How to do Managed Hosting – Without Managed Hosting

By Simon Varney on 9 Jun 2014

Multiethnic People with Startup Business Talking in a Cafe

Leaving your website unmanaged is like leaving a child unsupervised.

Don’t believe me?

Think about it – imagine that you left a child unsupervised for an evening. You wouldn’t stop worrying!

“Is she safe?” “Has he gotten into trouble?” “Is she playing with something she shouldn’t be playing with?” – or worse: “is he eating something he shouldn’t be eating?”

For businesses, a live website is an absolute must and without the peace of mind that comes with knowing it’s being taken care of, it can be as worrisome as leaving a child unattended.

Still unconvinced? Let me tell you the story about Octopus Incorporated…

Welcome to Octopus Incorporated.

Octopus Inc. is the app that helps you juggle eight (or more) tasks at once.

It launched in 2012 with the mission to help people manage their time more effectively, and get more done.

After receiving glowing feedback from a prominent app review site, the user base surged to over 250,000 users. This, understandably, attracted the attention of quite a few VCs, and after a one million pound investment, it all seemed to be smooth sailing.

The founders, Jack and Jasmine, are developers so they know what they’re doing, and decided to go with Amazon Web Services for their hosting to save some money and keep full control of their site. They’re confident that they can stay on top of their scaling, keep their site live, and (most importantly) keep their now 600,000 users happy!

Providing a reliable user experience

Jack and Jasmine are so caught up in delivering frequent updates to their app, that when uptime percentages start decreasing, they barely notice. After a stern talking to from their VCs who expect a more reliable user experience from their investment, Jack and Jasmine decide to offer their developers sys admin training. They’re thrilled when not one – but all nine of their developers volunteer. Problem solved!

Although their developers will have to take time away from developing to dedicate to training, Jack and Jasmine think it’s the easiest and most cost effective way to resolve their downtime issues.

What happened next: A week after the sys admin training, the app goes down and (almost smugly) Jack and Jasmine turn to their well-trained developers. It’s 2pm on a Friday, and they’re confident their trained devs will resolve the issues within minutes. The only problem is that the devs are brand new to this whole sys admin thing, and end up taking several hours to get the app back up and running.

At 7pm, burnt out and grumbling about having to work late, the devs head home.

Hiring an in-house systems administrator

After a few more occasions of app downtime (and another stern talking to from their VCs), Jack and Jasmine decide to hire a dedicated systems administrator.

They certainly have their work cut out for them seeing as neither of them are systems or network administrators, but they take it in their stride.

After two weeks of advertising the position and scouring through CVs, they finally find Alan – someone who seems to have the right credentials. The whole process has taken their focus off of their core business aims temporarily, but they figure it’s worth it because they’re pretty sure they’ve found the right person for the job.

What happened next: Alan starts his job on a high. Downtime is lower than ever, and he gets along great with the team. He correctly sets up monitoring for the website and all aspects of their servers, and things are going well!

A month after he joins the team, a stream of angry tweets wakes up Jasmine at 12am. Octopus Inc. has gone down and their American clients are fuming! All of their users’ well organized to do lists are gone and the app keeps crashing.

Jasmine calls Alan but can’t get through to him. Turns out his mobile had run out of battery and he hadn’t received the alerts. To ensure 24/7 uptime in future, she and Jack make the decision to hire another systems administrator to act as a backup. They put the bulk of their responsibilities on hold again for an additional two weeks to find the right person which is a setback, but the good news is that the website can now be monitored all day, every day.

Limited experience leads to more downtime

Three months pass with Alan and the new sys admin, Luna, taking care of infrastructure for Octopus Inc. They’re each getting paid £50,000 a year, which is a hefty investment, but it seems to be paying off as downtime has fallen dramatically – good going, Jack and Jasmine!

What happened next: After three months of no downtime, another problem arises with the Octopus Inc app. Performance has dropped dramatically, and Alan and Luna are struggling to work out why. They are pretty sure it’s a problem with the website’s code, but the developers are adamant it’s a systems problem.

Eventually they solve the problem, a complex one they’ve not encountered before. If this particular problem crops up again, they’ll be able to solve it, but who knows how many other problems may crop up that are new.

They keep having to learn on the job, and learning on the job leads to mistakes.

After a week of glitches, performance is back under control, but not before one of the VCs has pulled their money – nobody wants to be associated with an unreliable service. Jack and Jasmine begin to wonder if this story is one many startups face.

A different way to manage your infrastructure

It’s certainly possible to manage your infrastructure without managed hosting, but the time, resources, and money spent can add up.

Why leave your child in the hands of someone that you’re not 100% sure of to do the best job possible? Of course it’s better than leaving your child completely unsupervised – but only just.

Why not trust your infrastructure with a team of experts who can all benefit from and build on each other’s experience? Why not trust your infrastructure with a team who you know will be responsible for its uptime 24/7/365?

Not only that, but the cost can be significantly lower than hiring in-house systems administrators, or having your developers spend time away from concentrating on your app or website at a critical time when your business needs to move as quickly as possible.

If it were up to me, I’d want to leave my child in the hands of Mary Poppins – and that’s exactly what managed hosting is for your website’s infrastructure.