The fix is really simple, chkconfig wants to see 2 lines in your apachectrl that tell it what it is and what to do with it. At that point, chkconfig will recognize the service as your script name. So here it is:

1. Add the following lines to the top of the apachectrl script that you’ll find in the apache2/bin directory of your install.#!/bin/sh
   #
   # chkconfig: 35 85 15
   # description: Apache is a World Wide Web server.
   #
   # Licensed to the Apache Software Foundation (ASF) under one or more

   The first line:
   # chkconfig: 35 85 15
   The 35 tells chkconfig you want the service running at run level 3 and run level 5.
   The 85 is the start priority.
   The 15 is the stop priority.

   The second line:
   # description: Apache is a World Wide Web server.
   This is merely a description of the service.
   *note* Both lines start with a comment “#” (pound sign).

2. Once your lines are in place you can move or link your apachectrl to /etc/init.d/httpd
3. Once the script is in /etc/init.d/ you can run the following command to make it all go:
   chkconfig httpd on(this is of course assuming that you named the script in /etc/init.d/ “httpd”, if you named the script “chuck” then you’ll run: chkconfig chuck on, I think you get the point).

That is all.

I have 2 machines.  My main Ubuntu box which has a ton of storage on it, and a crappy old gateway thats got about a 20GB disk.

I do a clean load of whichever Ubuntu version I want to image on the gateway.

Next, I add any custom packages, network keys, etc etc, to the newly loaded box.

Once I have it like I want, I reboot the system to a recovery cd that you can download from this site .  This is a good image in that it has the tools we’ll need to make partimage work.

On my main server I create an NFS share to /mnt/images and export it.

Once booted to the recovery CD, I first add an ip address to eth0

root@sysresccd: ifconfig eth0 192.168.2.2

Then I mount the share on the main box

root@sysresccd: mount -t nfs 192.168.2.1:/mnt/images /tmp/images

Once this is complete, I first take a copy of my  MBR like so:

root@sysresccd: dd if=/dev/sda of=flash-sda.mbr count=1 bs=512

Next, I make a copy of the partition table:

root@sysresccd: sfdisk -d /dev/sda > flash-sda.fs (Note: sfdisk is a utility that comes by default on the resource disk refereced earlier)

At this point, we have the data we need to recreate our master boot record as well as our existing partition tables.  We now need to get images of the data.  This is where partimage does it’s thing.  To initiate the gui for partimage simply type partimage  at the recovery CD prompt.

Once you’re in the GUI, you should see the slices that are available on the disk.  In my case I have 3 sda1, sda2 and sda5.  This is the default configuration of an Ubuntu 9.04 install.  I  can see from the gui, that one slice is the majority of my disk, 18.25G, and is labeled as ext3fs.  This is the slice that is holding all of the data.  The second slice is labeled extended and has no file system info, and the third slice is labeled as swap.  All I need is an image of sda1 and I’m good to go.

In the GUI, I need to make sure that sda1 is highlighted, then tab to the “Image file to create/use” text box.

In this box I will put the name of the image that I will create of sda1.  I’ll use a path to the nfs mount so the image will be saved to my secondary machine, just as my mbr and fs information was.  Then hit F5 to continue

On the next screen I choose a compression level of “None”, because it’s the only happening on my LAN, and it’s fast.  Then hit F5 again to continue.

The app then asks for a description of the image.  Add whatever you like here.

Depending on the size of your slice, this next step might take some time.  Once it completes, you will have all the info necessary to do a bare metal restore of the machine.

To test, all you need to do is use fdisk to destroy the existing partition table.

From the sysresccd command prompt type in: fdisk /dev/sda  (or whatever your disk is).

From here, you’ll see a prompt similar to the one below:

From this interface you can use the “p” command to print the existing partition table:

Our goal here is to make it a blank disk, clear of partition information and then write our changes.

Step One is to delete what is there, the following screen shot shows this process, I delete all of the primary partitions, print to verify and then write the changes with the “w” command.

Now we are going to use the previously created MBR and FS files to recreate the partition tables.  The following screen shot shows this process.

1.)  Use DD to rewrite the master boot record.

2.) Use sfdisk to import the partition data.

At this point, you can check fdisk and see that the disk is back in the condition that it was after the Ubuntu load.

The last step here will be to reimage sda1 with the data from the partimage backup.  We’ll want to start partimage again from the recovery disk:

1.) Make sure your restore slice is highlighted, then tab to the file name textbox.

2.) Add the filename of the image that is on the NFS share.

3.) Make sure that you move the * to “Restore partition from an image file”.

Press F5 to continue.

In the following dialog, if you are confident in what we have done to this point, there is no need to change anything, just press F5 to continue and the image restore will begin.

Once the restore finishes, press enter on the “OK” and then reboot your box.  Try to hold back the excitment when it comes up exactly as you left it.