atom.xml

<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">

  <title><![CDATA[James Jefferies and ShedCode Ltd]]></title>
  <link href="http://jamesjefferies.com/atom.xml" rel="self"/>
  <link href="http://jamesjefferies.com/"/>
  <updated>2013-02-17T17:30:06+00:00</updated>
  <id>http://jamesjefferies.com/</id>
  <author>
    <name><![CDATA[James Jefferies]]></name>
    
  </author>
  <generator uri="http://octopress.org/">Octopress</generator>

  
  <entry>
    <title type="html"><![CDATA[Updating WordPress performance]]></title>
    <link href="http://jamesjefferies.com/2013/02/13/updating-wordpress-performance/"/>
    <updated>2013-02-13T14:05:00+00:00</updated>
    <id>http://jamesjefferies.com/2013/02/13/updating-wordpress-performance</id>
    <content type="html"><![CDATA[<p>Vanilla WordPress performance isn&#8217;t great, each page has to be generated via PHP per view, which soon starts to bog your server down under load. Fortunately there are some plugins and tweaks you can make to improve performance and resilience.</p>

<p>Any PHP or Apache gurus out there, feel free to comment and let me know where I&#8217;m going wrong, I don&#8217;t claim to be a PHP environment expert!</p>

<h2>Caching plugin</h2>

<p>I&#8217;ve generally used <a href="http://wordpress.org/extend/plugins/w3-total-cache/">W3 Total Cache</a> which provides a number of different levels of caching, including the PHP itself if the relevant apache modules are installed. You can even delegate data to a CDN if that&#8217;s your thing.</p>

<p>But before installing the plugin, let&#8217;s get everything ready for optimising it.</p>

<h2>PHP Performance</h2>

<p>To get the full benefit of a plugin like Total Cache, you need to add a few modules to Apache and PHP.</p>

<h3>Install DotDeb repository</h3>

<p>The DotDeb repository is set up for Debian boxes with up to date versions of common packages.</p>

<p><a href="http://www.dotdeb.org/instructions/">Instructions</a> are straightforward and involve adding their repository.</p>

<h3>Install fpm</h3>

<p><a href="http://php-fpm.org">fpm</a> is a FastCGI process manager.</p>

<pre><code>sudo apt-get install php5-fpm
</code></pre>

<h3>Install apc</h3>

<p><a href="http://www.php.net/manual/en/intro.apc.php">apc</a> is a PHP object cache. It&#8217;s required by W3 for doing the PHP caching.</p>

<pre><code>sudo apt-get install php5-apc
</code></pre>

<h3>Install mod_expires apache module</h3>

<p>The caching plugin works better with mod_expires installed.</p>

<pre><code>sudo a2enmod expires
</code></pre>

<p>You should get a response like</p>

<pre><code>Enabling module expires.
Run '/etc/init.d/apache2 restart' to activate new configuration!
</code></pre>

<h3>Install mod_headers apache module</h3>

<p>.. and with mod_headers installed</p>

<pre><code>sudo a2enmod headers    
</code></pre>

<p>You should get a response like</p>

<pre><code>Enabling module headers.
Run '/etc/init.d/apache2 restart' to activate new configuration!
</code></pre>

<h2>Do a graceful apache restart</h2>

<pre><code>sudo /etc/init.d/apache2 reload
</code></pre>

<p>Once you have restarted, have a look in your apache logs for any thing unusual.</p>

<h2>Now install W3 Total Cache</h2>

<p>Once installed and activated, the first thing to do is run the compatibility check to ensure that everything is ok.</p>

<p>Then you can start enabling features.. and hopefully see some response time improvements!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[DotForge Accelerator - Sheffield - updated 29th Jan]]></title>
    <link href="http://jamesjefferies.com/2013/01/18/dotforge-accelerator-sheffield/"/>
    <updated>2013-01-18T11:14:00+00:00</updated>
    <id>http://jamesjefferies.com/2013/01/18/dotforge-accelerator-sheffield</id>
    <content type="html"><![CDATA[<p>The other day I went to the first event for the new <a href="http://dotforgeaccelerator.com">Dotforge Accelerator</a> programme in Sheffield.</p>

<p>Having been following the progress of a couple of Sheffield companies, <a href="http://www.orderharmony.com/">OrderHarmony</a> and <a href="http://matchchat.co.uk/">Matchchat</a> during the <a href="http://ignite100.com/">Ignite100</a> accelerator programme in Newcastle Upon Tyne, I was interested to find out what is planned for Sheffield.</p>

<p>This is my write up.</p>

<h2>What is it?</h2>

<p>Dotforge is going to follow the same kind of pattern as Ignite100, TechStars, Y-Combinator and other programmes. A 13 week intense programme consisting of a number of cycles, pitches, mentoring, training, building, creating, destroying, tweaking and learning. The programme will be for 8 startup teams.</p>

<p>By the end of the programme, teams will be able to pitch to investors in Sheffield &amp; in London, to hopefully provide funding for their startup business to make the next big steps.</p>

<p>It is being run by <a href="http://twitter.com/jagusti">Jag Goraya</a> and <a href="http://twitter.com/germinates">Emma Cheshire</a>.</p>

<p>The list of mentors can be found on the <a href="http://dotforgeaccelerator.com/mentors/">Dotforge site</a></p>

<h2>Areas of Interest</h2>

<p>The programme will, in theory, work well for any businesses that are high-growth, scalable and (web) technology-enabled, regardless of market or industry sector. The programme’s focus is to identify startups that can get to the next stage of growth, and get them to it.</p>

<p>Having said that, the programme has the following as its main areas of interest.</p>

<ul>
<li>Take advantage of the &#8216;Cloud&#8217;</li>
<li>Work as Business to Business (B2B) products</li>
<li>use Big Data or Social data</li>
<li>have the capability for high growth</li>
<li>have global ambition</li>
<li>web technologies</li>
<li>consider Sheffield strengths, for example

<ul>
<li>e-learning</li>
<li>e-health</li>
<li>advanced manufacturing</li>
</ul>
</li>
</ul>


<h2>Teams</h2>

<p>Applications are encouraged from teams with more than 1 founder. There is no upper limit and typically there will be 3 to 4 people. Preferably people who already know each other or have worked together, but there are no real hard and fast rules.</p>

<p>A lone founder could apply and get on the programme, but usually people find that they are so busy during the 13 weeks, that the load needs to be shared out amongst other team members, you can&#8217;t be in two places at the same time.</p>

<p>A good mix of skills is required, having development skills as well as ideas is pretty much considered essential.</p>

<h2>The beans</h2>

<h3>Initial money</h3>

<p>Each company on the programme gets £5k investment, per person (up to three team members) for a 6% stake in the company. The 6% is made up of the various investors who are funding the programme.</p>

<p>Around half of all companies on these programmes end up getting decent investment offers at the end, i.e. hundreds of thousands of pounds.</p>

<h3>Location</h3>

<p>There will be an office in Sheffield city centre, where the teams will be located for the 13 weeks. The focus for the teams will be on the programme, trying to juggle other interests at the same time will always be possible, but you get out of the programme what you put in.</p>

<h3>Perks</h3>

<p>There are often benefits associated with being on an accelerator programme, credit for Amazon web services or Rackspace for example. This would enable growth without having to worry quite as much about infrastructure costs.</p>

<h2>When?</h2>

<p>Applications are now <a href="http://dotforgeaccelerator.com/programme/">open</a>, the programme will commence after Easter 2013. The application platform is hosted by <a href="http://www.f6s.com/dotforge-2013-spring">f6s</a> and will be their standard process.</p>

<h2>Process</h2>

<p>The selection process is in three stages, the online application review, telephone/skype interview, face-to-face interview, each of which reduces the number of applicants down to the final eight.</p>

<h2>Well?</h2>

<p>I&#8217;m certainly going to consider this very carefully to see whether I can form or join a team for this accelerator and whether it is the right time for me.</p>

<p>If you&#8217;re working on a startup, or have a cracking idea, then you really should have a look and consider applying!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Sheffield's High Speed Railway Station HS2]]></title>
    <link href="http://jamesjefferies.com/2013/01/08/sheffields-high-speed-railway-station-hs2/"/>
    <updated>2013-01-08T15:46:00+00:00</updated>
    <id>http://jamesjefferies.com/2013/01/08/sheffields-high-speed-railway-station-hs2</id>
    <content type="html"><![CDATA[<p>There has been a fair bit of discussion in the press about where Sheffield&#8217;s High Speed station is to be situated for HS2. The leaks indicate that Meadowhall is the likely location. Most recently in the <a href="http://www.telegraph.co.uk/news/uknews/road-and-rail-transport/9783576/HS2-route-Manchester-and-Liverpool-win-while-Sheffield-loses-out.html">Daily Telegraph</a></p>

<p>My local MP, Paul Blomfield also <a href="http://www.paulblomfield.co.uk/news/news-story/article/the-rail-minister-is-getting-it-badly-wrong-by-proposing-that-sheffields-hs2-station-should-be-at-m.html">supports a city centre station for HS2</a> instead of Meadowhall.</p>

<p>I&#8217;d like to make a few points about why Meadowhall makes some sense for a new station, Tinsley Yard makes even more sense and then I&#8217;ll  provide some other examples of getting the station in to the city centre instead.</p>

<h2>High Speed 2 (HS2)</h2>

<p><a href="http://www.hs2.org.uk/">HS2</a> is a new High Speed line planned to be built over the next 20 odd years from London Euston, initially to Birmingham, before splitting in to two arms, one to Manchester, the other to Leeds (via the East Midlands and South Yorkshire). Speed is projected to be up to 225mph with an estimated journey time from the South Yorkshire station of 1 hour 10 minutes. The route to Leeds &amp; South Yorkshire is due to be open by <a href="https://www.gov.uk/government/news/britain-to-have-new-national-high-speed-rail-network">2033</a></p>

<h2>Current and near future plans for Sheffield</h2>

<p>The fastest trains at the moment take 2 hours and 7 minutes to travel from Sheffield Midland station to London St. Pancras via Chesterfield, Derby and Leicester.</p>

<p>Following electrification of the Midland Mainline, projected times to London would be 1 hour 45 minutes within the next 5 years.</p>

