1and1 RootServer - Need to clarify...
1and1 RootServer - Need to clarify...
Hi all,
I have finally got around to moving from an old 1and1 root server (RH7, Ensim 3.1) to a newer (FC2, Plesk 7.5.4) one although it is still 1and1. Wow! I am impressed with the quality of Plesk.
Anyway, I have spent a few hours reading these forums and the ones at SWSoft and just need to confirm one question before I start upgrading and securing parts of the system.
I notice that there are a number of posts that warn you to exclude the "kernel*" when updating your /etc/yup.conf file. Is this still a requirement when working on a 1and1 root server?
I have finally got around to moving from an old 1and1 root server (RH7, Ensim 3.1) to a newer (FC2, Plesk 7.5.4) one although it is still 1and1. Wow! I am impressed with the quality of Plesk.
Anyway, I have spent a few hours reading these forums and the ones at SWSoft and just need to confirm one question before I start upgrading and securing parts of the system.
I notice that there are a number of posts that warn you to exclude the "kernel*" when updating your /etc/yup.conf file. Is this still a requirement when working on a 1and1 root server?
Yes.I notice that there are a number of posts that warn you to exclude the "kernel*" when updating your /etc/yum.conf file. Is this still a requirement when working on a 1and1 root server?
As I understand it, this is primarily due to customizations to facilitate the "Recovery Console". There may be other reasons as well.
See http://www.1and1faq.com/forums/ where there's a thread buried somewhere about how to properly update the kernel on a 1&1 box. Note: A re-image I did earlier this month included the current version of the FC3 kernel, so you may not have to do this right away.
Cheers!brucew wrote:A re-image I did earlier this month included the current version of the FC3 kernel, so you may not have to do this right away.
That sounds interesting about your re-image as my root server value is less than 1 week old. I shall have to look and see what kernels exist on it. Plesk says it is FC2 at the moment.
Well that update was painless. All nicely "yummed" including the mysql 3.xx to 4.xx bit.
So thanks to Scott and co for the information in these forums. It's saved my quite a bit of work.
Now to install awstats as I am bored with webalizer. I read about an install routine on SW-Soft so time to see how good it is.
So thanks to Scott and co for the information in these forums. It's saved my quite a bit of work.
Now to install awstats as I am bored with webalizer. I read about an install routine on SW-Soft so time to see how good it is.
All I know is that two years ago when up2date told me there was a new kernel and I let it update, the box wouldn't boot and I had to have it re-imaged. Fortunately, I hadn't moved clients to it yet. This was back in the day when that process involved phoning support followed by endless faxing rather than a mouse click. While on the phone, that's what the tech told me--that it had to do with their recovery console and their little partition that facilitates it.
Once bitten...
It's nice to know you have a way around that. Is that part of ASL?
Once bitten...
It's nice to know you have a way around that. Is that part of ASL?
-
- Atomicorp Staff - Site Admin
- Posts: 8355
- Joined: Wed Dec 31, 1969 8:00 pm
- Location: earth
- Contact:
Yeah, the atomicorp website is actually running on an ASL'd 1and1 box. In addition I also fixed their totally screwed up partition layout:
[root@u15188619 root]# uname -a
Linux u15188619.onlinehome-server.com 2.6.14-4.art #1 Sat Mar 4 18:07:49 EST 2006 i686 i686 i386 GNU/Linux
[root@u15188619 root]# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/hda5 74493500 8204900 62504492 12% /
/dev/hda1 489992 25914 438778 6% /boot
none 513920 0 513920 0% /dev/shm
[root@u15188619 root]#
Ive been meaning to do an article on how I fixed their partitioning layout, since I know a lot of people run into problems with it.
[root@u15188619 root]# uname -a
Linux u15188619.onlinehome-server.com 2.6.14-4.art #1 Sat Mar 4 18:07:49 EST 2006 i686 i686 i386 GNU/Linux
[root@u15188619 root]# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/hda5 74493500 8204900 62504492 12% /
/dev/hda1 489992 25914 438778 6% /boot
none 513920 0 513920 0% /dev/shm
[root@u15188619 root]#
Ive been meaning to do an article on how I fixed their partitioning layout, since I know a lot of people run into problems with it.
