CenturyLink Expanding our Cloud Managed Services to the UK

Today we are pleased to announce one-click access to cloud-based managed services on the CenturyLink Cloud in the UK. CenturyLink Cloud managed services provide automated provisioning of operating systems and applications – check the box when you need them, uncheck when you don’t – all billed hourly.  When paired with our easy-to-use cloud infrastructure, you get up and running quickly with the peace of mind you need from managed apps and operating systems. Now customers in the UK or global customers using UK nodes can mix and match cloud-based managed services, public and private cloud, and even colocation – all under one roof. This suits enterprises who want to embrace hybrid IT as part of their cloud journey, as well as those who just want to leave legacy applications where they are for the time being, but have access to the flexibility of cloud infrastructure when they need it. With this launch, customers can take advantage of the following services from our Slough data center: Windows and Red Hat Enterprise Linux managed operating systems, as well as managed applications Active Directory, Apache HTTP Server, Apache Tomcat, Microsoft Internet Information Services, Microsoft SQL Server and MySQL.

And, if you want even more assistance, our cloud experts stand ready to assist with the installation, configuration, monitoring and maintenance of your cloud infrastructure, so you can focus on other priorities.

Find out more about CenturyLink Cloud Managed Services on our website, reach out to me personally at, or follow me: @jimbattenberg

Designing an office for Collaboration and Culture

Q&A with James Newkirk, CenturyLink Cloud VP of Engineering

This week CenturyLink will celebrate the opening of our Cloud Development Center in Bellevue, WA.  I sat down with James Newkirk, Engineering VP for CenturyLink Cloud, who anchored the team designing and overseeing the build out of the new Center’s space to get his insights on what went into the design.

Why did we select Bellevue, WA as the home of our new Cloud Development Center?

Seattle and its surrounding area are the growing center of the cloud universe. We know that if we want to tap into the growing talent pool of cloud computing experts and developers, the greater Seattle area is a great place to be.

We chose our new location in Bellevue for a couple of reasons.  First, as a nod to our heritage with Tier 3 which was founded and headquartered in Bellevue.  Secondly, was access to transportation – a lot of employees use public transport and our new location is close to the transit center.  Finally it’s very central to great restaurants – we wanted employees to have options to get quality food on lunch breaks or easy access to social outlets. We all know engineers like food.

It’s a bit unusual for an engineering exec to lead office space design. Why did you get involved?

My take on this:  If you want a specific result you have to be involved.  Space design in important enough to both culture and developer productivity that I felt it was critical for me drive it versus someone may not bring the same perspective on our DevOps work style. Plus, I wanted to bring unique point of view to the project that combined function and style. I also saw it as a fun and challenging opportunity – how many engineers get to say they designed an office space?

What was your most important design philosophy for the site?

I wanted the site to bring people together – to not only foster collaboration, but to really connect us as people. There are so many benefits to connecting with people:  we build trust and understanding that enables us to interact at different levels. We collaborate better with people that we know and collaboration is essential at work.

As we were designing the space we focused on these goals of connecting and collaborating with our space planning team.  Selecting the space was the first hurdle - a space is either conducive to this or its not, so space and space planning are critical.  I know our efforts there are paying off because we are seeing the connections and collaboration happening in team rooms and common areas. In the common area people are eating lunch or enjoying a drink together where often the conversation evolves from work to social topics, so coworkers can relate as individuals. These are the types of things we wanted to happen, because ultimately it enables us to all work better together.

CenturyLink Cloud October Release: ‘Execute Package,’ SAML Options & a New UK Data Center

Our latest platform release went live earlier this week, release notes here.  As usual, we’re focused on increasing automation, security, and flexibility.  Here are a few of the highlights, along with some details that illustrate how users can take advantage of these new capabilities.

‘Execute Packages’ - Server Management using Automation

Cloud computing isn’t just about provisioning servers. You have to administer them too.  Ideally, you could administer these servers in bulk, and avoid the pain of one-offs.

CenturyLink Cloud has several options to help you manage servers in bulk (groups, scheduled tasks, alerts, power operations, and more).  But perhaps the most powerful solution available to users? The newly re-factored capability to execute Blueprint packages. With the latest enhancements to the Groups user experience, you can run a script package on any or all of the servers in a Group with just a few clicks.

CenturyLink Cloud includes several publicly available packages for customers to use. Need to install SQL Server? There’s a package for that. How about Active Directory? There’s a package for that too. Looking for the LAMP stack? Yep, there’s a package for that.  You get the idea.