<p><strong>UPDATE</strong> - actually, electrification from Derby to Sheffield will take place by December 2021 in <a href="http://www.networkrail.co.uk/browse%20documents/strategicbusinessplan/cp5/supporting%20documents/our%20activity%20and%20expenditure%20plans/route%20plans/east%20midlands%20route%20plan.pdf">Control Period 6</a></p>

<h2>Common Design Patterns for High Speed stations</h2>

<p>Although currently the UK has very few High Speed stations (St. Pancras, Stratford, Ebbsfleet, Ashford), Europe, especially France has a lot more experience of building and siting them. The common design patterns which are generally used are:</p>

<ul>
<li>Intermediate stations are on passing loops, allowing non-stop trains to pass straight through at maximum speed. This means the station is generally long and straight, requiring enough room for the loops too. Platforms for HS2 are planned to be about 415m.</li>
</ul>


<p>Here is a train passing at high speed!</p>

<iframe class="youtube-player" type="text/html" width="640" height="385" src="http://www.youtube.com/embed/9MMNlsatk18" frameborder="0">
</iframe>


<ul>
<li><p>Stations are often out of town to allow easy access for a wide variety of people, including huge car parks for drivers. The equivalent for the UK at the moment is the &#8216;Parkway&#8217; station, like at East Midlands.</p></li>
<li><p>Speed is rarely compromised, and when it is, it is usually because of important cross-connections and infrastructure, i.e. Lille Europe, where a massive new station was built with major disruption. So you rarely get through stations built on bends or in city centres!</p></li>
<li><p>A new UK example will be the station provided outside Birmingham, the plans are available for <a href="http://assets.dft.gov.uk/publications/hs2-maps-20120110/hs2arp00drrw05303issue2.pdf">download</a></p></li>
<li><p>HS2 stations are going to be designed for speedy embarkation, the trains are nearly half a kilometre long, so passengers will be directed to the right carriage <em>before</em> the train arrives. There will be no last minute jumping on with these trains.</p></li>
<li><p>Stations shouldn&#8217;t be too close together. A purist route would have a terminus at Leeds, a station outside Birmingham and a terminus at London Euston, that would be the most efficent and that&#8217;s your lot!</p></li>
</ul>


<h2>Why Meadowhall/Tinsley Yard makes sense</h2>

<ul>
<li><p>Space - there is far more brownfield space around Meadowhall for building new infrastructure. There is also room to move businesses around if required.</p></li>
<li><p>Existing infrastructure - the station would be on the tram network, serving Sheffield and by then Rotheram too. The M1 travels right by on the Tinsley Viaduct.</p></li>
<li><p>It meets the common design patterns outlined above.</p></li>
<li><p>Area - it would serve South Yorkshire, not just Sheffield City Centre. Most of the official documentation I have seen references South Yorkshire, not <em>Sheffield</em></p></li>
<li><p>If built in the old Tinsley Yard, there is a lots of good space for it to go. Access could be provided at either end.</p></li>
</ul>


<h2>Why it doesn&#8217;t make sense</h2>

<ul>
<li><p>Meadowhall is already busy as a transport hub, especially the M1.</p></li>
<li><p>It will add a good 15-20 minutes on to the journey time from the City Centre, giving a similar time to the new &#8216;classic&#8217; electrified service  anyway.</p></li>
<li><p>City Centre businesses need a city centre station for growth (so they say, I&#8217;ve not read up on any research in to this).</p></li>
</ul>


<h2>Route considerations</h2>

<p>At the time of writing we do not know which route HS2 is set to take through South Yorkshire. Consultation is due in 2014.</p>

<p>High Speed routes though, again, have common design patterns</p>

<ul>
<li><p>no sharp curves which would reduce speed unless absolutely necessary. These trains are not built to tilt.</p></li>
<li><p>gradients aren&#8217;t usually too much of a problem, compared to normal trains and routes. High Speed lines often have &#8216;motorway&#8217; style gradients.</p></li>
<li><p>High density population areas to be avoided otherwise extensive soundproofing &amp; demolishment is required.</p></li>
</ul>


<p>Consequently, I&#8217;d be surprised if HS2 was to be routed parallel to the existing mainline route via Dore and the Sheaf valley. You probably could squeeze the infrastructure in and you might have an ideal alignment for the existing Midland station.</p>

<p>However, what is more likely would be a route around the East Side of Sheffield where the trains from the South originally came from, before the route along the Sheaf valley was built towards Dore. This is the route you might have experienced when engineering works force a diversion. You leave Sheffield to the North, turn right through Darnall, Woodhouse, before bearing South back round to Chesterfield past Staveley and Barrow Hill.</p>

<p>So, I could see a route which heads round the East of Sheffield, avoiding the centre completely, before heading North to Meadowhall to a South Yorkshire Parkway station.</p>

<h2>City Centre station (Midland)</h2>

<p>Let&#8217;s assume that the route <em>can</em> be built along the Sheaf valley, parallel to the existing mainline. What do you do when you get to Sheffield Midland station?</p>

<ul>
<li><p>You can&#8217;t build underneath it, because that is where the <a href="http://www.flickr.com/photos/richboxfrenzy/sets/72157631911926282/">Megatron</a> is. Midland station is built on top of the river.</p></li>
<li><p>You could build on top of it but building a bigger station on top of a smaller station would make a bit of a mess.</p></li>
<li><p>You could completely knock down Midland station and rebuild it.</p></li>
<li><p>You could incorporate it in to the existing station, but some platforms would require demolishing and there would be extreme capacity issues for existing services.</p></li>
</ul>


<p>It would end up taking about <a href="https://maps.google.com/maps/ms?ie=UTF&amp;msa=0&amp;msid=203170883012533158238.0004d2c9c0c74fc9d440f">this much</a> space.</p>

<h2>City Centre station (Victoria)</h2>

<p>If the route of HS2 heads around the East of Sheffield, then why not use the old Victoria station? The alignment is correct for incoming trains and you could have a dedicated station there with, say a tramlink to Midland?</p>

<p>Well, I think the problem there is that trains heading for Leeds would then end up pointing towards Hillsborough, Oughtibridge and Stocksbridge. You would need some serious infrastructure (i.e. a big tunnel) to get the route back to a more Northerly direction to continue the journey to Leeds. Not impossible, but adds a lot of extra cost.</p>

<h2>Hybrid using Victora &amp; Midland Stations</h2>

<p>One off the wall idea, might be to re-open Victoria as a mainline station, re-open the line to Penistone, then divert some of the trains currently using Midland to a new Victoria Station.</p>

<p>Services to Lincoln, Leeds (Northern), Barnsley, Huddersfield could all use Victoria with Cross Country and East Midlands then using a reduced size Midland, creating extra space to build High Speed platforms there.</p>

<p>Of course, re-opening Woodhead would be the icing on the cake!</p>

<h2>Tinsley Yard</h2>

<p>This seems to be the best location for alignment and existing space. The long lost Marshalling Yard is in a great location for the Motorway, not that far from Meadowhall and would probably be good for noise and local housing. I think there is a good chance that it will be used as the final location.</p>

<p>Someone has done a great route example <a href="https://maps.google.com/maps/ms?msid=210155476846284269415.0004a7028d801cf5385bd&amp;msa=0&amp;ll=53.401243,-1.384234&amp;spn=0.013868,0.027423">here</a> - looks very plausible to me.</p>

<p>NB. I found that route on a Sheffield Forum <a href="http://www.sheffieldforum.co.uk/showpost.php?p=7828103&amp;postcount=72">post</a></p>

<h2>Meadowhall itself</h2>

<p>The more I look at maps and consider routes, the more I think that a route passing across the valley parallel to the M1 is likely, rather than one parallel to the current mainline at Meadowhall. So, imagine a station under the Tinsley Viaduct, at 90 degree angles to the existing main line. Probably then, with tunnels either side.</p>

<h2>My vote</h2>

<p>If we end up with a station in South Yorkshire, then my vote goes to Tinsley Yard. If you want to get quickly to London in 2032 from Sheffield City Centre. Take the Electrified Midland Mainline to St. Pancras in 1 hour 45 minutes. It&#8217;ll probably be cheaper than HS2, be more scenic and will not take much longer.</p>

<p>If you don&#8217;t live in the City Centre and you drive, park up at the HS2 station next to the M1, get there early and whizz down HS2 to Euston in 1 hour 10-15 minutes. Either way will get you in to the centre of London fairly sharpish. Win win!</p>

<h3>Further reading</h3>

<p><a href="http://www.syltp.org.uk/HS2.aspx">South Yorkshire Local Transport Plan</a></p>

<p><a href="http://beleben.wordpress.com/2013/01/08/hs2-and-south-yorkshire-part-three/">beleben blog</a></p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Olympic legacy server setup - part 3 - installing Ubuntu server]]></title>
    <link href="http://jamesjefferies.com/2013/01/03/olympic-legacy-server-setup-part-3-installing-ubuntu-server/"/>
    <updated>2013-01-03T15:22:00+00:00</updated>
    <id>http://jamesjefferies.com/2013/01/03/olympic-legacy-server-setup-part-3-installing-ubuntu-server</id>
    <content type="html"><![CDATA[<p>I was thinking about a Debian install on my server, but in the end, I thought I&#8217;d go with Ubuntu server.</p>

<h2>Ubuntu server install</h2>

<p>I&#8217;m not going in to all the details about installing Ubuntu server (debian would be similar), but I thought it would be worth noting the steps I made to get the RAID device partitioned ready for installing xen or KVM.</p>

<h3>RAID configuration</h3>

<p>When booting the server, you are able to access the MegaRAID configuration. As my server came with 2 unregistered drives, it was simple to create a new RAID1 configuration with those two 146Gb drives, creating a virtual 146Gb RAID1 drive.</p>

<p>I ummed and ahhed a bit about whether to use the RAID controller, but it looks like it might be ok kit.</p>

<h3>Install</h3>

<p>I usually use inetbootin to create a USB memory stick with the OS I want to install. In this case, I downloaded the 64bit Ubuntu image and used it to create the bootable USB stick.</p>

<p>Expert install followed.</p>

<h3>Grub2 gotcha 1</h3>

<p>Oh dear, I had a number of failed installs until I asked an expert (@wilbatron) for some help. The grub2 installation in to the MBR kept failing because… The grub2 installation doesn&#8217;t fit in the MBR does it? I wish it told me that rather than just failing.</p>

