There is an exciting expo happening next week. It’s the Cloud Expo Europe 2009, in London and it’s FREE if you pre-register.

Now we’ve all heard of the term cloud (in the computing sense), but I suspect many don’t really know what it involves. So this is your chance to find out.

As with many conferences there will be stands for companies and projects related to the event. There will also be talks from field pioneers and other contributors- in both business tracks and a technical track.

I’ve been invited to talk about Cloud offerings regarding Ubuntu, and will be talking about Eucalyptus (no, not the plant), the Cloud suite that is the next big thing. This replicates the Amazon EC2 and other offerings in a free and open source software.

Also, the Ubuntu community have been offered a stand. If you want to help man this, it would be appreciated! Let me know if you are available.

Looking forward to seeing you there.

For more information:
http://www.cloudexpoeurope.com/

This chap isn’t someone i’ve seen to be especially athletic nor a ’spring chicken’ (i’m sure he’ll love me for saying that) - which makes him quite unlikely to choose to ride 234 miles in 3 days. Even if someone is a keen rider, this is still quite a distance to cover in such a short time.

So why would he choose to do this?
We’ll he’s doing it to raise money for a rather special charity called “Big Issue”. Now this charity is a little bit different to some of the ‘traditional’ charities. They help people who are less fortunate, often homeless people by allowing them to sell, what many consider, to be a quality magazine, and these people earn from the proceeds.

He made a video which shows him training (hopefully you can see it below):
httpv://www.youtube.com/watch?v=ZMO6Iyno3BU

As I said, he’s rather unlikely to elect to do this, but he has been training hard. He’s done somewhere in the region of 700 miles since Christmas, around 90 miles per week. This training is gradually rising to 200 miles per week, as he gets nearer the event.

The worrying thing is, unless he can raise £1300 by May 23rd (10 days away!) - he can’t take part. This would be such a shame, as he’s put in so much effort to date and it would be an awful for the charity to miss out.

So, I wouldn’t normally ask - but please consider donating - even if it’s just a small amount, to help him take part, and complete his personal challenge!

Find out more, and sponsor online (paypal supported) via:
http://www.justgiving.com/chrisdakin

I’m sure a “during” video would make your donation worthwhile!

This was our first effort of having an event like this, and it was a huge success. As the UK is larger than people think, we actually held two venues, one was held in Birmingham and the other in our countries capital - London.

I traveled to Birmingham Thursday night, in preparation as it was a number of hours away. Luckily i did, as it mean’t I was able to attend the Birmingham Linux Users Group, who had the guest speaker of Scott James Remnant speaking about one of his pet projects, Upstart. This was really worth attending!

The Jamming started on Friday in the Birmingham venue, and in many ways I was kind of worried about having one during a working day - as not everyone can get away from employment commitments during the week. However, I was proved wrong - and was actually rather pleased with the turn out! It never ceases to amaze me how so many great people are willing to give up their free time to *want* to help make Ubuntu better.

As we had two venues, we eventually got a working video and sound link between the sites - things like this are really useful for a couple of reasons, primarily it was great to see the ‘other’ site at work on the large LCD television on the wall - but also it made it possible to share talks across sites. As we didn’t enable this until quite late on in the event, it wasn’t used as much as it could have been - however, it’s really something we can get working for the next one! I would also encourage other countries to consider this also, even if it’s perhaps snapshot pictures beaming across on the internet - just for others to see. It’s great watching others at work, surprisingly encouraging!

Both events were kicked off with me giving an Introduction to Launchpad, some people at both events hadn’t really used Launchpad (or in some cases heard of it), others believed that as they weren’t experienced developers (just users), that they were not entitled to have an account on Launchpad. This is a shame, and I hope the feeling isn’t widespread. The introduction talk I gave, covered the basics of launchpad including bug reporting, triaging, advanced triaging and then went onto fixing the bug by using the supplied information on the report to create a ‘debdiff’ and upload to the site pending review. I made it clear that someone didn’t have to work to the end, and each part of the process was really useful in itself. I’m pretty sure that to some people there, it wasn’t that useful, but I think it was generally useful to ‘get the day started’, and allow less experienced users to know where to get started. Following this the rockin’ Martin Meredith (Mez), showed how to locate bugs to continue work on, with the intention of getting them fixed, packaged and uploaded.

