You are here
Arwen - Tue, 2013/07/23 - 22:58
Does anyone have any experience or instructions on how to do a distribution upgrade of any of the 12.1 turnkey images (from squeeze to Wheezy). I could not even find the /etc/apt/sources.list file in this path. Are the repos paths in a different file or location? Is it not a good idea to even try this for some reason? Is it better to wait for the 13.0 versions built on Wheezy and port my production files at that point instead of doing a distribution upgrade?
Thanks in advance!!
Forum:
It should be possible
But TBH I've never tried... The sources lists are in /etc/apt/sources.list.d/
Having said that though if this is a production server and you don't actually have a need for Wheezy then I'd wait. Squeeze will continue to get security updates for some time... You want to migrate to a TKL v13 appliance anyway IMO (depending on the appliance, some have upstream non-packaged components which will not be updated except by migrating to the new version). As for migration (when the time comes) that should be handled easily by TKLBAM.
If you decide to press ahead, let us know how you go.
I just did this...
I was playing around on a test system and wanted to do this (actually, I was in the middle of the process when I found this post).
I followed the steps on this page:
http://www.howtoforge.com/how-to-upgrade-debian-squeeze-to-wheezy
But I did two things different compared to that doc:
1) On the first pass of things (making sure that squeeze was up to date before upgrading to wheezy), I did not comment out (I repeat, I KEPT) the following line in the /etc/apt/sources.list.d/sources.list:
2) On the second pass of upgrade commands (when modifying the sources.list to have wheezy sources), in addition to adding the Debian wheezy sources, I modifed the turnkey line to read:
Of course, during prompts that talked about configuration files (replace=y,keep=n), I always chose "N" to keep the current configuration files.
After I went through that process, all went well. Here is my "lsb_release -a" output.
Reboots fine. And everything appears (5 minute check here) to be running fine. Incidentally, it appears that "opsi" that I installed on squeeze (www.opsi.org) is still working too. Definitely do this in test first. And my testing of this process took a couple of hours to complete. Hence, I should have attempted with subsituting "us" for "de" in the source URLs to debian.org.
- NSparks
I just tried this, as support for Squeeze just ended...
All was going good until I ran into an issue where Grub would not install on any of the 3 locations given during the dist-upgrade process. I can say continue anyways & have a sucessful upgrade. At that point I need to somehow reinstall/configure Grub, as the OS is still stable. If I reboot, the server will not boot back up; & I have to revert to the snapshot I took right before upgrade & start over. Any help would be appreciated.
Thanks,
Jason
Jason
What have you tried? What is the error message you get?
E.g.:
And all should be well. If not, please share the error you get.I am getting errors...
Errors are below. I think this may have something to do w/ LVM being used in the original TKL appliance. I even added a 5GB new virtual disk to try to install grub; & I get similar errors.
Error... (sorry I forgot how to properly quote output/code in this forum)
root@fileserver ~# fdisk -l
Disk /dev/sda: 536.9 GB, 536870912000 bytes
4 heads, 32 sectors/track, 8192000 cylinders, total 1048576000 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0000b15e
Device Boot Start End Blocks Id System
/dev/sda1 * 1 1000000 500000 83 Linux
/dev/sda2 1001472 39061503 19030016 8e Linux LVM
/dev/sda3 39061504 1048575999 504757248 83 Linux
Disk /dev/sdb: 5368 MB, 5368709120 bytes
255 heads, 63 sectors/track, 652 cylinders, total 10485760 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x5cea82e9
Device Boot Start End Blocks Id System
/dev/sdb1 5221125 10474379 2626627+ 83 Linux
Disk /dev/mapper/turnkey-root: 535.1 GB, 535121887232 bytes
255 heads, 63 sectors/track, 65058 cylinders, total 1045159936 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000
Disk /dev/mapper/turnkey-root doesn't contain a valid partition table
Disk /dev/mapper/turnkey-swap_1: 536 MB, 536870912 bytes
255 heads, 63 sectors/track, 65 cylinders, total 1048576 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000
Disk /dev/mapper/turnkey-swap_1 doesn't contain a valid partition table
root@fileserver ~# grub-install /dev/sda
/usr/sbin/grub-setup: warn: This msdos-style partition label has no post-MBR gap; embedding won't be possible!.