<p>Ah well, got it sorted in the end, here is how I set up my partitions.</p>

<h3>Partition set up</h3>

<p>During the install, do manual partitioning. We&#8217;ll set up 2 physical partitions, one of which will be a Logical Volume Manager (LVM) Physical Volume (PV)</p>

<h4>Physical</h4>

<p>At this point, the menu will show your USB stick (still mounted) and your 146Gb virtual RAID drive. The Hardware RAID is hiding the fact that you&#8217;ve actually got two hard drives there and the installer just sees one.</p>

<ul>
<li>Select the Raid array - in my case, it was labelled <em>146.0Gb LSI MegaRAID 8708EM2</em></li>
<li>Once selected, create a new empty partition (and table) for this device</li>
<li>Choose GPT as the partition table type (not MSDOS!)</li>
<li>Now select FREE SPACE 146.0GB and create new partition</li>
<li>Set new partition size to be just 10Mb, this is where the installer will put the rest of grub2</li>
<li>Set to be at the beginning of the drive</li>
<li>Set type to be used as <em>Reserved BIOS Boot area</em></li>
<li>Create next partition to use as Physical Volume (PV) for LVM</li>
</ul>


<h4>Logical</h4>

<p>Now we configure the LVM</p>

<ul>
<li>Configure LVM (option at the top of the partitioning page)</li>
<li>When prompted, write partition changes to disk</li>
<li>Create volume group, give it a name and select the PV.</li>
<li>Create logical volume, call it ROOT or whatever, mine was 15Gb - this is where the OS will be installed. You could create all the separate partitions if you like of course.</li>
<li>Create swap volume, I chose 6Gb, although the server has 12Gb of server memory, I didn&#8217;t want to give it swap of that size. We&#8217;ll see how we go.</li>
<li>Finish up and head back to partition page</li>
<li>Set root partition to be root (ext4), doesn&#8217;t need to be bootable partition.</li>
<li>Set swap to be swap</li>
<li>Down to the bottom of the page and write changes to disk</li>
<li>Back to main menu and install!</li>
</ul>


<h4>Installing grub in MBR - gotcha 2</h4>

<p>Just another slight gotcha, if your USB stick is mounted as /dev/sda &amp; your RAID array as /dev/sdb you might have to tell it not to install to the MBA, but instead to where you want it installed, i.e. /dev/sdb</p>

<p>Sorted!</p>

<h3>BMC (Baseboard Managment Controller)set up</h3>

<p>To get the BMC up and running properly, I needed to do a few things, first, is to ensure that it is set up correctly in the BIOS.</p>

<p>Enter the BIOS and in the server management section, ensure it is set to use the correct LAN port, either <em>Dedicated</em> if you are using the single port on the back of the server, <em>On Board</em> for one of the 4 built in LAN ports. I used the latter with DHCP. Make a note of the IP address which is given to the BMC and fire it up in your web browser. All being well, you should see the log in page:</p>

<p><img src="http://jamesjefferies.com/images/smartconsole.jpg" alt="console" /></p>

<h4>Default authentication log in</h4>

<p>Apparently the default is either admin/admin or ADMIN/ADMIN so give that a go first of all to see if you get in. If not, we will have to do some jiggery pokery on the server.</p>

<h3>Resetting BMC credentials from the server</h3>

<p>The tool, ipmitool is required to access the BMC from the server. I just did apt-get install ipmitool to get it up and running. However, before I got it working properly, I had to load some kernel modules</p>

<h4>Kernel modules for ipmitool</h4>

<p>First, I&#8217;d see whether it is working straight after install:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool sensor</span></code></pre></td></tr></table></div></figure>


<p>if you get</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>Could not open device at / dev/ipmi0 or / or dev/ipmi/0 / dev/ipmidev/0: No such file or directory
</span><span class='line'>Get Device ID command failed
</span><span class='line'>Unable to open SDR for reading</span></code></pre></td></tr></table></div></figure>


<p>Then try manually loading the kernel modules with</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
<span class='line-number'>5</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>modprobe ipmi_devintf
</span><span class='line'>modprobe ipmi_msghandler
</span><span class='line'>modprobe ipmi_poweroff
</span><span class='line'>modprobe ipmi_si
</span><span class='line'>modprobe ipmi_watchdog</span></code></pre></td></tr></table></div></figure>


<p>You can add these to <code>/etc/modules</code> for permanent loading.</p>

<p>Now try again, you should get some results</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool sensor</span></code></pre></td></tr></table></div></figure>


<h4>Get correct channel</h4>

<p>By running</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool -I open channel info 1</span></code></pre></td></tr></table></div></figure>


<p>(where the latter 1 is the channel, so if 1 doesn&#8217;t work, cycle through)</p>

<p>I got</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
<span class='line-number'>5</span>
<span class='line-number'>6</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>Channel 0x1 info:
</span><span class='line'>  Channel Medium Type   : 802.3 LAN
</span><span class='line'>  Channel Protocol Type : IPMB-1.0
</span><span class='line'>  Session Support       : multi-session
</span><span class='line'>  Active Session Count  : 0
</span><span class='line'>  Protocol Vendor ID    : 7154</span></code></pre></td></tr></table></div></figure>


<p>which showed that Channel 1 was the one I needed.</p>

<h4>List users</h4>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool user list 1</span></code></pre></td></tr></table></div></figure>


<p>should give you a list of users, then you can reset the passwords. I ended up setting all I could to <em>password</em> then, once I was in the web admin, I deleted the users I didn&#8217;t need and set all the passwords to be more secure ones!</p>

<h4>Set password</h4>

<p>Hmm, not sure I got this right, but I tried</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool -I open lan set 1 password admin password                                                                                                                         
</span><span class='line'>Password set for user 1</span></code></pre></td></tr></table></div></figure>


<p>(the 1 here is the channel)</p>

<p>and</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ipmitool user set password 2 password</span></code></pre></td></tr></table></div></figure>


<p>for the various users until I could log in to the console!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Olympic legacy server setup - part 2 - spares]]></title>
    <link href="http://jamesjefferies.com/2013/01/03/olympic-legacy-server-setup-part-2-spares/"/>
    <updated>2013-01-03T14:40:00+00:00</updated>
    <id>http://jamesjefferies.com/2013/01/03/olympic-legacy-server-setup-part-2-spares</id>
    <content type="html"><![CDATA[<p>I&#8217;ve been having a bit of a look around to try and find out where you can buy caddies and Seagate drives for the server.</p>

<h2>Hard Drive Caddies</h2>

<p>My server came with two drives in caddies, but with no spares.</p>

<p>According to the <a href="http://www.supermicro.com/products/system/2U/2026/SYS-2026T-URF4_.cfm">SuperMicro</a> spec for the server, the caddy model number is MCP-220-00047-0B &#8220;Black gen 3 hot swap 2.5&#8221; HDD tray&#8221;</p>

<p>There seem to be a few UK suppliers around, although I wonder what has happened to all the drives removed from the servers before sale?</p>

<p>Anyway, here is one which I think I will try, from <a href="http://www.lambda-tek.com/MCP-220-00047-0B-Supermicro-BLACK-HOT-SWAP-2-5-quot-SAS-SATA-HDD-TRAY~cs/B577304&amp;viewOverview=y#productTop">Lambda-tek</a>, price currently £9.20 + VAT.</p>

<h2>Hard Drives</h2>

<p>Then for the Seagate drives, mine has a couple of Seagate Savvio 10K.3 10k rpm, SAS drives, there are a few on ebay, this one seems to be the cheapest for used drives on <a href="http://r.ebay.com/zq4oCB">ebay</a>. Of course, these drives may have gazillion hours of use on them, but hey ho.</p>

<p><img src="http://jamesjefferies.com/images/seagate.jpeg" alt="seagatedrive" /></p>

<h2>What next</h2>

<p>Well, I&#8217;ve ordered a couple of caddies and drives, will see how they turn out. Plus, I&#8217;ve got an Operating System installed, see next post!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Olympic legacy server setup - part 1b]]></title>
    <link href="http://jamesjefferies.com/2012/12/27/olympic-legacy-server-setup-part-1b/"/>
    <updated>2012-12-27T18:06:00+00:00</updated>
    <id>http://jamesjefferies.com/2012/12/27/olympic-legacy-server-setup-part-1b</id>
    <content type="html"><![CDATA[<h2>More details on the Olympic servers</h2>

<p>Thanks to a lot of investigation by <a href="http://twitter.com/interpipes">@interpipes</a> we now have some more details about these servers.</p>

<p>It looks like Acer/Gateway built the servers on top of one of these SuperMicro <a href="http://www.supermicro.com/products/system/2U/2026/SYS-2026T-URF4_.cfm">Super Servers 2026T-URF4+</a></p>

<p>consisting of one of these <a href="http://www.supermicro.com/products/motherboard/qpi/5500/x8dtu_.cfm">motherboards</a></p>

<p>with one of these <a href="http://www.supermicro.com/products/chassis/2U/213/SC213A-R740W.cfm">chassis</a></p>

<h2>Where was the server used</h2>

<p>Apparently, one of the three letter codes on the server can tell you where the server was located. Mine, seems to have SDC on it, but I haven&#8217;t been able to find out if that is a location or not.</p>

<p>This <a href="http://webcache.googleusercontent.com/search?q=cache:aiCDaaSXo1oJ:www.dalkeithtransport.co.uk/wp-content/uploads/Olympic-Venues-for-UP-Deliveries.xls+&amp;cd=1&amp;hl=en&amp;ct=clnk&amp;gl=uk">spreadsheet</a> is the UPS delivery cheatsheet, no SDC on there.</p>

<p>Mine also has PDC on one of it&#8217;s stickers, which might be the Primary Data Centre. Try searching for it in your favourite search engine with Locog.</p>

<h2>State of servers on delivery</h2>

<p>The RAID card was loose in the server case when I unpacked mine, which I&#8217;m glad I spotted before firing up for the first time. Once re-attached, all seems fine.</p>

<p>I also thought mine was broken, but actually, it was what you might call a user error! What I hadn&#8217;t realised was that if you only plug in one of the power supplies, then the server lets you know that one of the power supplies might be broken by sounding a continous BEEEEEEEP! So don&#8217;t make that error, plug both power leads in!</p>