The London Jam started on Saturday, this mean’t I had to get a train from Birmimgham to London Saturday morning which shouldn’t be too bad. However, London Transport decided it would be a good weekend to have major maintenance work making delays and rerouting quite a problem. However, i got there eventually.

Following my second introduction to Launchpad, the ever awesome James Westby gave a presentation on fixing a package based on a patch and uploading it. This was great on many levels, as for me, I found it useful to see someone elses work flow and other ways of using tools I wasn’t previously aware of.

We were also quite lucky to get a GNOME developer turn up - Simos Xenitellis, which was mutually useful as we were able to show him some of the more advanced features of Launchpad, and he was able to show us around the GNOME Bugzilla and between us we explored advanced triaging methods and linking to upstream bug reports, he was really keen - which was great to see.

For me, the aim of both events was to meet as many people as possible, and pass on knowledge to as many as I could, as well as learn a thing or five for myself. It wasn’t necessarily to just churn out touching of as many bugs as possible, it was to equip others with the knowledge and skill to be able to go home, and continue working on them throughout the year. Otherwise, we would all probably have sat in silence trawling launchpad and working as if we were on our own anyway, saying that I did get the odd message on IRC from across the table!

At the end of both events, i asked for feedback and it was really encouraging. Will we hold another Jam in around 6 months? Hell yeah! I have some ideas for some things we could improve on, but as this was our first effort I think we did really well!

So a big thank you (and hugs), to everyone that turned up, even if you chaps feel you didn’t achieve much on the day(s), you really did! Every little counts. Infact, the success of this reasonably technical event has given me the confidence to explore organising more events, including more social aspects. Especially as we did pretty well in the Bug Jam Stats.

A special thanks to Linux Emporium for providing us a venue for the Birmingham Jam, Martin Meredith (Mez) for allowing me to stay at his digs - as I was a guest in Birmingham. Also James Thomas (selinuxium), for getting his employer to allow us to use their lovely offices in the centre of London.

If you see yellow snow, no matter how free or appetising, don’t eat it.

As you can guess, most of the UK has been getting snow this week. As ever, we’ve practically come to a halt - with an estimated 25% of employees staying at home, doodling in the snow.

For those that don’t know what Twitter is, I define it an “open chat” method, where you can ‘microblog‘, and really just let the world know when you scratch your bottom.

Now to the point of this post; Some crazy British chap, whom i suspect has drunk too much Newkie Brown, has decided to try and travel as far around the world as he can can in 30 days, surviving on hand outs from the twittering masses. He’s doing this with the hope of raising money for a good cause/charity. I won’t try and explain the whole shebang, as he describes it pretty well on his website http://www.twitchhiker.com.

However the real challenge, and what prompted this posting is the sub-challenge he’s set. If he can get 2000 followers by the end of today (02/14/09), he “will walk out in the snow, in my pants, and dance“. Well as we all know kids, without a photo or video - “It didn’t happen” - . So help some crazy Brit’ make a fool of himself, follow him now (Oh and consider helping him on his little challenge thingy)

I was asked how I solved Challenge 2, i did it largely the same way as the example given when the winner was announced, with some variation of the script. So there didn’t seem much point.

This challenge I decided i would document how I did it.

Challenge 3:

So what do we know?

• An encrypted “mount passphrase” inside 128-bits of random data, symmetrically encrypted using a wrapping passphrase and the standard eCryptfs salt
• The MD5 hash as a shadow of the wrapping passphrase
• The “challenge_3.txt“, which is the underlying data ecrypted using eCryptfs - basically the encrypted form of data that would otherwise be in your ~/Private directory.

Clearly the mount passphrase is the strongest link, as shown in challenge 2. As Dustin pointed out cracking 4 digits for the wrapping passphrase will take an estimated 40 days (based on 4.3 attempts a second), using multiple threads, on many servers with lots of cores will significantly decrease this - however, it’s still a long time for this challenge.