But best of all, you can create your own custom package to provide consistent automation and repeatability for your most commonly performed tasks.  You can configure, update, and patch your servers in the most efficient way possible.

You can use a package to join an entire set of servers to a domain, deploy application code from a repository to all your web nodes, install a new performance monitoring agent to a bunch of servers, or add a software patch to a subset of machines. Anything that is scriptable from the Windows command line or Linux shell can be turned into a package and executed across a number of servers.

Packages also support ‘parameters’ to maximize reuse and utility. Maybe you have two slightly different configurations for your database installation script.  Include a parameter to indicate ‘master’ or ‘slave’ at run time. If you need to include the server’s IP address in a config file, pass that in as a parameter too. With parameters, the possibilities for what a package can do are endless!

Option to ‘Require SAML’ at Login

We’ve supported SAML in CenturyLink Cloud for nearly two years.  For our customers and partners, it’s a great way to extend existing IT systems and policies to the public cloud (and now the private cloud).

In this release, we add a key feature for administrators – the option to require SAML at login.

This capability closes the ‘back door’ login option at, and instead automatically re-directs the user to the SAML login page, as configured by the user’s administrator.

Administrators enable this feature with a single click (shown above), and can also easily apply to all (or opt-out) to sub-accounts.

Many customers have given us interesting ideas on how we can improve our SAML support, and this “option to require” is the next step for the platform.

As far as what lies ahead, identity management requires a bit of a balancing act.  We want the nuts-and-bolts of IdM configurations to live in the customers’ IdM system, while building in flexible, standards-based capabilities in our platform.

Our goal is to make it easier, and frictionless for customers to access our platform – while focusing on what we do best. Consequently, we have a proverbial ‘line in the sand’ for what we will do in the platform and what we won’t.

New UK data center – GB3

Santa Clara, Sterling, Toronto – and now Slough, UK.  These are the four new locations for CenturyLink Cloud that we’ve opened so far this year.  Each site offers the ability to deploy resources closer to users and employees, to maximize performance.

Another benefit? Hybrid IT becomes far simpler for customers already deployed in these facilities.  For the UK market specifically, our EMEA leader Fiona Cullen explains more in a recent blog post here.

These new features – and others – are available at your fingertips.  Take advantage of them today!

DevOps Gets Physical (Office Space)

There’s no shortage of DevOps experts these days.  These luminaries are everywhere - blogs, social media, trade shows, you name it.  Most agree that DevOps best practices fall into three categories: People, Process, and Tools.

You can read all about those in this space, and elsewhere. But once you look past the slideware, the conferences, and the consultants – one question lingers.

What does DevOps look like in a workplace?

How do the principles of collaboration, instant feedback, cross-functional teams, and shared objectives actually come together in a physical office?

As we designed our new CenturyLink Cloud Development Center in Bellevue, a few delegates from the CenturyLink Cloud team had the opportunity to create the perfect environment for DevOps. And, to do it from scratch, with few restrictions.  After all, this Center was one of the major components underpinning CenturyLink’s acquisition of Tier 3 nearly a year ago.

So what are some of the results?

Team rooms, organized by function.  The majority of our space is dedicated to “team rooms” – large, open rooms where employees are grouped by workstream (platform team, application services team, service engineering, etc.). As we build new features and push them to production, all the people you need to interact with are literally a few feet away. Some employees prefer to float from team room to team room, depending on what they are doing that day, or “just because.”

Desks suited to pairing.  Pairing is a powerful tool for distributing knowledge across an organization – and you need wide desks to sit shoulder-to-shoulder to maximize that concept.  Whether it’s a security expert, network architect, product manager, or tester, you need the physical space to work effectively and knowledge-share. And because our teams are using a single set of tools, each ‘pair’ can review dashboards and data in common repositories, to inform their decision-making.

Built-in collaboration spaces. Each team room is outfitted with gigantic whiteboards.  This generous allocation of space means people are more inclined to start drawing and illustrating an idea, problem, or scenario.  Near each whiteboard is a high-def TV connected to AirPlay.  This wireless setup is frictionless, and far faster than fumbling with adapters and cables. Each TV is near a couch, two chairs, and coffee table.  There’s even whiteboards in hallways – because sometimes hallway conversations are the most important ones.

A recurring theme: “having fun is just as important as working.”  It’s easy for employers to say that work should be fun.  But how do they show it?  At the Center, team rooms have embedded speakers in the ceiling, so everyone can listen to music during the day.  There’s even satellite TV in each team room, as well as in the largest shared spaces, too.  And of course, there’s a dedicated gaming room when it’s time for a much-needed break.