<p>Once again, @interpipes to the rescue! &#8220;beeping is just to tell you one of the supplies is &#8220;failed&#8221; - pull it out (an inch)/plug it in, and it&#8217;ll stop beeping.&#8221;</p>

<p>@interpipes also had some interesting configurations too on his servers
&#8220;Half of mine had bios passwords set, half didn&#8217;t, and 3 had a configuration on the BMC controllers.&#8221;</p>

<h3>Advice on getting in to the BMC via @interpipes</h3>

<p>The <a href="http://en.wikipedia.org/wiki/Baseboard_management_controller">BMC</a> is the Baseboard Management Controller, an extra microcontroller which servers to manage server kind of things at a lower level.</p>

<blockquote><p>&#8220;Getting back into the BMC (which among other things gives you full IP KVM access), </p><p>first, set up the network configuration for it in the BIOS as if it has been configured it&#8217;ll have a static IP set the network config is under Server Management, and then the first option in that menu - I forget what it is called - I recommend using the dedicated LAN port setting, which is the solitary lan port above the two usb sockets. </p><p>Now you can get to it over your lan (pointing a web browser at it being the easiest way) but you&#8217;ll need to use ipmitool to change the root user password back to something you know (then you can log in and reset the BMC to factory defaults). Using the BMC it&#8217;s possible to mount an ISO from a windows file share (I think you can also mount one from a local disk when connected to the remote control interface) to install the system from entirely remotely.&#8221;</p></blockquote>


<h2>Installing an OS</h2>

<p>Now that I could start mine up without it beeping the place down, I entered the RAID configuration and set up a simple 146Gb RAID1 set up, using the 2x146Gb 10k RPM 2.5&#8221; disks it came with.</p>

<p>I also used <a href="http://unetbootin.sourceforge.net">Unetbootin</a> to get a server ISO set up on my old USB memory stick. Of course, it&#8217;s up to you to choose your favourite Linux distro, Unetbootin will allow you to download a customised ISO too which it will then set up on the USB, so you are not limited to the ones it comes with by default.</p>

<p>The server quite happily booted off the USB stick, so no optical drive is required (my server didn&#8217;t have one).</p>

<h2>What&#8217;s next?</h2>

<p>Getting the OS installed and, I think Xen set up.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Olympic legacy server setup - part 1]]></title>
    <link href="http://jamesjefferies.com/2012/12/20/olympic-legacy-server-setup-part-1/"/>
    <updated>2012-12-20T15:14:00+00:00</updated>
    <id>http://jamesjefferies.com/2012/12/20/olympic-legacy-server-setup-part-1</id>
    <content type="html"><![CDATA[<p>For the Olympics, a whole lot of techie kit was required to do lots of clever things. As a lot of the kit wasn&#8217;t required after the games were over, they have been made available for purchase at super cheap prices! Kit was originally for Educational Establishments, but once that batch were sold, I could buy one. So I did!</p>

<h2>Server Model - Gateway GR380 F1</h2>

<p>Well, I hadn&#8217;t realised that the Gateway brand had been brought back to life, but it has by Acer! The Gateway GR380 F1 - you can have a look at the <a href="http://tinyurl.com/d2xvncz">spec</a> if you like.  It is a 2U (i.e. bigger) version of one of <a href="http://www.pcpro.co.uk/reviews/servers/359875/gateway-gr360-f1">these</a> coming with 2x146Gb drives, lots of memory and a big cardboard box.</p>

<p><img src="http://jamesjefferies.com/images/big-server.jpeg" alt="bigserver" /></p>

<h2>What are you going to do with it?</h2>

<p>Good question, I&#8217;m thinking of setting up a virtualisation kind of thing, so my big server can pretend to be lots of smaller servers, useful for building client sites, experimental things, build servers etc. We&#8217;ll see</p>

<h2>Where do you start?</h2>

<p>Another good question. It comes with the Smart Manager application, which you can install on it to do lots of clever things, however, at the moment, I&#8217;ve not even fired it up to see if it has the remnants of an Operating System on it. I&#8217;m suspecting not!</p>

<h2>I&#8217;ve got one too, where do I start?</h2>

<p>Well, I&#8217;ve managed to find the user manual, you can download it yourself from the <a href="http://global-download.gateway.com/GDFiles/Document/User%20Manual/User%20Manual_Gateway_1.0_A_A.zip?acerid=634213526746016513&amp;Step1=Server&amp;Step2=Server&amp;Step3=GR380%20F1&amp;OS=303&amp;LC=en&amp;BC=Gateway&amp;SC=EMEA_27G">Acer site</a>.</p>

<p>If nothing else I&#8217;d like to have a look inside to see what sort of state it&#8217;s in. Then I need to work out how to attach the funky rack mounts to it, then I need to work out how to install Debian or Ubuntu on it (plus Xen probably for virtualisation).</p>

<p>Oh - it&#8217;ll need a name too!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Steam powered Arduino part 1]]></title>
    <link href="http://jamesjefferies.com/2012/11/11/steam-powered-arduino-part-1/"/>
    <updated>2012-11-11T16:22:00+00:00</updated>
    <id>http://jamesjefferies.com/2012/11/11/steam-powered-arduino-part-1</id>
    <content type="html"><![CDATA[<p>Following on from getting <a href="http://jamesjefferies.com/2012/08/30/the-story-of-cathy-and-heathcliff">Cathy &amp; Heathcliff</a>, our internet enabled thermal printers up and running as part of the <a href="http://happenstanceproject.com">Happenstance Project</a>, I wanted to see if I could get them running on steam power.</p>

<h2>Steam power</h2>

<p>Of course most electricity we get from the national grid has been generated by a steam turbine in a power station. I though, was thinking more of this kind of steam engine!</p>

<p><img src="http://jamesjefferies.com/images/mamod-in-the-kitchen.jpg" alt="mamod in the kitchen" /></p>

<p>This is stationary steam engine, hand built in the West Midlands by a company called <a href="http://mamod.co.uk">Mamod</a> founded in 1937.</p>

<h2>Mamod 1335D SP5</h2>

<p>This specific model comes with a dynamo, dual cylinders and a generator, a 1335D. I wanted to see how much power it would output with the standard dynamo. The bulb was a 1.5V bulb, so I wasn&#8217;t expecting it to power an Arduino out of the box. I realise as well that the output would require stabilising and regulating. Cathy &amp; Heathcliff&#8217;s Arduino&#8217;s require 7-16V.</p>

<p><img src="http://jamesjefferies.com/images/testing-the-voltage.jpg" alt="testing the voltage" /></p>

<p>And I was right! I lit the boiler using three of the solid fuel pellets and waited for pressure to build in the boiler.</p>

<h2>Pressure builds</h2>

<p>Once the pressure did start to build, it soon became obvious that there were a few leaks which meant that some pressure was going to be lost. I kept on attempting to start the engine, but it was a while before it was ready, around 10 pounds per square inch of pressure.</p>

<p>After starting, the generator produced a rather paltry, 0.7V, hardly enough to light the light!</p>

<h2>Best results</h2>

<p>I managed then to stop the engine and hold it, to let the pressure build. Once it reached about 15 psi, I started the engine again and tested the voltage. This time it made it up to 1.5V</p>

<p>After that run, the pressure dropped and the fuel began to run out, so that was that.</p>

<h2>Next Time</h2>

<p>The plan for next time will be to not even start the engine until the pressure reaches 15 psi. I might also try and extra fuel pellet to see if that helps maintain the pressure.</p>

<p>I&#8217;ve ordered some <a href="http://sugru.co.uk">sugru</a> to try and fix the leak too! Will be interesting to see if it works.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[A great idea for Amazon deliveries]]></title>
    <link href="http://jamesjefferies.com/2012/10/28/a-great-idea-for-amazon-deliveries/"/>
    <updated>2012-10-28T15:14:00+00:00</updated>
    <id>http://jamesjefferies.com/2012/10/28/a-great-idea-for-amazon-deliveries</id>
    <content type="html"><![CDATA[<h2>Amazon Pickup Lockers</h2>

<p>This year, Amazon launched it&#8217;s Pick up lockers and Collection locations. A great idea for a mail order company, providing delivery alternatives for customers so they don&#8217;t have to put up with the, <a href="http://www.guardian.co.uk/money/2012/aug/24/amazon-complaints-yodel-delivery-service">generally awful</a>, parcel delivery services. Having had parcels left outside houses and in rubbish bins, it&#8217;s good to know you can collect, at your convenience from a location you have chosen.</p>

<h2>The Oxford - Cambridge Railway</h2>

<p>Then, the other weekend, I was on a railway line I&#8217;d never been on before, the Bletchley to Bedford line, which joins the West Coast Main Line, to the Midland Main Line, serving lots of small stations on route. A lot of work has been done on the infrastructure recently, as <a href="http://www.railfuture.org.uk/ox-cam/">services get extended</a> from Bletchley towards Oxford, and hopefully one day, back towards Cambridge. The line will also be electrified as part of the <a href="http://jamesjefferies.com/2012/07/16/the-electrified-spine-and-the-rest-of-the-news/">&#8216;electric spine&#8217;</a>.</p>

<h2>How about&#8230;</h2>

<p>About half way along this route, is a massive Amazon warehouse. Although it is next to the railway line, it isn&#8217;t rail served, there are no sidings set up for transferring goods. It got me thinking though, if local and major railway stations had Amazon pick up lockers, then they could be rail served in a really efficient way. Of course there would be infrastructure costs, getting Amazon warehouses linked up, but once that was done, a really efficient way of delivering goods.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Scenic Railways - our railway hack]]></title>
    <link href="http://jamesjefferies.com/2012/10/16/scenic-railways-our-railway-hack/"/>
    <updated>2012-10-16T16:22:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/10/16/scenic-railways-our-railway-hack</id>
    <content type="html"><![CDATA[<p><img src="http://jamesjefferies.com/images/scenic-railways.png" alt="screen shot of application" /></p>

