Nick Pikarsky's picture

help. I have two systems that are virtual lamp appliances running on a VMware ESX server 4.0.0, they have been running for the past 6 months with no issues. Today I shutdown the appliances via the console, and  they will not reboot. The last message on the console boot screen is no init found try passing init = bootarg.

It then drops into BusyBox v1.13.3 shell.

How can I fix??

Thanks in advance


Jeremy Davis's picture

You'll need to find out what the LVM is called. OTTOMH I think it's called something like dm-0, perhaps try something like:

fsck -t ext4 /dev/dm-0

If that doesn't work then my memory isn't as good as I'd hoped and you'll need to do some research. I think the lvscan command should tell you the name of any LVMs.

There's a pretty cool (although a little dated) LVM tutorial on HowToForge here which may be helpful (don't just blindly type commands in though unless you understand them or you can destroy your data). There should also be plenty of info found on the Ubuntu forums and/or google.

Jeremy Davis's picture

ie what you were trying to do intially. Hopefully that should fix it... Otherwise I'm all out of ideas.

Jeremy Davis's picture

Firstly fsck needs to be run without the filesystem mounted. Generally this is done via a live CD or similar. I suspect that is your issue. I think it can be scheduled to run on boot (similar to how Win runs chkdsk on boot) but OTTOMH I'm not sure how to do this, perhaps google will be your friend!? :) (or just use a live CD/ISO)

Secondly, assuming you are using TKLBAM, It is probably easier (and less prone to error) to just launch a fresh VM and restore your data there. Don't destroy your old instance though until you are 100% sure that all your data is there and it all works as expected.

L. Arnold's picture

I don't know about the difference between ESX and ESXI but I have been running 4.0 ESXI without problems.

When you create a new Server from withing VSphere Client you need to ultimately get through the install and either first be able to get a DHCP IP, or go to Network Settings and configure a STATIC IP manually.  This can be a little tricky if you have multiple Network adapters but my experience is to simply configure the first Network adapter with all the gateway and dns settings you need.

Once you do that you should be able to access your WebShell or Webmin from a browser.

If this is not working see if you can install a Windows Instance from an ISO  (You can build an ISO with ISODisk, then upload that to your ESX Datastore via VSphere Client.  If you can get Windows Running with an appropriate IP you can get TKL running.  If you cannot then your issue is within your ESX setup, which I think, generally, is the case here.

Good Luck.

I have same problem see my post

I am getting error after second line

mount -n -o remount,rw /

mkdir /var/run/network /etc/init.d/networking restart

cannot create directory /etc/init.d/networking

Any Ideas



Jeremy Davis's picture

The vHDD is corrupted (common when not shut down properly, also seems to be quite common when the underlaying FS (on the host) is a RAID config... My inclination would be to install (from ISO - rather than VM image) and avoid LVM (whilst it is a great thing IMO it adds an extra layer and increase possibility/probability of issues...).

TBH I'm not sure whether it's just because ESX/ESXi is so popular, whether it is closed source/proprietry, whether it's just not totally compatible with Linux/Debian/LVM or whether it is just a bit flakey (or a combo of above) but users seem to have a disproportionate experience of issues with TKL combined with ESX/ESXi.

Personally my solution would be to ditch it and go for a (IMO) better hypervisor such as Proxmox... :p I have been using Proxmox in multiple locations for years and had little to no issues with TKL (and I can run servers as OVZ containers where they run like bare metal - ~5% loss to hardware abstraction...)

Add new comment