Two installations of Debian (Jessie) on same HD? [Solved]

[davidcarass]

I tried it, and I do have two different boot options in grub. I can boot into either (I've created different usernames and root passwords for both) but my first install (now second option in grub) displays the following on startup:

"A start job is running for..." for 90 secs ...' then "Dependency failed for
/dev/disk/by*" and "Dependency failed for Swap"

* I don't know how to get boot error messages, did some research but I cannot find any log file in /var that displays this info.

But in order to install a second version of Jessie, it seems I have to have another root partition. Wouldn't that also be true if I went for a dual boot with Jessie and another distro? Isn't this the crux of the problem?

My partition scheme for the first Jessie install was:

/
/boot
/usr
/home
var/
/tmp
swap

I purposefully left unallocated and unformatted disk space when I used the debian installer in the above scenario.

In the second install, it was clear that the installer couldn't proceed without a root partition, so I allowed the installer to create this using all the remaining unallocated disk space. This felt counter-intuitive, since I now had two root partitions on the same drive - but I presume this is necessary when dual booting two linux - regardless whether they are the same distro or different, the same versions or different?

What I have now is what I wanted - provided I'm prepared to wait 90 seconds for the boot process to continue. I wanted to have one version of Debian for experimenting with different software. I tried virtualbox, but found the guest OS too slow for serious work (need a RAM upgrade I guess).

I wondered if the error message on boot is something I should worry about? If there was a work around? The second install of Debian Jessie I performed gives no such errors. Both installs 'seem' stable - though early days.

Thanks in advance.

Dave

[kenwong]

The issue arose because the UUID of the swap partition changed as a result of the second installation of Linux on the same drive. Updating the UUID in /etc/fstab will fix it.

To get the UUID of the swap partition, use the blkid command.

[kiyop]

Did you (or the debian installer) make the preformed swap partition used by the second debian at the second debian installation?
If so, do as kenwong suggested.

Do you use hibernation? If yes, I suggest you preparing 2 different swap partitions and making each of them used by each of the 2 installed debians.

But in order to install a second version of Jessie, it seems I have to have another root partition. Wouldn't that also be true if I went for a dual boot with Jessie and another distro?

Yes, it is true. You did right thing. You should prepare each root partition for each linux. (Maybe my English is not good.)

My partition scheme for the first Jessie install was:

/
/boot
/usr
/home
var/
/tmp
swap

Why did you prepare so many separate partitions? Do you know the advantage and disadvantage of separating partitions?

I wondered if the error message on boot is something I should worry about?

You need not worry about the error message, if the first debian need not use swap partition.

If you prepare 2 swap partitions in future, read below.

With default configuration of systemd, all the detected swap partitions are automatically mounted.
To disable this, execute with root privilege,

Code: Select all

systemctl --type swap

and check the UNIT of the swap (for example, dev-sda11.swap) and

Code: Select all

systemctl mask dev-sda11.swap

Reference:
https://wiki.archlinux.org/index.php/sw ... by_systemd

[davidcarass]

The issue arose because the UUID of the swap partition changed as a result of the second installation of Linux on the same drive. Updating the UUID in /etc/fstab will fix it.

To get the UUID of the swap partition, use the blkid command.

Thank you kenwong for this useful response. I did wonder whether the problem was having two root mount points or swap - or both. UUIDs are new to me, so I'll do some more research. I presume you mean that I should make the UUID for the swap partition of the first install to be the same as the UUID for the swap partition in the second install.

Note: I did not create a new swap partition in the second install (although I did when I first tried this out), More research needed, but you have given me enough to play with. I came across this:

https://liquidat.wordpress.com/2013/03/ ... d-to-know/

What do think?

Dave

[dasein]

I presume you mean that I should make the UUID for the swap partition of the first install to be the same as the UUID for the swap partition in the second install.

What kenwong quite correctly said is that you need to check and/or edit /etc/fstab in both installs, to make sure that both files refer to a single UUID for swap--the current UUID of the swap partition. You need to at least review both, and change (at least) one.

[sunrat]

You can find the UUID of your swap partition by doing (as root) -

Code: Select all

# blkid |grep swap

Check that it matches the entry in /etc/fstab.

[davidcarass]

Success - thank you all for the assistance and clarification.

Some useful info for future reference.

Best regards - Dave

[davidcarass]

I tried it, and I do have two different boot options in grub. I can boot into either (I've created different usernames and root passwords for both) but my first install (now second option in grub) displays the following on startup:

"A start job is running for..." for 90 secs ...' then "Dependency failed for
/dev/disk/by*" and "Dependency failed for Swap"

* I don't know how to get boot error messages, did some research but I cannot find any log file in /var that displays this info.

But in order to install a second version of Jessie, it seems I have to have another root partition. Wouldn't that also be true if I went for a dual boot with Jessie and another distro? Isn't this the crux of the problem?

My partition scheme for the first Jessie install was:

/
/boot
/usr
/home
var/
/tmp
swap

I purposefully left unallocated and unformatted disk space when I used the debian installer in the above scenario.

In the second install, it was clear that the installer couldn't proceed without a root partition, so I allowed the installer to create this using all the remaining unallocated disk space. This felt counter-intuitive, since I now had two root partitions on the same drive - but I presume this is necessary when dual booting two linux - regardless whether they are the same distro or different, the same versions or different?

What I have now is what I wanted - provided I'm prepared to wait 90 seconds for the boot process to continue. I wanted to have one version of Debian for experimenting with different software. I tried virtualbox, but found the guest OS too slow for serious work (need a RAM upgrade I guess).

I wondered if the error message on boot is something I should worry about? If there was a work around? The second install of Debian Jessie I performed gives no such errors. Both installs 'seem' stable - though early days.

Thanks in advance.

Dave