<p>Thanks to the <a href="http://jamesjefferies.com/2012/10/16/railway-hackday-off-the-rails/">Off The Rails</a> hackday, I was able to team up with some fellow software engineers to build something in a day. Our team consisted of <a href="http://twitter.com/frankieroberto">Frankie Roberto</a>, <a href="http://twitter.com/jnicho02">Jez Nicholson</a>, and <a href="http://twitter.com/joooe">Joe Hughes</a>. Frankie had had the idea for building an application which would show sights of interest visible from the train route, indicating rough placements and which side of the train to look on. I knew Frankie from Sheffield, Frankie had worked with Jez before on a previous hackday and Joe was interested in what we were doing, our team was formed!</p>

<p>Interestingly, during the &#8216;speed dating&#8217; section at the beginning of the day, when we chatted to other people about their ideas and what they were thinking of building, one chap, whose name I think was Phil, was thinking about Railway iSpy, which sounds pretty similar to what we actually built (without Big Chief i-Spy though!). Our application also bears an uncanny similiarity to the Peppa Pig episode <a href="http://www.youtube.com/watch?v=Htl6NYcBE-k">The Train Ride!</a></p>

<p>After some discussion on ideas, scope, implementation and the <a href="http://www.railperf.org.uk/">Railway Performance Society</a> we started building a Rails web application and an iOS app. Although I know a lot about railways, I&#8217;m still in the early days of learning Rails and iOS, so the patience of Jez and Frankie whilst building <a href="http://www.scenicrailways.org.uk">Scenic Railways</a> was much appreciated. So whilst the three of us built the web application, Joe got cracking with building an iOS application.</p>

<p><img src="http://jamesjefferies.com/images/scenic-railways-on-iphone.png" alt="ios-1" /></p>

<p>We got an initial version built in the time we had (from about 11am until 7pm) before the presentations were due to start. I worked on a bit of map integration using <a href="http://leaflet.cloudmade.com">leaflet.js</a> whilst Jez and Frankie implemented the main structure.</p>

<p>Although the version we finished was relatively basic, there are some nice features, the best one being that the sides to look out of the window on, reverse depending on direction of travel!  It could easily be extended in the future to add extra routes and extra points of interest (POIs). Joe had some ideas on making the iOS app competitive, by providing incentives for people to take photos of the points from the train!</p>

<p><img src="http://jamesjefferies.com/images/scenic-railways-1.png" alt="screen shot of application" /></p>

<p>So, feel free to have a look at the site <a href="http://www.scenicrailways.org.uk">Scenic Railways</a> and you can even checkout the code from <a href="https://github.com/frankieroberto/scenic-railways">github</a> and laugh at my Rails code (or lack of it!).</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Railway Hackday - Off The Rails]]></title>
    <link href="http://jamesjefferies.com/2012/10/16/railway-hackday-off-the-rails/"/>
    <updated>2012-10-16T15:12:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/10/16/railway-hackday-off-the-rails</id>
    <content type="html"><![CDATA[<p><img src="http://jamesjefferies.com/images/off_the_rails_sponsors1.png" title="Off The Rails" alt="off the rails image" /></p>

<p>After being Technologist in Residence at <a href="http://sitegallery.org">Site Gallery</a> earlier in the year, as part of the <a href="http://happenstanceproject.com">Happenstance Project</a> I started to think of some ideas I wanted to follow for later in 2012.</p>

<h2>Hackday</h2>

<p>One of those ideas was to run a <a href="http://rewiredstate.org/hack-days">Hackday</a>, but not any old Hackday, one centred around Railways. Network Rail had just opened up their realtime <a href="http://datafeeds.networkrail.co.uk">Datafeeds</a> and as far as I knew, no one else had ever run one. Sure, there had been Transport Hackdays, but not Railway specific ones.</p>

<h2>First steps</h2>

<p>So I got to work, I went to experience a 24 hour hackday for myself, where I discovered that sometimes technical issues with the building can really slow people&#8217;s progress and that working overnight doesn&#8217;t really suit me anymore!</p>

<h2>Clickerty Hack</h2>

<p>I also bought a domain and registered a Twitter account. The hackday would be called Clickerty Hack. Then, having a good read of <a href="http://hackdaymanifesto.com/">The Hackday Manifesto</a> made me realise that a lot of work was required to get a decent Hackday up and running.</p>

<h2>Loco 2 and Off The Rails</h2>

<p>It was then that I found out about somebody else&#8217;s plans for a Railway Hackday, <a href="http://loco2.com">loco2</a> was a company I was already watching with interest. A startup based around European Train travel, ticketing and guides.</p>

<p>Their hackday, Off the Rails, was in a more advanced state than Clickerty Hack, so putting it on hold, I volunteered to help.</p>

<h2>ShedCode Sponsorship</h2>

<p>It was a real privilege to be able to help with a small part of the organising and for my company <a href="http://shedcode.co.uk">ShedCode</a> to be one of the sponsors, providing some of the food for the masses!</p>

<h2>Meeting the experts</h2>

<p>During 2012, I&#8217;d found a number of people who had built some incredible pieces of software using Railway data, especially <a href="http://twitter.com/poggs">Peter Hicks</a> and <a href="http://twitter.com/swlines">Tom Cairns</a> who built, amongst other things <a href="http://www.opentraintimes.com/">Open Train Times</a> and <a href="http://traintimes.im">Train Times</a> respectively. So it was brilliant to meet them in person at Off The Rails. There were also many other people knowledgeable about Railways, Open Data and Mapping - I could have spent all day just talking to people!</p>

<h2>The teams</h2>

<p>During the day, the teams built some inspiring things, including an application which builds spotify playlists to match the length of your train journey and many others. Most of them will hopefully be available at some point on github for perusal.</p>

<h2>Thanks</h2>

<p>All in all, a fantastic day, thanks to <a href="http://loco2.com">Loco2</a> and <a href="http://bethnalgreenventures.com">Bethnal Green Ventures</a> and here&#8217;s to the next Railway Hackday, whether it&#8217;s Off The Rails 2, or Clickerty Hack! With cake too of course ;)</p>

<p><img src="http://jamesjefferies.com/images/cake.jpg" title="Cake" alt="off the rails cake" /></p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Tech companies - accelerate your employees]]></title>
    <link href="http://jamesjefferies.com/2012/09/30/tech-companies-accelerate-your-employees/"/>
    <updated>2012-09-30T10:29:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/09/30/tech-companies-accelerate-your-employees</id>
    <content type="html"><![CDATA[<h2>Corporate Social Responsibility</h2>

<p>Corporate Social Responsibility is a huge field in modern times. Businesses vary in their response to the challenge of responsibility, from box ticking a checklist, to doubling up charitable donations, to actively encouraging employees to be involved in Charitable works.</p>

<p>My old company, <a href="http://www.technophobia.com">Technophobia</a> ran a FedEx day, where one of the challenges was for a team to rebuild a website for a charity, which they did. You can <a href="http://www.technophobia.com/blog/read/another-successful-innovation-day-407">read all about</a> it, good for them!</p>

<p>This year though, I&#8217;ve been thinking more about how companies could donate their employees and employee time to make a difference in more unusual ways.</p>

<h2>Technologist in Residence</h2>

<p>Having had the privilege of being a Technologist in Residence at the <a href="http://sitegallery.org">Site Gallery</a> earlier this year, I was able to see first hand what a difference a problem solving techie can make in a relatively small art&#8217;s organisation. With a Research and Development remit, as well as helping with some of the day-to-day tasks, a techie can enjoy the excitment of trying new things, whilst also being able to fix problems without the usual gubbins of working for a large corporation.</p>

<p>Of course, many arts organisations have a resident techie, but usually they have to spin many plates, from producing exhibitions, to maintaining the website, leaving little room for them to catch up with some of the more strategic things which need doing.</p>

<p>There is limited funding to run <a href="http://www.happenstanceproject.com">Happenstance</a> style residencies, and this is where local businesses could make a difference. By adopting a local arts organisation, getting involved, running events and allowing staff to spend a week once a year, then a day a month for the rest of the time, for example, the benefits would be felt by all.</p>

<h2>Who gains?</h2>

<ul>
<li>The arts organisation - they gain some help in fixing some things, as well as being inspired with new and interesting technology. Building relationships with the local tech community</li>
<li>The techies - they can hopefully have time to do experiments, R&amp;D, as well as doing some of the stuff which needs doing. From tidying networks, configuring firewalls, mending printers, redesigning websites, helping with copy… you get the picture!</li>
<li>The company - they help the local community, building relationships with organisations around and about, increasing their reach and network. Their employees are more inspired and excited about new things.</li>
</ul>


<h2>Who pays in these difficult times</h2>

<p>It would be completely understandable for tech companies to claim that they cannot afford to let their billable employees take time out of their billing cycle to do &#8216;good&#8217;. Understandable, but short sighted. It&#8217;s the same argument about sending employees on courses, or to conferences - &#8220;where is the business benefit?&#8221;.</p>

<p>Well, I believe that good, motivated employees are expensive to recruit and maintain, why risk them getting demotivated, bored and unhappy, when by giving them time to innovate and collaborate, they grow! So the risk is far greater, that you lose good people by not giving them the room to grow. Best to pay out of a sense of innovation and employee investment, than to pay the price for losing them, either by them leaving or by turning them in to faceless, billable resources.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Happenstance Newspaper]]></title>
    <link href="http://jamesjefferies.com/2012/09/26/happenstance-newspaper/"/>
    <updated>2012-09-26T20:50:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/09/26/happenstance-newspaper</id>
    <content type="html"><![CDATA[<p>Following on from the amazing <a href="http://happenstanceproject.com">Happenstance Project</a> I worked on earlier in the year, we had the final Showcase at Nesta HQ. For the Showcase, <a href="http://finalbullet.com">Leila</a> produced a Happenstance Newspaper, designed by <a href="http://alexparrott.co.uk">Alex Parrott</a> including articles by yours truly! You can download a <a href="http://jamesjefferies.com/Happenstance_2012_Newspaper.pdf">PDF copy of it</a>.</p>

<p>Enjoy!</p>

<p><img src="http://jamesjefferies.com/happenstance_newspaper.jpg" title="Happenstance Newspaper" alt="Newspaper" /></p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[The Creativity Tap]]></title>
    <link href="http://jamesjefferies.com/2012/09/25/the-creativity-tap/"/>
    <updated>2012-09-25T22:15:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/09/25/the-creativity-tap</id>
    <content type="html"><![CDATA[<h2>The Creativity Tap</h2>

