For 10 years Bytemark been technically-focussed, and our engineering staff have vastly outnumbered everyone else (the opposite ratio to most hosting companies). But with the growth of BigV and our data centre, it’s time to tweak that balance a little bit. I’m looking for people who can write and talk about what we do, showing the advantages of hosting with Bytemark to both technical and business-minded people and parading our company culture to the world.

Both positions will break new ground inside Bytemark – within 12 months, the people we hire will be defining the way we sell and market ourselves. If you’re looking to jump out of the rank & file, this could be your chance.

As usual for jobs at Bytemark, there is a screener question that you’ll need to answer and submit with your CV.

The two positions are:

• Head of External Communications (£25-40k): an expert in short-form writing, advertising & web site maintenance.
• Sales Director (£60k): someone who loves juggling technical planning for a plenty of incoming customers and enjoys the challenge of a performance-driven position.

Follow the links for full details of both; they are based in York and feature our usual excellent benefits. Good luck to everyone applying!

Of course, servers can contain lots of different pieces of hardware (many of which can be emulated by QEMU), and different operating systems have varying levels of support for them. When deciding how to configure QEMU for our BigV virtual machines, we wanted the defaults to be as fast as possible, but we also wanted to be able to support a wide variety of operating systems.

We met these contradictory goals by introducing the idea of ‘hardware profiles’. Until today your virtual machine was given a profile called virtio2011 (or, if you asked for it, compatibility2011). By picking one hardware profile or another, you were deciding what hardware QEMU would emulate for your VM - fast-but-narrowly-supported, or slower-but-widely-supported. The hardware profile also decides which version of QEMU is used to run your virtual machine, with the 2011 profiles using a modified version of QEMU 0.15 – which actually dates from 2008!

Of course there have been a lot of changes to QEMU in 5 years. While we’re happy with the hardware that QEMU presents, and its stability there are still loads of bug fixes, performance enhancements, and behind-the-scenes features we’d like to start using in BigV.

As of right now, any new virtual machines you create on BigV will use the new virtio2013 hardware profile, which uses the brand new QEMU 1.4.1. If you didn’t specify a hardware profile when creating a virtual machine, it will have defaulted to virtio2011, and next time it’s stopped (by running bigv vm shutdown or bigv vm restart, for instance), it will be automatically upgraded to virtio2013. This is almost invisible to your guest – when it checks the CPU’s model name, it will see “QEMU Virtual CPU Version 1.4.1″ instead of “0.15.0″, but that should be all.

VMs that were created with a hardware profile of compatibility2011 will, by constrast, not be upgraded to compatibility2013 just in case there’s a very subtle change that somehow breaks your OS. We also won’t automatically upgrade a VM if you’d specified virtio2011 explicitly, just in case you had a reason to do so, and we’ll stick to that for any future VMs where you choose a profile.

If you’d previously specified compatibility2011 or virtio2011 for a VM and you’d like it to be upgraded to the 2013 release, the client can’t do it yet – so pop in a support request to support@bigv.io and we’ll do it for you (it will need a reboot). We’d also like to hear from you if there’s new virtual hardware you’d like to see added to BigV. For now we think the virtio and compatibility profiles cover most use cases but please tell us if you think we’re missing a trick.

Finally, we recently identified a bug in QEMU that was harming disc I/O speeds. We’ve fixed this for all our hardware profiles. This reduces I/O request latency by a factor of 40-200. This translates to some fairly drastic improvements in the performance of everyday tasks – compiling a typical C++ application was seven times faster, in one example – and differences between the various disc storage grades should now be more obvious. So far feedback has been positive, but if you’ve had I/O problems now is the time to reboot your VM, check again and let us know.

With Associates have built the site for Code Club, “a nationwide network of free volunteer-led after-school coding clubs for children aged 9-11″ and host it on BigV.

We love the idea of giving schoolchildren a fun after-school activity which can easily turn into a lifelong interest and marketable skill. Most of us at Bytemark got started that way, but experience of computers in the 80s and 90s came from generous parents, uncommonly ambitious schools and teachers going above and beyond the call of duty. We like that Code Club is helping to level the field.

Jamie White of With Associates:

Code Club came to us at the start of the year to see if we might be interested in building the next version of their website. We jumped at the opportunity, and the past few months have seen us working closely with co-founder Clare Sutcliffe to realise her vision. It launched last month, and so far it’s been a great success.

The Code Club site (pictured above) uses a pretty recent set of technology:

• Slices, With Associates’ in-house content management system (built on Rails);
• The database is MongoDB;
• It uses queues extensively, via MongoDB and the Qu gem;
• It plugs into a bunch of other services, including:
• Google for geocoding (soon to be switching to Postcode Anywhere)
• Mandrill for email delivery
• Intercom for CRM/support

Jamie will be at the Scottish Ruby Conference next week courtesy of Bytemark, and we look forward to seeing more children programming from theirs, and Code Club’s efforts.