/usr/sbin/grub-setup: warn: Embedding is not possible. GRUB can only be installed in this setup by using blocklists. However, blocklists are UNRELIABLE and their use is discouraged..
/usr/sbin/grub-setup: error: will not proceed with blocklists.
root@fileserver ~#
Jason
Update...
I get a different error when attempting to install grub on a seperate virtual disk. This is the 5GB disk I just added to try to fix my issue. Here is what happens...
root@fileserver ~# grub-install /dev/sdb
/usr/sbin/grub-setup: warn: Attempting to install GRUB to a disk with multiple partition labels or both partition label and filesystem. This is not supported yet..
/usr/sbin/grub-setup: error: embedding is not possible, but this is required for cross-disk install.
root@fileserver ~#
Jason
Hmm, ok. A few more questions...
Can you please post the error messages that you were getting trying to install grub to the original vHDD? You gave the output of fdisk, but not of install-grub! :)
Also I have a hunch, I'm not sure if it's a good one, but can you try the following commands and give me the output:
If that gives an error message, that's ok but you may wish to remount /boot if you want to try other things.As an alternative approach, using TKLBAM you can just migrate your data to a new server. You will need to manually make some adjustments to get it all running but we have a doc page which provides a suggested workflow and notes a few gothchas which you may encounter.
I am not in the office now to
I am not in the office now to try your suggestion, but I did post the output of the install-grub in a previous post. Here it is...
root@fileserver ~# grub-install /dev/sda
/usr/sbin/grub-setup: warn: This msdos-style partition label has no post-MBR gap; embedding won't be possible!.
/usr/sbin/grub-setup: warn: Embedding is not possible. GRUB can only be installed in this setup by using blocklists. However, blocklists are UNRELIABLE and their use is discouraged..
/usr/sbin/grub-setup: error: will not proceed with blocklists.
root@fileserver ~#
I will be back in the office on Wednesday to try your suggestions. I agree not instaling grub on the newly added 5GB disk, i was just shooting from the hip there after many failed attempts.
Thank you,
Jason
Thanks...
It makes me suspect that you had grub-legacy installed previously. If so, perhaps that's still there and you can leverage that to bootstrap grub2. Maybe try installing grub to /dev/sda1?
Failing that, I suspect that you'll need to move all your partitions and leave a small gap at the start of the drive. It's doable but will be a PITA.
IMO whilst migrating your data won't be quite as quick as an in place upgrade should be, it will give you the latest build of TurnKey & will avoid this issue...
If this appliance is one of the more generic ones (such as Core or LAMP) then it should be pretty straight forward. Other appliance upgrades may get a bit more involved...
If you go that route, please start a new thread (feel free to link to it here if you want). That way we can keep this thread on topic. I get notifications of all new threads and posts so I'll see it...
I'm leaning as well as having
I'm leaning as well as having to move all partitions and leave the small gap at the start; if so then I will probably get the latest TLK appliance & rebuild.
No go on installing grub to /dev/sda1...
root@fileserver ~# grub-install /dev/sda1
/usr/sbin/grub-setup: warn: Attempting to install GRUB to a partitionless disk or to a partition. This is a BAD idea..
/usr/sbin/grub-setup: warn: Embedding is not possible. GRUB can only be installed in this setup by using blocklists. However, blocklists are UNRELIABLE and their use is discouraged..
/usr/sbin/grub-setup: error: will not proceed with blocklists.
This appliance is an sftp server for us & I just may need to rebuild it from scratch.
I'll try a few more things, then update this thread.
Jason
I am starting over...
Got the latest TKL FileServer appliance up & running & going to rebuild this application. This was the lesser of the two evils. Still not a bad solution, won't take too long; & its always nice to have a fresh slate from time to time.
Jason
Add new comment