<p>Creativity is just like a tap right? You need to do some thing a bit woo and a bit wah, so you switch on the creativity tap and all these incredible ideas just pour out of your brain. There are two dangers though if you use the creativity tap.</p>

<p>One is that if you leave that tap switched off for a while, then all the creativity gets stored up in your brain, without an outlet until&#8230; Pop! One day it’s gone, those ideas had to go somewhere, so they leaked out, usually somewhere behind your ear, hopefully picked up by someone else. You have writers block, coders deadlock, songwriters&#8230; err, song block. Oh dear</p>

<p>The other danger is that you leave that tap switched on all the time until your brain empties. This too is a bad idea, sure, you’ve made some room in your noddle for new things, but you need a few left in there, to keep the ideas factory ticking over. Oh double dear&#8230;</p>

<h2>Writing Songs</h2>

<p>Knowing people who are songwriters, I get the privilege of seeing songs appear, until they end up being recorded and performed, or they mutate into a different song, or they end up on the shelf, maybe for another day. I’m sure not every songwriter works like this, but those I know have to <em>work</em> at honing a song. They rarely appear fully formed but they often appear when there is space in their lives to switch on the creativity tap for a bit to see what happens.</p>

<p>If they don’t have the chance to get the songs <em>out</em>, into the open, then you can see how those song ideas start to play merry hell. The frustration builds, no time for the songs, too busy having to do other stuff. It’s almost a physical pain, the song <em>has</em> to be brought to life. There is a creative imperative which must be fulfilled.</p>

<h2>Writing Code</h2>

<p>Fortunately, for us software engineers, coders, developers, tapping at our keyboards day by day, telling the computer what to do and when to do it, we don’t need the creativity tap. We get stuck? We ask a search engine. We need to do something from scratch? Well, <em>someone</em> will have done something like it, we just need to find it and copy them. We need another pair of eyes? Well, ask a colleague. It’s all there for us, if we use a clever framework or development environment, it’ll probably autocomplete what we’re thinking too. We are coding machines..</p>

<p>Or are we? Having spent many years writing code, I started to realise that when I was mechanistic in what I was doing, I became frustrated that I wasn’t able to come up with new ideas or ways of doing things. I was constrained by getting this piece of work finished in as short a period of time that was possible. No time to switch on the creativity tap, just <em>get it done</em>.</p>

<p>The big problem is when you work for someone else, you can find yourself under a lot of pressure to crank the sausage machine handle rather than think things through. I worked with an engineer once, who would deliberately stop what he was doing and just <em>think</em>. Rather than thrashing through a jungle of code&#8230; stop first and think. It worked too, giving a bit of time to turn on the creativity tap to see what happened.</p>

<p>I struggled in my day to day job, so I tried to make time in the evenings and the weekends. But you’re so tired then that switching on the creativity tap just floods you with ideas which you don’t have the energy to either make note of or do anything with. Like the frustrated songwriter I would feel the pain of not learning new skills, not playing with languages and technology. Not actually chuffing making anything which made you say, even if it is just to yourself “I made that and I’m pleased with what I’ve done”.</p>

<p>So, what can be done?</p>

<p>Well one answer, in the workplace, where people have a pressurised, day to day job and a boss, is for the bosses to give people time when they have the headspace to play. With their work, with something new, with wooden blocks, with a great idea they’ve had. Not only do they get chance to switch on the creativity tap, but usually it knocks on in to their ‘day to day’ work too. Wins all around.</p>

<p>If you are your own boss, then plan in time where you put down the work you’re doing for clients or your startup and have a play. You never know what amazing ideas might come to life!</p>

<p>Disclaimer - the creativity tap premise is a load of nonsense, if only it was a tap we could switch on and off!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Troubleshooting WordPress connectivity issue - could not connect to host]]></title>
    <link href="http://jamesjefferies.com/2012/09/05/troubleshooting-wordpress-connectivity-issue/"/>
    <updated>2012-09-05T14:57:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/09/05/troubleshooting-wordpress-connectivity-issue</id>
    <content type="html"><![CDATA[<h2>The problem</h2>

<p>After migrating a WordPress site, auto update and plugin install wouldn&#8217;t work. The former errored with <code>could not connect to host</code> which seemed very odd.</p>

<h3>First attempts</h3>

<p>WordPress allows you to download the file manually or autoupdate if configured correctly. Clicking on the manual download link and the file downloaded fine to the local machine. So first, I thought I&#8217;d try and download the file from the command line on the new server.</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>wget http://wordpress.org/wordpress-3.4.1.zip</span></code></pre></td></tr></table></div></figure>


<p>Zip, it all downloaded fine.</p>

<h3>Extra help</h3>

<p>There is a good trouble shooting/dev plugin for WordPress called <a href="http://wordpress.org/extend/plugins/core-control/">Core Control</a>. By installing this and enabling the HTTP control, you can test the various transports WordPress uses for downloads. Curl, PHP File open, PHP Fsockopen etc.</p>

<p>Testing all three default transports, starting with cURL came up with the same message, <code>could not connect to host</code></p>

<h3>Head Scratching time</h3>

<p>So WordPress was telling the truth. Now, as WordPress was running as the <code>www-data</code> user, what happens if I tried the wget from the command line for that user?</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>wget http://wordpress.org/wordpress-3.4.1.zip
</span><span class='line'>Permission denied!</span></code></pre></td></tr></table></div></figure>


<p>Ah-ha, so it was nothing to do with WordPress, the <code>www-data</code> user couldn&#8217;t download anything.</p>

<h3>Problem solved!</h3>

<p>Firewalls, I started thinking firewalls and lo and behold, <a href="https://projects.bytemark.co.uk/projects/24/wiki/Firewall">BytemarkUK block outgoing http traffic</a> for the www-data user by default! This makes sense for when sites get compromised, but it also stops some of the core functionality for updates etc working.</p>

<p>Enabling http traffic for the www-data user will make things worse if the site is compromised, but also makes it a lot easier to keep the site up to date, reducing the chances of a WordPress exploit.</p>

<p>The following removes the firewall rule</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>rm /etc/symbiosis/firewall/outgoing.d/50-www-data</span></code></pre></td></tr></table></div></figure>



]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Poor old java]]></title>
    <link href="http://jamesjefferies.com/2012/09/04/poor-old-java/"/>
    <updated>2012-09-04T12:11:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/09/04/poor-old-java</id>
    <content type="html"><![CDATA[<p>There seems to be a lot of noise at the moment about the latest security vulnerabilities in Java, especially Java 7. I&#8217;d go as far as to say that some of the articles are <a href="http://www.pcadvisor.co.uk/news/security/3379150/time-give-java-boot/">scare stories</a>. Is this what we need, to give Java the boot?</p>

<h2>TL;DR</h2>

<p>All computers connected to things have security vulnerabilities of various forms.</p>

<p>Desktop Java installed computers where not required is an unnecessary security risk and should be uninstalled, just as Adobe Flash, Adobe Reader, Silverlight etc. Windows 8 and Mac OS X don&#8217;t bundle Java runtimes via default anymore. This is a good thing. Oracle should be far quicker out of the blocks fixing security vulnerabilitles.</p>

<p>Giving Java the boot full stop is an over the top reaction. However, giving Java the boot on the desktop when it&#8217;s not required is a wise move.</p>

<h2>History</h2>

<p>Java as a platform has a reputation for being secure, reliable and frequently patched. Being able to run Java server side on Sun software provided the building blocks for many web businesses at the end of the last millenium and in to this one.</p>

<p>It also has many branches in to other parts of the internet ecosystem. As well as running server side, the other two main areas are mobile Java, J2ME, for example and desktop Java, either via browser plugin Applets or as runnable GUI applications using a desktop JVM (virtual machine).</p>

<p>Over the last few years, mobile Java has basically been nobbled by the rise of smartphones and their respective application frameworks and stacks, Android, iOS, Windows whatever it&#8217;s called, BlackBerry and the rest.</p>

<p>However, the legacy of desktop Java has continued by some applications still requiring a JVM installed on the local PC and some people still writing or using Java applets. Althoguh Sun and now Oracle have continued to try and encourage desktop development with their JavaFX initative, they&#8217;ve not got very far.</p>

<h3>Applets</h3>

<p>The idea behind applets was that you visited a site via a web browser, which could always deliver the latest version of the application code to your browser, either to run in a plug-in or using the JVM installed on your machine. They were useful for some niche applications, you sometimes see them now replacing flash file uploaders for example, but the User Experience was often a bit shonky.</p>

<h3>Desktop Applications</h3>

<p>Write Once, run anywhere was the mantra behind Java Desktop applications, an attractive proposition. One set of code would run on Windows, Linux, Unix variants &amp; Macs. Many Java development environments have been written in Java and are used today. CrashPlan is an example cross platform application which requires desktop Java to work.</p>

<h2>Legacy</h2>

<p>Unfortunately, a lot of the Java runtimes (and Flash installs, adobe reader installs etc), installed in days of yore on people&#8217;s machines will have security vulnerabilities, either because they are out of date or because of recent new exploits.</p>

<p>It was interesting to see Apple&#8217;s response to recent JVM exploits, although they may have been slow in responding, they, in effect, switched off people&#8217;s Java installations, providing a prompt if it was required. In one OS X update (assuming it was installed!) Macs became better protected from Java exploits.</p>

<h2>Summary</h2>

<p>If you don&#8217;t use Java on your machine - uninstall it, it&#8217;s not worth having a potential vulnerability. Same with Flash or Adobe reader - if you don&#8217;t use them, uninstall them.</p>

<p>If you do need it installed on your machine, there is some great advice <a href="http://www.macworld.co.uk/macsoftware/news/?newsid=3378690&amp;pagtype=allchandate">here</a></p>

<p>Of course, keep things up to date that you do use, Operating systems, browsers etc</p>

<p>Java server side is a different story. If you do use the Java Virtual Machine, whether it is for Java apps, Clojure, Scala or whatever. Keep your eyes on the Oracle security bulletins, update when you can and be aware of any vulnerabilities in the latest version.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[The Story of Cathy and Heathcliff]]></title>
    <link href="http://jamesjefferies.com/2012/08/30/the-story-of-cathy-and-heathcliff/"/>
    <updated>2012-08-30T16:29:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/08/30/the-story-of-cathy-and-heathcliff</id>
    <content type="html"><![CDATA[<p>Back, back, back in the day&#8230; well, January 1978, Kate Bush released the classic song Wuthering Heights, inspired by the Emily Bronte novel. As well as launching her career, and boosting the sales of flowy white dresses, the song provided inspiration for us when it came to one of our projects.</p>

