options snd-card-0 index=0
options snd-ens1371 index=0
remove snd-ens1371 { /usr/sbin/alsactl store 0 >/dev/null 2>&1 || : ; }; /sbin/modprobe -r --ignore-remove snd-ens1371
Any suggestions? I tried skipping USB support with mkinitrd, and then it came up with the same error but for RAID, and I can't skip RAID because the server has 2 x 160GB drives in a RAID1 config. I'm a bit confused because it's as if the modules don't exist yet I selected them from menuconfig and the kernel appeared to compile fine.
I tried AOOI earlier, set it up for network install, rebooted after it prompted and then it couldn't download the kickstart image. Looked as if it hadn't brought the NIC up... I ended up re-imaging the box.
I'd prefer just to replace the kernel if it's possible as everything else is fine on the server (Plesk, etc) but i'd also be happy to AOOI the box if it was fairly straight-forward but it doesn't seem so easy on a 1&1 box.
Posted: Wed Apr 30, 2008 6:42 pm
by scott
try local mode with AOOI at 1and1.
Posted: Wed Apr 30, 2008 7:41 pm
by cmaxwell
Have 1&1 really done something which makes it so difficult to replace the kernel that I need to reinstall the OS?
Not that i'm bothered, but it's quite drastic when all i want is a new/stable kernel
lol.
I'd be interested if anyone knows why it won't upgrade "as is" ie what's causing the error i previously quoted...
TIA.
Posted: Thu May 01, 2008 6:49 am
by cmaxwell
I tried AOOI again and here's what happened:
Code: Select all
Network or Local installation (network/local) [Default: network] ? local
Tranferring CentOS image to swap: mke2fs 1.39 (29-May-2006)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
245280 inodes, 489982 blocks
24499 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=503316480
15 block groups
32768 blocks per group, 32768 fragments per group
16352 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912
Writing inode tables: done
Creating journal (8192 blocks): done
Writing superblocks and filesystem accounting information: done
This filesystem will be automatically checked every 37 mounts or
180 days, whichever comes first. Use tune2fs -c or -i to override.
Done
Done
Getting CentOS ISO: --11:39:02-- http://mirror.1and1faq.fr/centos/5.0/isos/x86_64/CentOS-5.0-x86_64-bin-1of7.iso
Resolving mirror.1and1faq.fr... 87.106.179.206
Connecting to mirror.1and1faq.fr|87.106.179.206|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 655493120 (625M) [application/octet-stream]
Saving to: `CentOS-5.0-x86_64-bin-1of7.iso'
100%[======================================================================================================================================================================>] 655,493,120 11.2M/s in 57s
11:40:05 (11.1 MB/s) - `CentOS-5.0-x86_64-bin-1of7.iso' saved [655493120/655493120]
--11:40:05-- http://mirror.1and1faq.fr/centos/5.0/isos/x86_64/CentOS-5.0-x86_64-bin-2of7.iso
Resolving mirror.1and1faq.fr... 87.106.179.206
Connecting to mirror.1and1faq.fr|87.106.179.206|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 665100288 (634M) [application/octet-stream]
Saving to: `CentOS-5.0-x86_64-bin-2of7.iso'
100%[======================================================================================================================================================================>] 665,100,288 10.0M/s in 61s
11:41:06 (10.4 MB/s) - `CentOS-5.0-x86_64-bin-2of7.iso' saved [665100288/665100288]
Done.
Setting up installer kernel: Done
Not Done yet!
Code: Select all
[root@server ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/md1 950M 237M 665M 27% /
/dev/md5 4.7G 2.2G 2.6G 46% /usr
/dev/md7 137G 131M 137G 1% /var
/dev/md6 4.7G 4.4M 4.7G 1% /home
none 1.0G 12K 1.0G 1% /tmp
/dev/sda2 1.9G 1.3G 493M 73% /mnt
I rebooted and the system came up normally (ie booted CentOS, not the AOOI/CentOS installer) but I noticed the following error on boot:
Code: Select all
Remounting root filesystem in read-write mode: [ OK ]
Mounting local filesystems: [ OK ]
Enabling /etc/fstab swaps: swapon: /dev/sda2: Invalid argument swapon: /dev/sdb2: Invalid argument
[FAILED]
Any suggestions?
