Docker – What Problem Does It Really Solve

 

by Lukasz Guminski

I like Docker. It’s easy to use and fast. Simply brilliant. So brilliant that every day we find new possible applications. But what actually is the real problem it solves? Stable environments? Configuration management? Effective virtualization?

I think it is something else.

In the past I used to work for a company providing (and buying) business-to-business (B2B) services. Looking from that perspective, it was always surprising to me that B2B services did not move far beyond simple content syndication in times when technology changed so much: ultra-fast broadband networks, expanding clouds, distributed databases, etc.

There are some exceptions: a few B2B services mostly in areas of advertising (Google AdWords, Facebook Ads), analytics (Google Analytics), and streaming (Netflix). The vast majority of the B2B market, however, still revolves around slow, asynchronous data transfers.

Why is it so? I think that businesses expect a much higher quality of service than end users do. So it is harder for the market of B2B services to grow. Only a handful of companies were actually able to win customers’ trust. I’ve just mentioned a few, namely Google, Facebook and Netflix.

Yet, if you were a customer, there is somebody else you would trust – yourself. So what if you could get the service you need delivered in a black box, which you could deploy wherever you want – to your data center or a cloud? You decide on network latency and how much resources (computing power, memory etc.) it is given. Even if you don’t know how the service actually works, you already trust it more, don’t you?

Docker is a technology, which perfectly solves the problem about how to package services. Using Docker, business customers could deploy their services very closely to the applications that need them. Same model as in high frequency trading, where network latency has been brought to minimum.

Technologically-wise, the time is just right. The family of aaS (IaaS, PaaS, SaaS) solutions is mature enough; Docker enables the next step: Service as a Product. But we are still missing a few things.

  • Marketplace – where buyers can find containers with services they need.
  • Property protection mechanisms:
    • Licensing infrastructure – allowing providers to benefit when the demand for their services in customer’s cloud grows (federated license servers?).
    • Intellectual property protection (encrypted containers?).
  • Container orchestration – to manage complex multi-container services.

However, I am optimistic. The B2B space, I think, is going to change, as Docker opens a new market of standard, easily deployable B2B services. And I hope that soon we will see that happening.

The following two tabs change content below.

Lukasz Guminski

Lukasz specializes in system architectures. At Container Solutions his role is to deepen the understanding of technological trends and making forecasts on what is to come.

6 Comments

    • It is surely one of most advanced solutions. But I think it is too early to say. Docker itself is very fresh.

      For sure the need for orchestration is there. So the market will respond somehow.

  1. Lukasz, thank you for the article. I like the idea of services as a product and agree on your wishlist to make this happen.

    Deciding on the environment of the black box might increase trust, but I’m not sure if that solves much of the lack of trust issue in B2B markets. If I was managing risk from software products in a security conscious sector, I wouldn’t be comfortable of running black boxes without thoroughly inspecting them in a safe environment before deploying them.

    • I wouldn’t be comfortable of running black boxes without thoroughly inspecting them in a safe environment

      This is exactly the point. With containers you are getting an option to inspect a service in an environment, which is fully controlled by you. You can easily check opened connections, data transferred etc. I believe it makes a difference.

      And if not you personally, then there is a chance that somebody else would do that. Same way as it happens currently with mobile applications that are reverse engineered and analysed (example).

      Surely it won’t solve all problems, but it will help. I think Docker creates a new channel of software distribution. And its potential is really big in my opinion.

    • You can argue about that. Many TV stations use its distribution platform and clearly you can talk here about B2B model. The market is growing (Netflix CEO: Broadcast TV Will Die Within 16 Years), and already some shows are available exclusively on Netflix.

      However if you are an independent publisher, it is hard to use their distribution platform.

      http://mediakitchenfilms.blogspot.nl/2012/09/how-i-got-my-movie-in-itunes-hulu.html :
      Netflix was the first one I wanted to land in. I couldn’t find anything online that was useful, everything was a dead end. I even called the Netflix office and asked around a little. No one was helpful. It was obvious that their acquisitions office was not open to the public. So basically a dead end.

      So as I said, you can argue, but I think this is not a B2B service in the sense I had originally in mind. I was thinking of services for small companies as well. Thanks for pointing this out.

Leave a Reply

Your email address will not be published. Required fields are marked *