<h2>Go Free Range</h2>

<p>James Adam of <a href="http://gofreerange.com">GoFreeRange</a> has done a lot of work on transforming simple thermal receipt printers in to intelligent, ethernet enabled printing devices. He’s done an incredible job of writing software and working out the electronics, based on the Arduino device, to get them up and running.</p>

<p><img src="http://jamesjefferies.com/images/nyancat.png" alt="nyancat" /></p>

<h2>Cathy &amp; Heathcliff</h2>

<p>When our two printers arrived to play their part in the Happenstance project, their default names were a bit of a mouthful, they made a good password, but not a name which slipped off the tongue - 4j8k6i5x5q7a7b5y.</p>

<h2>Up and running</h2>

<p>So, when we finally got the printers up and running, after a few head scratching hours, the first message we sent to 4j8k6i5x5q7a7b5y was, ‘It&#8217;s me, Cathy, I&#8217;ve come home’ – a line from Cathy to Heathcliff in the song. So Cathy named herself, and 4j8k6i5x5q7a7b5y became Heathcliff.</p>

<p><img src="http://jamesjefferies.com/images/output.png" alt="output" /></p>

<p>That was in the final week of the Bill Drummond exhibition, “Ragworts” at Site Gallery and we gave the two printers a part to play in the project. Some local Sixth Form students were tasked with sending messages to Cathy &amp; Heathcliff as they jumped on the nearest bus they could find, and journeyed into the wilds of outer Sheffield.</p>

<p>At first we thought they could tweet their progress and the printers would pick up on those tweets and print them, but we decided to mash up a lot of technology to provide a simpler solution for the students. With the clock ticking, we glued together Twilio, Gmail, ifttt (‘If This Then That’) and some custom code we wrote. Cathy &amp; Heathcliff now had their own Sheffield phone numbers, and now you could send them a text message and the contents would be printed and tweeted.</p>

<p>When the students set off on their journeys, they could send SMS text messages to Cathy and Heathcliff’s numbers and the magic of technology would take care of the rest. Their messages would appear on the printers, back at Site Gallery and on their Twitter account. It was fascinating to read the messages sent back to base by the students, many of which were very poetic and evocative.</p>

<p>After all this excitement, our plan was to get Cathy and Heathcliff mobile so they could roam the streets of Sheffield whilst remaining available for printing. With the aid of dongle, a mac and a giffgaff sim card they were freed from the shackles of mains power and wifi, ready to explore!</p>

<p><img src="http://jamesjefferies.com/images/getonabus.png" alt="get on a bus" /></p>

<h2>What&#8217;s next</h2>

<p>Although Happenstance has now finished, the printers have been used in Site for other purposes. Cathy has been on reception acting as a comments book (text her and she’ll print your comment) whilst Heathcliff has been reading twitter and printing mentions of Site Gallery!</p>

<p>We still hope to use them for one final experiment where we get them to run off steam power!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Setting up WordPress]]></title>
    <link href="http://jamesjefferies.com/2012/08/22/setting-up-wordpress/"/>
    <updated>2012-08-22T15:07:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/08/22/setting-up-wordpress</id>
    <content type="html"><![CDATA[<p>So, your server is set up and raring to go. Please see my <a href="http://jamesjefferies.com/2012/08/22/setting-up-a-fresh-debian-server/">earlier blogpost</a> on getting the server to this point.. next!</p>

<h2>Install Apache2</h2>

<p>Let&#8217;s get our web server up and running by installing apache</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install apache2</span></code></pre></td></tr></table></div></figure>


<p>and check it is up and running with <code>http://123.456.789.012</code> in your friendly web browser.</p>

<h2>Configure Apache2 for low memory usage</h2>

<p>With previous virtual servers, I&#8217;ve found that Apache needs a bit of tweaking to support WordPress load. I found this useful <a href="http://imperialwicket.com/tuning-apache-for-a-low-memory-server-like-aws-micro-ec2-instances">post</a> about tweaking apache. The nuts and bolts are to amend some of the settings in your <code>/etc/apache2/apache.conf</code> as follows:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
<span class='line-number'>5</span>
<span class='line-number'>6</span>
<span class='line-number'>7</span>
<span class='line-number'>8</span>
<span class='line-number'>9</span>
<span class='line-number'>10</span>
<span class='line-number'>11</span>
<span class='line-number'>12</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>Timeout 30
</span><span class='line'>KeepAlive On
</span><span class='line'>MaxKeepAliveRequests 50
</span><span class='line'>KeepAliveTimeout 10
</span><span class='line'>
</span><span class='line'>&lt;IfModule mpm_prefork_module&gt;
</span><span class='line'>    StartServers          5
</span><span class='line'>    MinSpareServers       5
</span><span class='line'>    MaxSpareServers      10
</span><span class='line'>    MaxClients          150
</span><span class='line'>    MaxRequestsPerChild   0
</span><span class='line'>&lt;/IfModule&gt;</span></code></pre></td></tr></table></div></figure>


<h2>Setting up virtual host with apache</h2>

<p>A very simple set up would be to create the following <code>/etc/apache2/sites-available/myacedomainname.com</code></p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
<span class='line-number'>5</span>
<span class='line-number'>6</span>
<span class='line-number'>7</span>
<span class='line-number'>8</span>
<span class='line-number'>9</span>
<span class='line-number'>10</span>
<span class='line-number'>11</span>
<span class='line-number'>12</span>
<span class='line-number'>13</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>#Ensure that Apache listens on port 80
</span><span class='line'>#Listen 80 (note uncomment this with ports.conf set to 80 and you will experience a WORLD of pain!)
</span><span class='line'>
</span><span class='line'># Listen for virtual host requests on all IP addresses
</span><span class='line'>NameVirtualHost *:80
</span><span class='line'>
</span><span class='line'>&lt;VirtualHost *:80&gt;
</span><span class='line'>DocumentRoot /var/www/myacedomainname.com
</span><span class='line'>ServerName www.myacedomainname.com
</span><span class='line'>
</span><span class='line'># Other directives here
</span><span class='line'>
</span><span class='line'>&lt;/VirtualHost&gt;</span></code></pre></td></tr></table></div></figure>


<h2>Enable site (and other useful commands)</h2>

<p>Create that file in sites-available and use the following commands to enable/disable - they will give you options when they run as to which sites you want to enable/disable</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>sudo a2ensite
</span><span class='line'>sudo a2dissite</span></code></pre></td></tr></table></div></figure>


<p>Force reload configuration?</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>sudo  /etc/init.d/apache2 force-reload</span></code></pre></td></tr></table></div></figure>


<h2>Install PHP</h2>

<p>Ok, well to get WordPress running, we&#8217;re going to need some PHP, so get it installed</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install php5 libapache2-mod-php5</span></code></pre></td></tr></table></div></figure>


<h2>Install MySQL</h2>

<p>And we&#8217;ll need a database too</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install mysql-server mysql-client php5-mysql</span></code></pre></td></tr></table></div></figure>


<p>If already installed, but you want to reset it so you can set root password etc, give it a bit of this:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>sudo dpkg-reconfigure mysql-server-5.1</span></code></pre></td></tr></table></div></figure>


<h2>Migrating a site from another provider</h2>

<h3>Database gubbins</h3>

<p>Get a database dump from your old site, either using PHPMyAdmin if installed, or doing a dump from the command line using <code>mysqldump</code>. Hopefully at the end of that process you&#8217;ll be able to import the dump in to your new set up with a bit of:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>mysql -uroot -p &lt; databasedump.sql</span></code></pre></td></tr></table></div></figure>


<p>The dump may contain a create database statement to create the database, right at the top along the lines of</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>CREATE DATABASE `db1234` DEFAULT CHARACTER SET latin1 COLLATE latin1_german2_ci; USE db1234;</span></code></pre></td></tr></table></div></figure>


<p>Now you&#8217;ll need to create a database user for WordPress to use. So make sure you&#8217;ve got a file dump of your WordPress installation, either by command line, or the hosting admin suite, or an FTP application or whatever you&#8217;re comfortable with. You&#8217;re going to need your wp-config.php file which will tell you how your existing WordPress site is set up. This is the bit you need to look at for now</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
<span class='line-number'>5</span>
<span class='line-number'>6</span>
<span class='line-number'>7</span>
<span class='line-number'>8</span>
<span class='line-number'>9</span>
<span class='line-number'>10</span>
<span class='line-number'>11</span>
<span class='line-number'>12</span>
<span class='line-number'>13</span>
<span class='line-number'>14</span>
<span class='line-number'>15</span>
<span class='line-number'>16</span>
<span class='line-number'>17</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>/** The name of the database for WordPress */
</span><span class='line'>define('DB_NAME', 'db1234');
</span><span class='line'>
</span><span class='line'>/** MySQL database username */
</span><span class='line'>define('DB_USER', 'dbuser');
</span><span class='line'>
</span><span class='line'>/** MySQL database password */
</span><span class='line'>define('DB_PASSWORD', 'dbpassword');
</span><span class='line'>
</span><span class='line'>/** MySQL hostname */
</span><span class='line'>define('DB_HOST', 'localhost');
</span><span class='line'>
</span><span class='line'>/** Database Charset to use in creating database tables. */
</span><span class='line'>define('DB_CHARSET', 'utf8');
</span><span class='line'>
</span><span class='line'>/** The Database Collate type. Don't change this if in doubt. */
</span><span class='line'>define('DB_COLLATE', '');</span></code></pre></td></tr></table></div></figure>


<p>So in your existing set up you have a database user called <code>dbuser</code> with password <code>dbpassword</code> so create that user in your new setup as follows</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>grant all on db1234.* to 'dbuser'@'localhost' identified by 'dbpassword';</span></code></pre></td></tr></table></div></figure>


<p>Then make sure it works with a bit of <code>mysql -udbuser -Ddb1234 -p</code> enter your password and hopefully you are in.</p>