I, for one, sure would appreciate such an article. My new box there has been up for three weeks and the stupid miniscule /var partition is 68% full as of this morning. Just where the hell do they think logs, caches, mail and databases are supposed to go?
I was going to do something with symlinks, but I'd rather change the partitioning--solve the problem, not just the symptoms.
I was going to do something with symlinks, but I'd rather change the partitioning--solve the problem, not just the symptoms.
-
- Atomicorp Staff - Site Admin
- Posts: 8355
- Joined: Wed Dec 31, 1969 8:00 pm
- Location: earth
- Contact:
I felt your pain, became enraged, and then fixed it Its not a particularly complex operation, but there are a bunch of steps and you will obviously have to take your system down for a few hours while you do it.
Basically:
1) Back the system up (to the 1and1 backup server)
2) Boot in rescue mode
3) zap all the partitions except for sda1/hda1 (this will become /boot)
4) clean all the garbage out of sda1/hda1, move /boot into it
5) fdisk, delete all the other partitions, make swap on sda2/hda2, and put everything else in sda3/hda3
6) mkfs (I used ext3 for lazy reasons) on sda3/hda3
7) mount up the new partition, restore your data from the backup server
e2label your new /boot a respectively
9) update /etc/fstab
LABEL=/ / ext3 defaults 1 1
LABEL=/boot /boot ext3 defaults 1 2
none /dev/pts devpts gid=5,mode=620 0 0
none /dev/shm tmpfs defaults 0 0
none /proc proc defaults 0 0
none /sys sysfs defaults 0 0
/dev/hda2 swap swap defaults 0 0
10) update lilo, or install grub
11) reboot
Basically:
1) Back the system up (to the 1and1 backup server)
2) Boot in rescue mode
3) zap all the partitions except for sda1/hda1 (this will become /boot)
4) clean all the garbage out of sda1/hda1, move /boot into it
5) fdisk, delete all the other partitions, make swap on sda2/hda2, and put everything else in sda3/hda3
6) mkfs (I used ext3 for lazy reasons) on sda3/hda3
7) mount up the new partition, restore your data from the backup server
e2label your new /boot a respectively
9) update /etc/fstab
LABEL=/ / ext3 defaults 1 1
LABEL=/boot /boot ext3 defaults 1 2
none /dev/pts devpts gid=5,mode=620 0 0
none /dev/shm tmpfs defaults 0 0
none /proc proc defaults 0 0
none /sys sysfs defaults 0 0
/dev/hda2 swap swap defaults 0 0
10) update lilo, or install grub
11) reboot
Awstats installed success. Followed the install instructions on SW-Soft forum and apart from one directory not created, it was very straightforward.
Hmm, changing the server partition sizes is a little too much to take on at the moment. I'm not going to be making too many changes to the system at the moment and I don't see myself growing higher than 50 low traffic websites. At least not on one of these value servers.
If I do then I should be able to afford a higher spec server.
Hmm, changing the server partition sizes is a little too much to take on at the moment. I'm not going to be making too many changes to the system at the moment and I don't see myself growing higher than 50 low traffic websites. At least not on one of these value servers.
If I do then I should be able to afford a higher spec server.
-
- Forum Regular
- Posts: 196
- Joined: Tue May 10, 2005 1:24 pm
Scott, I have run into so many problems with the 1and1 partitioning that I have decided I need to go ahead and repartition. I think I can make it through your suggested steps, but I have a question. Do I need to backup the complete server? I assume I need to do. If so, what is your suggested way to backup to keep all file permissions the same? I don't use 1and1's backup service. I backup remotely using ssh and 4psa's Total Backup. I assume this doesn't copy everything I would need. As always, thanks for the help.