We also know the md5 hash of the wrapping passphrase, there has been recent publicity about md5 being insecure for passwords - however, there are various other aspects. This will be our vector for attack.

The MD5 Hash is quicker to brute force than the unwrapping of the passphrase, as documented in the solution challenge 2.

So how do we do it? The initial obvious method is to brute force, every range within the known constraints. These are, that the passphrase is 4 character alpha numeric string (a-z A-Z 0-9).

So, lets look at this script based on Dustin’s previous challenge:

#!/bin/bash
# crackmd5.sh

MD5HASH="74b87337454200d4d33f80c4663dc5e5"

CHARS="a b c d e f g h i \
j k l m n o p q r s t u v \
w x y z A B C D E F G H I \
J K L M N O P Q R S T U V \
W X Y Z 0 1 2 3 4 5 6 7 8 9"
for i in $CHARS; do
     for j in $CHARS; do
        for k in $CHARS; do
           for l in $CHARS; do
                  md5=$( echo -n "$i$j$k$l" | md5sum |  awk '{print $1}' )
                   if [ $md5 = $MD5HASH ]; then
                         echo "Cracked: $i$j$k$l"
                          exit 0
                   fi
done  ;    done  ;   done  ;   done

This will find the result by brute force, it’s SLOW. mainly because it’s a single process on one server, using one core.

The next method to attack this would be using MD5 rainbow tables, now not many people have these - but if you do *GREAT*. They are basically a table, making cracking of MD5’s really fast and efficient. To have all combinations “laying” around could easily consume a few terrabytes. So this isn’t really practical.

The next method is to ask your friend, that knows everything and will know the answer. For the benefit of this exercise we will call this friend “Google”. So inserting the md5 hash from the shadow file into Google, we get the answer.

Okay, so we have effectively recovered the wrapping passphrase, and so we can now unlock it. I’ll miss that step out as Dustin covered that in the solution for Challenge 2.

The plain text challenge_2.txt contained (amongst other things):

Start with the sha512sum of the following seed:
$ echo “Daemon” | sha512sum
Write the computed sum to the first line of a new file.
sha512sum that new file.
Write the resulting sum to the second line of the new file.
sha512sum that new file.
Write the resulting sum to the third line of the new file.
…
Continue doing this until you have exactly 1,000,000 lines in the file.

Hmm.. this sounds tricky. Let’s try doing this as described.. the /logical/ way. Generate the sha512sum of “Daemon” and output that to a file, then append the sha512sum generated of that file to the bottom of the file, repeat until file has 1000000 lines.

#!/bin/bash
# sha512brute.sh
echo "Daemon" | sha512sum |  awk '{print $1}'   > test.log
for ((i=2;i<=1000000;i+=1)); do
sha512sum test.log |  awk '{print $1}' >>  test.log
done

The above will work, but running on one server could easily take 70 hours - as the whole sha512sum is being calculated from scratch, even though only the last line of the text file is being changed. If only you could cache the calculation round so far, and append to the end. Well you can!

I hadn’t used Perl in anger recently, and decided to use it to solve this aspect. So i knocked up a script to do just this. Surprisingly,it was pretty trival in perl. With this new number, find out “Who own’s this number?”.

#!/usr/bin/perl
# Run as:
# ./genSHA512.pl > sha512sums.txt

use Digest::SHA; # qw( sha512 sha512_hex );

$keyword = "Daemon";
$sha = Digest::SHA->new(512);

$sha->add($keyword,"\n");
$digest = $sha->hexdigest;
print $digest . "\n" ;

for ($i = 2; $count <= 1000000; $i++) {
        $sha->add($digest,"\n");
        $digest = $sha->clone->hexdigest;
        print $digest . "\n" ;
}

This will generate the 1000000 line sha512sum file in around 15-20 seconds on an average!

Now lets extract the lines asked in the next phase of the challenge, convert them from hex to decimal, and subtract the difference.

