I tweeted recently about our same day setup for colocation services in our primary London data centres. After speaking to a few people in the industry, the general consensus was – nice marketing Stef, but it can’t be done. Can it?

The tweet in question.

I thought I’d share some examples of situations where our customers have needed that kind of turn around, and how we were able to
provide it for them. Generally we’ve found the root cause of this urgency boils down to a couple of categories; poor project management, emergencies, or a combination of the two.

EXAMPLE 1 – OFFICE MOVE

On Boxing Day 2013, we received a call from a customer based in Scotland who were in the process of moving to new offices. They had a couple of onsite racks, VMware clusters, which some fairly important retail websites relied on for a backend.

The trouble arose when they realised their ISP had failed to meet the delivery deadline for the Leased Line in the new office. As a result, they had no connectivity in and out.

We were very close to driving up to Scotland and moving the equipment down to London which would have meant hours of downtime, but was still possible. While discussing the plans, it was made clear that the critical machines were all inside VMware, at which point we proposed that it would be a lot quicker and safer to simply move the VMware Virtual Machines from their on-site cluster to our enterprise VMware Virtual Data Centre based in the docklands.

This was done overnight, and by the morning, backend services for these sites were running out of the EXN vDC, where they remain to this day.

While in this case we didn’t actually end up moving the physical equipment to London, we certainly were prepared to take the van up to Scotland (with a turkey sandwich), and pull an all-nighter to get the customer up and running. Instead we managed to migrate the VMs over to our platform remotely, out of hours, in a controlled manner.

EXAMPLE 2 – FLOODING

For the last 6 years, I can recall something dramatic happening during the Christmas holidays every year. We’re just lucky like that, I suppose. It all started in 2009, when we lost 4 out of 6 disks in a RAID5 array, and since then there has always been something interesting occur.

In December/January of 2014, the Thames burst its banks around the Maidenhead area, damaging buildings… and in-house server rooms. We had one customer whose basement had started flooding and they needed to get their equipment out. Now. Ordinarily we could set this up for a few servers, no problem, but the customer needed an entire rack, which does take some time to provision.

That being said, within 24 hours the customer was installing their equipment and picking up their permanent access badges at their new data centre.

Care to join us in Global Switch 2?

EXN can provide same day setup on single server (or multiple single servers) at all of our core sites: Global Switch 2, Telecity Meridian Gate, Virtus Enfield, Volta Great Sutton Street and as of August 2015, this will include colocation options in Telehouse North.

All these sites have plenty of spare capacity, and our office location in Greenwich makes all these sites easy to reach. We have several checklists in place to ensure procedures are followed accurately and efficiently – these range from the allocation of the rack space, to the network and DNS setup.

When a colocation order is placed, the first step is to get the customer signed up with our EXtraNet Client Portal, which gives the customer access to our support team, service details, reboot ports, bandwidth details etc. Once the customer is signed up and the initial invoice is paid, several teams start working in unison to get the customer set up. First of all, our data centre team allocate the rackspace and power ports. These details are added to the EXtraNet Portal, giving the customer the ability to reboot remotely. They also liaise with the customer and arrange all relevant delivery IDs and access to the data centre.

During this time, the network team is busy setting up the IP services. If the colocation requires IP transit, this is done as a priority to ensure the customer can have the IP range as soon as possible. Once the network, power and space allocation has been completed, all that is required is for the equipment to be installed and cabled up. At this point an EXN engineer meets either the customer or the courier on site, racks the devices, cables them up and tests.

Job done.