Here’s the generator, an SDMO V550:

and here’s the crane:

And here’s the lift; the wind picked up slightly scarily before they started:

Once we’ve commissioned it, our data centre will have a “prime” 500kVA generating capacity, keeping us going off the grid for 20 hours at our initial full load without needing the tank refilled. We’ll be adding a new tank as data centre load requires and we won’t require its full capacity for a few more months at least.

There’s also space behind the data centre for two more generators, covering us for the full possible data centre load at N+1 as we build out the ground floor further.

We’re giving away an all-expenses paid trip to ScotRuby 2013 to one lucky person and a friend.

Do you have the “most interesting use of BigV”?

Are you running a PBX/VPN service, perhaps you’ve written an exciting plugin for your CMS, or perhaps something else? It doesn’t need to be related to Ruby or be a website but it does have to be hosted on BigV.

If you’re not on BigV yet, whether you’re a new or an existing customer, getting online is simple – there’s still time to enter.

One winner will get:

• Travel & food expenses
• 2 Conference Tickets
• 2 single occupancy hotel rooms Crieff Hydro from the day before the conference, until the day after (11th-14th)

And of course, we’ll blog about the winners

To enter, simply email competitions@bytemark.co.uk before 15:00 BST Friday 03/05/2013 with details of the thing you’re entering.

Unfortunately sponsored customers and Bytemark staff are ineligible for this competition – sorry guys, we love you already!

Winners will be notified by email & announced on this blog at the start of the week beginning 6th of May. There is no cash alternative. The judge’s decision is final. We might decide to award two individual winners if the first person doesn’t have a friend they want to bring.

The Crieff Hydro hotel where the conference will be held:

Eagle-eyed Linux users may have noticed the (very quiet) introduction of a Qt-based front-end in the last release; some work has gone into polishing this for 0.7, and we think it’s good enough for casual use now. Equally eagle-eyed Mac users will have noticed that a DMG-style installer has also been introduced for BigV, replacing the previous manual installer. Unfortunately, we couldn’t get the GUI client packaged in this form for 0.7 – hopefully, we’ll overcome the relevant hurdles in the next release or two.

The command-line client is still the reference implementation, and the GUI doesn’t expose every feature, but it may be more convenient for the cases it does support – do have a play. You’ll find it in the “bigv-gui” package for the Linux distributions we support, and it’s installed alongside the command-line client if you use the Windows installer. Speaking of packages, our Linux ones should now work properly on Debian Wheezy (Ruby 1.8 or 1.9) and Fedora Core 19 (Ruby 2.0).

That’s not all we’ve been working on, of course – bigv vm shutdown and bigv vm stop now do different things (the former is graceful, like pressing the power button; the latter is like pulling the power out) and you can send a hard reset signal too (bigv vm signal –vm-signal=reset). You can also now rename VMs (bigv vm rename), delete accounts (bigv account delete), and run the client behind a HTTP proxy server (the –http-proxy option to any command). When creating or reimaging a VM, Debian Wheezy, Ubuntu Precise, and Windows Server 2012 have been added as options (we don’t have an Ubuntu Raring image yet!)

Finally, a bushel of bugs have been squashed, although I’m sure many remain. If you happen to spot one in the wild, please do let us know – especially if it’s to do with the graphical client – and we’ll do our level best to fix it. Otherwise, enjoy!

Those are the first two cabinets of our Uninterruptible Power Supply system (with another two behind), the buffer between the servers and the mains supply or the diesel generator. We can pull the plug on the mains and switch to the generator, and batteries will keep us going for a few minutes.

We can even turn some very big switches to take a UPS out of the loop for maintenance:

Here’s one of many shelves of batteries:

In the early days, these batteries will give us more of a buffer than we need – probably an hour or more depending on how quickly we fill up the main room. And we will expand both the UPS unit themselves and the batteries without interrupting the supply to the equipment.

Here’s the hole where our electricity substation will be, out the front of the building.

at the back here’s the rear compound, and you can see the concrete pad where the diesel generator will be hoisted in a few days time:

There’s room there for two more generators, to cover us for full site load, and also in the case one fails. It’s a bit over-the-top but by having N+1 generators (and not just power sources) there are particular external data centre standards that we can meet fairly easily, which impresses government and other compliance types.

Also in the UPS room is the cabinet for our fancy building management system:

That’s where the controller will go that has access to every switch and measurement in the data centre: fire suppression, air conditioning power, fresh air cooling vents, missile defences etc. For obvious reasons, it’s not connected to any other part of our network (look what happened to all the other Battlestars…).

One of the main things it does is decide when to switch between fresh air cooling and direct expansion (DX) cooling methods. So here are the air handling units in the main room which push the air back and forth:

Now they can pass air over these external units, which basically are enormous refrigerators that work by direct expansion cooling (thanks wikipedia):

But, as we mentioned before, we’ve done our sums and the same units can also simply open the vents to the chilly York air, as long as the outside temperature is lower than 16 degrees:

Whoosh. That’s most of the year, in fact. So we’re not using the power-hungry direct expansion cooling most of the time, and we’re hoping for a Power Usage Effectiveness figure of 1.1 for the cooling side of things (i.e. we’re hoping to only have to use 10% extra power to run all the cooling).