The best example of fun, though, comes from drudgery of on-call engineering that’s required in cloud.  Every engineer at CenturyLink Cloud pulls pager duty, a role that rotates weekly. How do you make this at least a little tolerable?  By having a cool name.  The on-call engineer that week gets to be “Batman.”

Spaces for “spikes.” Most R&D efforts by our team involve a deep-dive on a topic over the course of a day or two.  Those engineers working on spikes do so in a dedicated space where they can immerse themselves in the task at hand.

No phone calls allowed in team rooms. Need to attend a meeting, or take a call from a remote colleague?  Do it elsewhere.  In general, our team eschews phone calls, since they are 1x1 conversations that exclude others by definition.  And, phone calls can’t be audited the same way other channels can.

Meeting rooms of all shapes and sizes. This is the flip side of the ‘no phone call’ rule. The office has “fishbowl” rooms with space for one, as well as medium and large rooms to fit the need.  And there’s plenty of them, so there’s never a shortage of space.

Large Projection Space.  Adjacent to the large break area is a projection space, designed for technical ‘brown bags.’  We’ve made the space available to start-up and dev events as well, like the upcoming Seattle Cloud Foundry Meetup.

Open lunch room, complete with catered lunches. Our “no eating” in the team room rule means people actually take real breaks.  When dozens of engineers sit down together to break bread, lots of ideas surface.

Few cubes.Cubes and rapid development don’t go together.  The walls just seem get the way of the constant collaboration, analysis, and troubleshooting.

There are many, many other details that reflect the unique nature of this space, and the people who work there.  Each room in the office references our favorite movies (“Jack Rabbit Slim’s”, “Mos Eisley Cantina,” and my favorite, “Thunderdome”), but the highlights above show how working at web scale has influenced our workplace.

Safehaven Run Book Automation – A small change with a big impact

In an earlier post, we discussed SafeHaven for CenturyLink Cloud Disaster Recovery- as-a- Service (DRaaS) solution and the benefits it offers IT Administrators.  As we noted, failing over a multi-tiered application when executing a disaster recovery plan is critical. However, it’s not always as easy as it seems.  In order for a multi-tiered application to recover correctly, the VMs upon which it depends start up according to a prescribed “recovery plan.”  For instance, it is usually necessary for the database to be running before application servers boot and, similarly, necessary that application services be running before webservers boot. Safehaven for CenturyLink Cloud’s latest feature enhancement, Run Book Automation, allows end users to configure custom shut-down and bring-up plans for each group of IT systems that received disaster protection.  For example, delivering web services often involves a set of interdependent workloads that need to start in a specific order and taking into account time intervals between applications.

For each group of IT systems within the CenturyLink Cloud, users can pre-configure and test recovery plans in the Safehaven Console that identifies bring-up and shut-down order, actions, delays, as well as any custom script to be executed as part of the recovery operation.  Once the plan has been loaded, executing becomes as simple as a point-and click operation.

For each group of IT systems within the CenturyLink Cloud, users can pre-configure and test recovery plans in the Safehaven Console that identifies bring-up and shut-down order, actions, delays, as well as any custom script to be executed as part of the recovery operation.  Once the plan has been loaded, executing becomes as simple as a point-and click operation.

Conversely, when the original data center recovers after a disaster event, the challenge becomes how to efficiently restore protected IT systems so that they resume normal operations in the production environment.  With this “failback” challenge, SafeHaven also provides full Run Book Automation. Users use the same interface shown above to specify custom recovery plans for failback to the original production data center.  So why is this seemingly minor feature so important?  Well, because in business time is money.  And when your applications are down, you are losing money.  Aberdeen recently released a report that stated mid-size businesses experience on average 3.5 downtime events per year, lasting 3.4 hours and costing over $880,000.¹ This equates to about $1,233 for every minute your business applications are not running when they should.  Let’s say, for example, that through the nature of Safehaven’s Run Book Automation allowing multiple applications to be brought up at the same time and nearly eliminates the risk of human error during the process, you save 15 minutes of downtime a per hour.  Using the assumptions above, this means that implementing a disaster recovery (DR) plan with Safehaven for CenturyLink Cloud could save your business $18,500 per hour in downtime costs.  And, when you take into account the average number of events and downtime, your estimated savings in this example become $220,150.  That’s no small chunk of change.

Make sure your DR solution is intuitive enough to protect your business both technically and financially.  Learn more about Safehaven for CenturyLink Cloud here.