#!/bin/sh
# ./sha512_process.sh
#
# Use the txt generated from genSHA512.pl
SHA512SUMS=sha512sums.txt

LastHex=$(tail -n1 $SHA512SUMS | sed -e 's/^.*\(......\)$/\1/'  )
LastDec=$( printf "%d\n" "0x"$LastHex )
echo  "Last 6 digits of the 1,000,000th line: " $LastHex " (Decmimal: ${LastDec})"

FirstHex=$( sed '961325q;d' $SHA512SUMS | sed 's/\([0-9]\{6\}\).*/\1/g' )
FirstDec=$( printf "%d\n" "0x"$FirstHex )

echo "First six digits of line 961,325: " $FirstHex " (Decmimal: ${FirstDec})"

diff=$(( $LastDec - $FirstDec ))
echo "      Difference = " $diff
echo "            Who's number is this?"
echo "                  * HINT: http://letmegooglethatforyou.com/?q=$diff&l=1  *"

So lets run them!

daviey@katana:~$ time ./genSHA512.pl > sha512sums.txt

real 0m15.567s
user 0m15.090s
sys 0m0.480s

daviey@katana:~$ ./sha512_process.sh

Last 6 digits of the 1,000,000th line: da7080 (Decmimal: 14315648)
First six digits of line 961,325: 561093 (Decmimal: 5640339)
Difference = 8675309
Who’s number is this?
* HINT: http://letmegooglethatforyou.com/?q=8675309&l=1 *

As we can see, this is very efficient (probably could be more so without saving to disk, but it’s certainly enough!), solving this part of the challenge in under 20 seconds!

So, what can we learn from this? If you plan to do something like this, don’t jump in too quickly, spend a good amount of time thinking through a good solution! Using a thought out solution, this challenge could be cracked in under 5 minutes!

I would like to say a big thank you to Dustin for setting this challenge, it is my understanding that he spent a great deal of time setting these tasks! I do hope he sets some more, and more people take part.

I guess i started to feel something was dwindling around 5 months ago, but didn’t really understand what it was until Jono pointed it out in his blog post, and some nattterings at the Ubuntu Developers Summit.

I feel sometimes we get too sidetracked with infighting, governance, procedure and councils to remember that we are all here (in the community) because we love the people, and we love what they do.

I think sometimes we need to remember that thousands of people make effort towards Ubuntu, largely to make things better for the community around them.  Ubuntu, for most, isn’t “work”, as in employment - it’s a hobby ( read addiction ;)  ).

One thing that truly amazes me about the Ubuntu community - it’s large, and with any large community you expect it to attract some people you really can’t stand to be around.  Surprisingly, this has never been the case for me.  Everyone i’ve actually met in real life, seem to be great people.

Why is the community full of ‘good’ people?
Does Ubuntu only attract ‘good’ people?
Is there, in actual fact, good in everyone?
Does Ubuntu bring out the ‘good’ in everybody?
Do only ‘good’ people want to get involved in the community?

Who knows.

So here it is, a hug for anyone reading this.  You rock my world, yes *you*.  In summary, the community rocks - let’s keep it fun.

You decide.

This is our chance chaps!  The UK hasn’t had an ‘big’ Linux expo for two years.  Sadly last year had to be cancelled.

I’ve literally /just/ had confirmation from the organisers that the Ubuntu UK LoCo has been reserved a stand in the .Org community village.  Hurray!

Two years ago it was a whopping success, and this year stands to be even better.  The actual expo will be co-existing in Olympia with the Mac expo, so it should give us a great chance to show off Ubuntu to people who wouldn’t normally attend a Linux Expo.

It seems Canonical haven’t yet decided if they will be able to attend.  I do know that the organisers did want a speaker to talk about Ubuntu for the main stage, so if you are that person - apply now!  The event could not have had better timing as it is around the same time as the next Ubuntu version, Intrepid Ibex (8.10) release!

Entry to both ‘.org’ exhibitors and attendees is currently free (£12 on the door), so apply here for your FREE tickets.

If you’re a UK Ubuntu Fan, join the community, come along!  If you fancy helping out on the stall, add your name to the volunter section of this page