For the times when the building management system decides it’s too hot and closes all the vents to the outside world, using DX cooling the whole time, it would be nearer 1.5

There will also be losses introduced by the UPS system, but are hoping that overall the PUE will be under 1.2 on a normal chilly day.

Back inside, here’s the guts of the air handling unit, including the big pumps necessary to push the cooling fluid in and out:

The building management system also has the crucial decision of when to let these babies off:

That’s the Kidde FM200 Fire Suppression system, masses of high-pressure liquid which turns into a fire-suppressing gas when blasted out of the tubes criss-crossing the main room. If they have to go off, the gas doesn’t hurt anyone in the room (there used to be carbon-dioxide systems which basically suffocated people along with the fire), and the power doesn’t need to be cut, so the data centre soldiers on. But the bottles are one-shot, and refilling them costs quite a lot. So there are various measures to make sure it doesn’t happen too automatically.

To decide whether to sound the alarm or fire the suppression system, the building management system takes inputs from a set of sampling points across the ceiling, little tubes that look like this one:

There’s a box that sucks the air samples down, and using a laser decides how “smokey” things are getting:

Ultimately it’s the building management system that decides what to do, and that’s down to some high-stakes configuration and policy decisions by Peter and me.

They’re actually building this data centre fast than I can write and take pictures, so it looks even better now. And there’s an engineering treat in store for Friday 26th.

Our businesses are built on an enormous foundation of free software.

Web hosting remains solid, but we’ve all lost customers in the last 10 years to proprietary email platforms, and have just collectively just accepted it. Mail solutions based on exclusively free software are looking second rate compared to GMail, Exchange and other vertical systems where we can’t add value through our expertise, or give users the choice of where they host.

Facebook and Twitter continue to use their muscle to wind internet messaging back to the 1980s. That was a time when sending a message between big commercial networks was a privilege and not a right. So you had multiple addresses, or people you had to pay to talk to, or people you just couldn’t talk to because those big networks wanted to lock people in. We’re getting back to that state again now.

In 2012, Mozilla publicly pulled staff from future development of Thunderbird. It used to be the best email client, bar none. But it had fallen behind before, and now looks like it’ll continue to do so, given how large a code base it is.

All of this bad news means hosting companies lose business. Users see a better experience with one of the big guys, moving towards them and their proprietary platforms. Among many other things, we all need to see a better free email client, and the prospect of anyone starting afresh seemed pretty remote.

Geary: A beautiful, modern, open source email client.

Geary is a brand new desktop email client – in the early stages of development.

It’s being developed by a San Francisco-based non-profit called Yorba – you may already know them – they’re the talented and proven group of hackers who built the slick photo manager for the Linux desktop called Shotwell.

Yorba simply want to put beautiful, functional, software out there, for free, with no strings attached, and with no plans to lock away the best features for paying customers. They’re funded by donations and consulting work and have turned to a crowdfunding campaign to pay for the development.

That means Yorba are writing more of the type of software that will continue to build all our companies. When clearly talented people come together to help us, just for the love of what they do, we must support them to see results.

They’re looking for $100,000 to finish Geary. Bytemark have pledged $2500, and if only a handful of us do the same, we have a great chance of seeing an amazing new email client coming to fruition in 2013.

Geary will help us all to sell more servers, and to grow our industry on open standards, not limited interoperability with giants. So please take a look at their project and donate what you can.

Thanks for reading,

Matthew Bloch
Managing Director, Bytemark Hosting

Debian is the universal UNIX operating system, built by thousands of volunteers across the globe since 1993, and bringing together the best of the free software movement since that time. Over 10% of web servers on the internet run it. And since Bytemark launched in 2002, Debian has been been a major plank of our success – all our internal systems use it, and most of our customers do too.

We’ve even built our Symbiosis easy hosting system on top of Debian. That’s a Bytemark product in use by thousands of domains across our network, and (because we released it as free software), we’ve seen it adopted outside too.

While we can’t match the unpaid efforts of Debian’s thousand of volunteers, we’re at least happy to be providing such a substantial part of Debian’s infrastructure and paying back a little of our debt to the project. Debian’s success will continue to spur ours.

As you can see, the loading bay precisely fits a Moffett’s turning circle – with a generous few inches to spare.

In the same delivery there was a pair of Airedale Easicool units - these make up a separate cooling system for our UPS room.

Both our communication ducts now contain fibre. The fibre isn’t lit yet, but we have some LC patch bays waiting to have our kit plugged into them.

The concrete plinth for the backup generator has finally been poured and set after weeks of delay caused by the cold, wet weather.

Lastly the floor is fully finished in both rooms, and the lights have been switched on! They’ve been hung up and unplugged for weeks, but what a difference it makes to the room over the temporary safety lighting. It suddenly looks like a proper data centre!

That’s it for now – the list of tasks remaining keeps shrinking.  In the next couple of weeks we should have the main incoming LV panel, UPS system, Generator and sub-station arriving. More later…