System unbootable after yesterday's upgrade.
 

Yesterday I upgraded my main Unstable Debian PC, including systemd which I had put off for a couple of days (the removal of sysv-init scared me a little).

However the systemd upgrade worked on another PC so I included it in yesterday's rather large upgrade (>300 packages)

However now, when I boot, I get a message:
After a few seconds the following line is repeatedly printed
Occasionally I have been able to stop the scrolling to see:
Could somebody please tell me what's wrong and how to get out of it?'m in over my head and Google hasn't helped!

Thanks

If it were me, I would reinstall the operating system, but not before I took out the hard drive, hooked it up as a slave to another system and made a backup of everything I needed on that drive. During the reinstall, you should be able to tell the install to spare at least /home, since it is doesn't contain files the system needs to boot.

I've thought about that. It is what I consider the "microsoft solution".

This Debian installation is nine years old and quite possibly a reinstallation will clean up a lot of clutter. However I prefer to do this on my own terms, not because I have to because of external circumstances.

So thanks for the reply but I'll hang on for a little time in case somebody can at least tell me what went wrong.

"looks" like you got an unwanted upgrade to NFS/rpcbind that is causing some issues. I suggest finding the offending package and downgrading it to the previous version.

Is there a reason you are using the unstable branch? Generally that's for developers for testing and is broken often.

I moved to unstable from testing a long time back when something I needed badly was only available in unstable.

However I find that I can sometimes make my own little contribution by providing feedback and bug reports.

Regarding the "unwanted update" I have got some more symptoms and info together and will post a modified question tomorrow - too tired tonight, my 18 month old grandson was here today:D.

Thanks for the reply.

That is very kind of you. Communities are always looking for good feedback like this.

Congratulations friend. Go easy.

http://www.sabayon.org/download
http://www.slackware.com/getslack/
http://www.gentoo.org/main/en/where.xml
http://crux.nu/Main/Download

All perfectly good alternatives to distros infected with the systemd bloatware.:D

As to fixing your problem, I haven't a clue, I am using an actually sensible distro with a legit init system.

@vl23: This is a technical thread where somebody seeks a solution for a problem with his Debian system. If all you have to offer is not constructive bashing of an init system then don't post.

@davcefai: Is it possible that you either boot from a live-CD containing journalctl or to the rescue.target of the installed system, so that you have access to the logs to get better information what is going on?

Solved
 

I solved it!

I narrowed the problem down to the existence of NFS mounts in my fstab file. Removing those removed the problem (and vice versa!)

It then turned out that a bug report had been filed against systemd.

https://bugs.debian.org/cgi-bin/bugr...cgi?bug=746358

The patch given in the report works.

szboardstretcher and TobiSGD, thanks for the input.

vl23, I'm sure that somebody, somewhere, finds you useful.

Going forward, you might want to mirror your system in a virtual machine and do these upgrades on the virtual machine first to see how they work out, before applying them to your actual system.

Xode, your suggestion makes perfect sense but is unfortunately impractical. I have 3 machines to maintain and, in Unstable, there are daily upgrades. Not all 3 PCs react the same to a bug. In this case, for example, only one of them has NFS mounts specified in fstab.

In practice I always hold one of them back from an upgrade for a day or so so that I always have some means of communication working. Then when the refuse hits the air impeller I can use the at least one working machine to sirt out the problem.

Also Knoppix is a life saver. Often working around a problem involves editing conf files so boot with Knoppix, edit, reboot and see if it works, rinse and repeat.

Living on the edge is never boring!

Just a thought: current computer hardware and the latest version of VMWare Workstation might just be up to snuff to allow you to virtualize all 3 of your current physical machines under one stable linux host system (e.g. SUSE 12.X or 13.X).

If you want to try out my suggestion here, I would suggest the following:

(1) Build a completely new computer and put a stable up to date linux host system (e.g. SUSE 12.X or 13.X) on it. Given that you will have 3 or more virtual machines on it, you probably want to have at least 2 hard drives and lots of RAM.

(2) For each physical machine with debian unstable that you have, install an equivalent virtual machine on the new stable linux machine.

Linux (stable, that is) as a host system for virtual machines is wonderful. Apart from being rock solid stable, you can tell it to just host resources (e.g. network cards) for the virtual machines and otherwise ignore them.

Perfect, I had the same problem and remove the NFS mount in my fstab has solved it!

Thanks davcefai

Not sure why you had to "upgrade" to systemd? I believe there is still a choice of init system in Debian with systemd being the default?

That may be true for Jessie, but no one can say if it will be the same for Jessie+1.