<h3>Filesystem</h3>

<p>Now get that dump of the filesystem you have and unpack it in <code>/var/www/myacedomain.com</code> or whatever you set in your virtual hosts configuration earlier. You may need to change the ownership of everything, including that directory to www-data with a bit of</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>sudo chown -R www-data:www-data /var/www/myacedomain.com</span></code></pre></td></tr></table></div></figure>


<p>
and check the permissions too, directories should be set to 755 and files to 644. You can enforce this with</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>sudo find . -type f -print | sudo xargs chmod 644  # Files
</span><span class='line'>sudo find . -type d -print | sudo xargs chmod 755  # Directories</span></code></pre></td></tr></table></div></figure>


<p>It is entirely possible that after all this, you can now hit <code>http://myacedomain.com</code> or <code>http://123.456.789.012</code> and see your migrated WordPress site!</p>

<h3>UTF-8 issues</h3>

<p>Now, one issue I&#8217;ve had in the past is problems with UTF-8 conversion. I.e. the posts you&#8217;ve migrated end up with funny characters like – “ ’ †- you can tell if this is causing you grief because you see the characters mixed up in your posts, but if you comment out these lines (by adding // in front) in your WordPress config, then they.. ahem, magically go away</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>define('DB_CHARSET', 'utf8');
</span><span class='line'>define('DB_COLLATE', '');</span></code></pre></td></tr></table></div></figure>


<p>Best not to do that though, you <em>can</em> use the <a href="http://wordpress.org/extend/plugins/search-and-replace/">search and replace</a> WordPress plugin to replace all the dodgy characters in the database. Find &#8220;â&#8221; and replace with &#8220;&#8221; - hooray!</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[Setting up a Fresh Debian Server]]></title>
    <link href="http://jamesjefferies.com/2012/08/22/setting-up-a-fresh-debian-server/"/>
    <updated>2012-08-22T14:41:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/08/22/setting-up-a-fresh-debian-server</id>
    <content type="html"><![CDATA[<h2>Accessing the server</h2>

<p>Once set up, you will have the IP address, in the form 123.456.789.012 and the root password. This means you will be able to log on to the server using ssh (secure shell), i.e.</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ssh root@123.456.789.012</span></code></pre></td></tr></table></div></figure>


<p>You will get a message along the lines of</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>The authenticity of host '123.456.789.012(123.456.789.012)' can't be established. RSA key fingerprint is blah blah bla
</span><span class='line'>Are you sure you want to continue connecting (yes/no)?</span></code></pre></td></tr></table></div></figure>


<p>to which you will reply</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>yes</span></code></pre></td></tr></table></div></figure>


<p>and you will then get</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>Warning: Permanently added '123.456.789.012' (RSA) to the list of known hosts.
</span><span class='line'>root@123.456.789.012's password: </span></code></pre></td></tr></table></div></figure>


<p>which is when you can enter your password. Hopefully then you&#8217;ll be logged on to the server.</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>blah blah blah snip…
</span><span class='line'>servername:~# </span></code></pre></td></tr></table></div></figure>


<h3>First things first</h3>

<p>Root access is the super user for your server. You really don&#8217;t want to have people being able to log on to your server as root. What you really need is to have user accounts which people can use to log on, and if/when required, they can get super user privileges to do whatever they need to do.</p>

<p>What we are going to do is to stop root being able to secure shell straight in to the server, create a new user account to use for our admin (with upgradable privileges) and run ssh on a different port to deter basic attacks.</p>

<h3>Change root&#8217;s password!</h3>

<p>So you have a default root password - CHANGE IT to a new one! From a root prompt, use <code>passwd</code> to change the password</p>

<h3>Adding a new user</h3>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>adduser james</span></code></pre></td></tr></table></div></figure>


<p>Will create a new user, called James - you can add in extra details as you go like full name etc.</p>

<p>Now we want to allow james a bit of privilege, not by default though, but by using the sudo command.</p>

<h4>Aside - setting default editor to be vi</h4>

<p>By default, for all users, you may want to use vi as your editor (ok, you may not, but I do!) so add the following to <code>/etc/profile</code></p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>EDITOR=vi
</span><span class='line'>export EDITOR</span></code></pre></td></tr></table></div></figure>


<h3>Giving the new user sudo privileges</h3>

<p>You may need to install sudo if it isn&#8217;t already installed on your server. It&#8217;s a simple case of</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install sudo</span></code></pre></td></tr></table></div></figure>


<p>Then, the application <code>visudo</code> is your friend. You edit the config file using visudo and then sudo does the rest. When you open the file, you&#8217;ll see a section with:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
<span class='line-number'>3</span>
<span class='line-number'>4</span>
</pre></td><td class='code'><pre><code class=''><span class='line'># User privilege specification
</span><span class='line'>root    ALL=(ALL) ALL
</span><span class='line'>admin ALL = (ALL) ALL
</span><span class='line'>james ALL = (ALL) ALL</span></code></pre></td></tr></table></div></figure>


<p>By adding my own entry, it allows me to upgrade my user to super user privileges for all actions. You can limit the commands which users can run if you like using this, but if you are going to be the super user, you probably want to leave it as all.</p>

<h2>Sorting out access</h2>

<p>Let&#8217;s stop people attempting to log on to the server as <a href="http://www.howtogeek.com/howto/linux/security-tip-disable-root-ssh-login-on-linux/">root via ssh</a> and also to run ssh on a different port than the default. There is a lot of good advice <a href="http://www.foogazi.com/2006/11/29/modify-ssh-to-maximize-security/">here</a></p>

<h3>Amend port</h3>

<p>The default port is 22, change this to something else, which isn&#8217;t being used by anything else</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'># What ports, IPs and protocols we listen for
</span><span class='line'>Port 60</span></code></pre></td></tr></table></div></figure>


<h3>Stop root login</h3>

<p>Now stop people logging in as root</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>PermitRootLogin no</span></code></pre></td></tr></table></div></figure>


<h3>Some other bits</h3>

<p>Give maximum number of log in attempts to be 3, only allow james to login.</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
<span class='line-number'>2</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>MaxAuthTries 3
</span><span class='line'>AllowUsers james </span></code></pre></td></tr></table></div></figure>


<h3>Restart sshd</h3>

<p>As root (or sudo)</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>/etc/init.d/ssh restart</span></code></pre></td></tr></table></div></figure>


<p>Now try logging in remotely from a different shell (i.e. keep the one you&#8217;ve just restarted sshd on open in case you have any problems!)</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ssh -p60 james@123.456.789.0123</span></code></pre></td></tr></table></div></figure>


<p>obviously set the p to be the port you set earlier.</p>

<h3>Stop chancers getting in</h3>

<p>It&#8217;s also a good idea to stop people attempting to log on to your server using common passwords, usernames etc. A good way to do this is to install the <a href="http://en.wikipedia.org/wiki/Fail2ban">fail2ban</a> which goes some way to banning people trying to brute force their way in to your server.</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install fail2ban</span></code></pre></td></tr></table></div></figure>


<p>Default settings are pretty good here.</p>

<h2>Firewall configuration</h2>

<p>UFW or Uncomplicated Firewall is a good first port of call for securing your shiny new server. You can install it as root with</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>apt-get install ufw</span></code></pre></td></tr></table></div></figure>


<p>Now, unlike fail2ban, ufw is installed switched off. You need to configure a few things before getting it up and running and providing that extra security. By default ALL ports are shut, so make sure you&#8217;ve opened up the ones you need before switching the firewall on!</p>

<p>For example, if you&#8217;ve set up your ssh to run on port 60, then you need to run</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ufw allow 60</span></code></pre></td></tr></table></div></figure>


<p>If you&#8217;re going to be running apache on port 80 (the usual)</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>ufw allow 80</span></code></pre></td></tr></table></div></figure>


<p>Here are a few <a href="http://www.mmncs.com/2011/07/this-guide-shows-howto-install-and-setup-uncomplicatedfirewall-ufw-on-linux-ubuntu-10-04-and-11-04-and-other-debian-distributions/">notes</a></p>

<p>Don&#8217;t forget to run <code>ufw enable</code> when you are ready! <code>ufw status</code> tells you what is set up.</p>
]]></content>
  </entry>
  
  <entry>
    <title type="html"><![CDATA[From MySQL datadump to mongo]]></title>
    <link href="http://jamesjefferies.com/2012/08/16/from-mysql-datadump-to-mongo/"/>
    <updated>2012-08-16T14:42:00+01:00</updated>
    <id>http://jamesjefferies.com/2012/08/16/from-mysql-datadump-to-mongo</id>
    <content type="html"><![CDATA[<p>Let us say that you have a MySQL dump of  a database table, with a load of data which ideally, you&#8217;d like to pull into a Mongo DB as json.</p>

<p>What you could do is…</p>

<h2>Import data in to MySQL database</h2>

<p>Usual kind of thing here, assuming you have a MySQL username for a created database:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>mysql -uusername -p -Ddbname &lt; mysqldumpfile.sql</span></code></pre></td></tr></table></div></figure>


<h2>Export data from MySQL database in json format</h2>

<p>There is a cracking ruby gem <a href="https://github.com/seamusabshere/mysql2xxxx">mysql2xxxx</a> which will export from a mysql database to various formats, including json.</p>

<p>You can drive it via code or a number of binaries are provided. So:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>mysql2json --user=username --password=password --database=dbname --execute="select * from my_special_table" &gt; my_special_file.json</span></code></pre></td></tr></table></div></figure>


<h2>Importing the json straight in to your mongodb</h2>

<p>Then you can use mongoimport to pull in the json. If your export from <code>mysql2json</code> is surrounded by the square brackets, you&#8217;ll need the <code>--jsonArray</code> option. Your command will need to be something like:</p>

<figure class='code'><div class="highlight"><table><tr><td class="gutter"><pre class="line-numbers"><span class='line-number'>1</span>
</pre></td><td class='code'><pre><code class=''><span class='line'>mongoimport --type json --file my_special_file.json --collection mycollection --db mymongodb --jsonArray</span></code></pre></td></tr></table></div></figure>


<h2>Summary</h2>

<p>Simple three step transformation from MySQL dump to Mongo. You may know a better, more straightforward way of doing this, please add a comment if you do!</p>
]]></content>
  </entry>
  
</feed>