-
- Atomicorp Staff - Site Admin
- Posts: 8355
- Joined: Wed Dec 31, 1969 8:00 pm
- Location: earth
- Contact:
If you use:
tar zcvfp /home/backup.tar.gz /
that "p" flag will keep the permissions, links, etc intact. This is from memory, but what I did was:
1) back up all the whole box with that, transfer the backup.tar.gz to the 1and1 ftp server. (Everything was shut down at this point)
2) boot it into rescue mode
3) fdisk /dev/sda (you may have hda), delete all partitions except for /
4) create sda2 as swap, set it to 100% of ram (1G, or 2G)
5) create sda3 with all the remaining disk space
6) reboot, this resets the partition table
7) mkswap /dev/sda2
mkfs.ext3 /dev/sda3
9) mount /dev/sda3 somewhere, like /mnt/
10) cd /mnt, mkdir /boot, mount /dev/sda1 boot/
11) cd boot/, mv boot/* . (you do this because this was the old /, and /boot was a dir in that partition. You're fixing the fs layout)
12) cd /mnt, and grab your backup from ftp
13) tar zxvfp backup.tar.gz
14) cd /mnt/etc, vi your fstab, and fix your paths
15) vi /etc/lilo.conf, fix your paths
16) lilo -r /mnt (this installs lilo back in). If you get errors here, thats bad. You might need to relabel the disks with e2label.
17) cd /, umount /mnt/boot, umount /mnt
18) set it to boot into the OS, and reboot
And yes, this will be in the book, so please post your experiences here, I want to make sure all the angles and problems are covered.
tar zcvfp /home/backup.tar.gz /
that "p" flag will keep the permissions, links, etc intact. This is from memory, but what I did was:
1) back up all the whole box with that, transfer the backup.tar.gz to the 1and1 ftp server. (Everything was shut down at this point)
2) boot it into rescue mode
3) fdisk /dev/sda (you may have hda), delete all partitions except for /
4) create sda2 as swap, set it to 100% of ram (1G, or 2G)
5) create sda3 with all the remaining disk space
6) reboot, this resets the partition table
7) mkswap /dev/sda2
mkfs.ext3 /dev/sda3
9) mount /dev/sda3 somewhere, like /mnt/
10) cd /mnt, mkdir /boot, mount /dev/sda1 boot/
11) cd boot/, mv boot/* . (you do this because this was the old /, and /boot was a dir in that partition. You're fixing the fs layout)
12) cd /mnt, and grab your backup from ftp
13) tar zxvfp backup.tar.gz
14) cd /mnt/etc, vi your fstab, and fix your paths
15) vi /etc/lilo.conf, fix your paths
16) lilo -r /mnt (this installs lilo back in). If you get errors here, thats bad. You might need to relabel the disks with e2label.
17) cd /, umount /mnt/boot, umount /mnt
18) set it to boot into the OS, and reboot
And yes, this will be in the book, so please post your experiences here, I want to make sure all the angles and problems are covered.
-
- Forum Regular
- Posts: 196
- Joined: Tue May 10, 2005 1:24 pm
Scott, I really appreciate this very detailed explaination. I am going to be giving this a shot in the next couple of weeks. I see that the reimages of 1and1 servers now come with plesk 8, fc4, and a new partitioning. I am thinking of
1. Doing a backup
2. Upgrading to Plesk 8
3. Doing another backup
4. Re-image the server
5. Restore my data
6. Be sure everything is ok
7. Do the backup as you explain here
8. Fix the partitions
I will let you know of how things turn out in the future. Again, thanks for all the help you provide.
1. Doing a backup
2. Upgrading to Plesk 8
3. Doing another backup
4. Re-image the server
5. Restore my data
6. Be sure everything is ok
7. Do the backup as you explain here
8. Fix the partitions
I will let you know of how things turn out in the future. Again, thanks for all the help you provide.
I wouldn't try the 1and1 reimage to FC4/P8. eWeb tried it and he's had all sorts of problems. If eWeb won't put his name behind it then there's something seriously wrong with it.Troy McClure wrote:Scott, I really appreciate this very detailed explaination. I am going to be giving this a shot in the next couple of weeks. I see that the reimages of 1and1 servers now come with plesk 8, fc4, and a new partitioning. I am thinking of
1. Doing a backup
2. Upgrading to Plesk 8
3. Doing another backup
4. Re-image the server
5. Restore my data
6. Be sure everything is ok
7. Do the backup as you explain here
8. Fix the partitions
I will let you know of how things turn out in the future. Again, thanks for all the help you provide.
http://www.1and1faq.com/forums/showthread.php?t=438