Thanks.
Posted: Thu May 01, 2008 9:02 am
by scott
The answer to your first question (is re-imaging necessary)... you tell me
For the latter, it looks like they finally switched to grub. Ive updated AOOI to handle that (I hope). Go ahead and run it again. It might complain about sda2 (this is where the ISO's get stored).
Posted: Thu May 01, 2008 11:10 am
by cmaxwell
Hi Scott,
Thanks! It seemed to work ok, it rebooted and loaded the CentOS installer (couldn't download the kickstart file) so i've chosen interactive installation. It's formatting now...
Job well done on AOOI Scott!!
PS: what's your recommended disk partitioning layout? I went with the default but i've no idea if that included software RAID1... no doubt i will be reinstalling this again anyway
Posted: Thu May 01, 2008 11:22 am
by cmaxwell
OK the installation went so far and then came up with:
Code: Select all
Package Installation
Missing ISO 9660 Image
The installer has tried to mount image #3, but cannot find it on the hard drive.
Please copy this image to the drive and click Retry.
Click Reboot to abort the installation.
Should I have done the kickstart option instead of interactive? If so, any suggestions for being able to access the kickstart file (it initially came up with the message saying it couldn't download the kickstart file)?
Posted: Thu May 01, 2008 5:36 pm
by scott
My partitioning layout is:
/boot, 256M, partition 1
swap, x2 ram, partition 2
/ <- everything else
AOOI isnt the most friendly thing for running more than once in local install mode. I'll bet it didnt grab the ISO's the 2nd time. What you can do is convert the swap partitions back (or re-image it if you're lazy. I know I am) and run it again.
The process it uses is:
1) figure out where your swap partition are
2) format it vfat (clever reasons why, which I wont go into)
3) mount that to /mnt
4) grab the isos, dump to /mnt
5) grab the ks, dump to /
6) change lilo/grub, set to boot once
If it cant figure out 1, then 2, 3, 4 and probably 5 fail.
Posted: Thu May 01, 2008 5:43 pm
by cmaxwell
Thanks Scott. I'm really struggling getting passed the kickstart file issue as it can't download it because it's not got internet access. I chose the local install method, it downloaded the 2 CentOS ISO's, when it reboots it starts the installer and says it can't download the kickstart file.
I've tried downloading the kickstart file locally (it didn't download it after i chose the local install method - not sure if it is meant to or not) and referencing it at hd:sda2/ks-x86_64.cfg, which worked, but then it asks for the HTTP address to get the CentOS files which I don't appear to be able to enter as a local path.
If it hasn't got internet access, but it needs this if it uses the kickstart file, is the only option to do an interactive installation? And if that's the case, how do we avoid it needing the third CentOS ISO which it asked for when i did an interactive install earlier (or was that because I chose the server components when it asked me what i wanted to install)?
The finish line is in sight i think!! Lol.
TIA.
Posted: Thu May 01, 2008 7:19 pm
by cmaxwell
Holy batpoop... i managed to get CentOS installed by selecting local install, then rebooting and doing an interactive install but make sure to deselect ALL packages when it lets you select them (make sure to choose 'Customize software selection' and uncheck ALL options).
Just literally fallen off my chair...
Thanks scott!
Posted: Sun May 04, 2008 11:25 am
by cmaxwell
Scott,
Just a thought, is it possible for the AOOI local install mode to download the kickstart file so the process could be more automated? I know you said this was what the local install mode should do in one of the other threads but it's not downloading the kickstart file when i choose local install.