Due to the main previous archiving script falling over itself I have started to write a new version with the help of #bash on freenode.

Conceptual things I have changed:

• Using a whitelist of UUIDs instead of mountpoints  to determine if a drive is plugged in.
• Keyfiles instead of password so an expect script is not required.
• Tar instead of rsync, even with limitations rsync was playing havoc with asterisk

This version uses gnupg instead of truecrypt, but I have plans on seeing if truecrypt can mount without requiring any sort of expect script. Also the script doesn’t have incremental archiving implemented, but tar does support incremental archiving.

In short expect a different version to emerge soon.

#!/bin/bash

#UUID used for drives (seperate drives using " " )
drives="9a3ffd9e-ee84-4475-ab0f-60b98910eeef dd519d4f-a63b-4579-bcc7-0feb073ba030"
mntloc="/mnt/bkdrive"

#check if drive is active: fuser ...

#step one umount folder
echo "attempting to unmount $mntloc"
umount "$mntloc"

#finds first acceptable uuid
found="0"
for i in $drives; do
    #sets the dev location
    devloc="$(blkid -U $i)"
   #checks drive is found
    if [ -n "$devloc" ]; then
        echo "suitable drive found. (UUID=$i $devloc) ..now mounting.."
        found="1"
        break;
    fi
done
#if drive not found. exit
if [[ ${found} -eq "0" ]]; then
    echo "Warning no suitable drive found!"
    exit 1
fi

#check mounting worked.
if (mount -t auto "$devloc" "$mntloc"); then
   echo "mounted.";
else
   echo "Failed to mount. Exiting."
   exit 1;
fi

#backup /ANAT/Common in a screen
file="$mntloc/ANATCommon_bk_$(date +%Y-%m-%d).tar.xz"

#screen -S Backup -d -m 
tar -cvJ /ANAT/Common | gpg -r [email protected] -e -o "${file}.gpg"

The situation

For a while now I have been considering changing from using Ubuntu server, whilst Ubuntu server does provide great support with updates.

Distribution upgrades are known to fail with Ubuntu and with a cycle of 2 distribution releases per year, it is not that feasible to be able to be upgrading by installing from scratch all the time.

Ubuntu does offer some other advantages such as a GUI  (gufw) for iptables which is not outdated and a pretty good community of support.

Why Debian?

The idea of using debian is centered around stability with updates. Distribution upgrades with Debian are known to work, distribution releases tend to have a much longer cycle than other linux distributions but at the same time it takes a lot longer too.

It might be argued that openSuse, fedora and arch linux are also known to work with a distribution upgrade. However, both fedora and arch linux tend to be more focused on cutting edge software than stable software and openSuse is no longer being supported by novell which may have a long term negative effect on the linux flavor as a whole.

By the way gentoo is out of the question, because hosting a server with the ability to compile is a bad idea from a security point of view.

So by a process of elimination of “core” linux flavors is the reason for choosing Debian.

How to migrate?

To migrate, I will have to use the backup server (this server only exists just incase the hardware on the existing server fails).

Firstly I will install all pre-existing and required software packages from the exisiting server.

Secondly I will take a dump of the /etc directory on the existing server, as the exisiting server is running ubuntu in theory the configurations should be similar enough to just dump the /etc folder into a debian system.

Finally I’ll transfer the data across.

Problem

Trying to add the embed code for something like YouTube or Vimeo gets stripped out by WordPress’s editor. On top of that, the standard embed option of leaving the URL to the website on a line of it’s own gets screwed up if you edit a published web page as the various text filters add the paragraph <p> </p> tags around the URL, meaning it’s no longer on the line on it’s own, and instead of being an embedded video it’s now

Fix v1 : Shortcode

If you simply want to embed videos from well known sites like YouTube, then you can use one of the plethora of WordPress plugins which use some sort of short code, like |YouTube#84631| to insert the embeded video.

Fix v2 : Allow the Embed Code

This process involves allowing the HTML tags used in the embed code into the HTML tab of the TinyMCE WordPress editor.

Step 1. Install the TinyMCE Valid Elements plugin.

Step 2. Add the following :

# embed
* allowfullscreen, allowscriptaccess, height, src, type, width

# iframe
* frameborder, height, marginheight, marginwidth, scrolling, src, width

# object
* embed, height, name, param, src, type, value, width

# param
* name, value

Step 3. Go to the video page and copy the embed code.

Step 4. Go to edit a page or post and do a [Shift] + Refresh [F5], this clears the browsers cache.

Step 5. Select the HTML tab of the page or post, scroll to the correct location and paste in the embed code.

Step 6. Preview the page and update or publish it if correct.

Note : When getting the embed code you may find the size of the embedded video is larger than the size of the theme’s column and you’ll have to use a smaller size.

Found this snippit of code and thought someone else would find it usefull

This was used to rename a bunch of log files with inconsistant naming schema to all lower case and with a .log extention instead of _log.

#!/bin/bash
### Rename all uppercase directories to lower case ##
# NB : Addapted from from http://www.linuxweblog.com/node/306
 
for x in `find * -type f -maxdepth 0`;
do
  # Translate Caps to Small letters
  y=$(echo $x | tr '[A-Z]' '[a-z]');
  if [ "$x" != "$y" ]; then
    mv $x $y;
  fi
done
 
for x in `find * -type f -maxdepth 0`;
do
  # Rename log files using .log instead of _log to conform to new naming schema
  y=`echo $x | sed "s/_log/.log/"`
  if [ "$x" != "$y" ]; then
    mv $x $y;
  fi
done

Enjoy

The ANAT Tech Department