Everyone is talking about Docker, and for good reason. The container technology "makes it possible to get far more apps running on the same old servers, and it also makes it very easy to package and ship programs," as ZDNet's Steven J. Vaughan-Nichols points out. That's a big deal, and it has driven constant chatter about Docker on Twitter and in the press.
But is that chatter translating into real deployments? Just how hot is Docker relative to other hot technologies?
First, let's take a look at the hotness.
How hot is hot?
In early June, there were 14,000 Dockerized applications available in the Docker repository. Today, that number has surpassed 35,000. In that same time period, Docker downloads have exploded from 3 million to 21 million. More than 13,000 Docker-related projects have mushroomed on GitHub.
More importantly, Docker is being pushed into production at companies like Gilt, Groupon, eBay, and Spotify, but also at large financial institutions and other "old school" enterprises. Meanwhile, software vendors as diverse as Google, Red Hat, and Microsoft have all announced Docker initiatives.
Docker's timing is impeccable, as I've written before. While containers are nothing new, Docker hits the market at a time when Linux is soaring, application development has become distributed and complex and developers are looking for an easier, lighter way to virtualize their infrastructure.
But how does Docker compare to other application development technologies?
What would your startup do?
Using Leo Polovets' AngelList data to track the technologies that okay/good/great startups use, Docker seems to be doing just OK (Figure A):
Leo Polovets' AngelList data.
However, as Polovets notes, the sample size for these DevOps technologies was small and hence perhaps not meaningful.
Broadening the data sample a bit to Ryan Williams' Hacker News hiring trends, Docker is listed as one of the fastest-rising job skills and is surpassed only by Chef in terms of total number of comments (Figure B):
Ryan Williams' Hacker News hiring trends.
Still, we don't want to just compare Docker to DevOps analogs — let's also look at the wider technology market. While an imperfect measure, Dice recently announced the top-10 fastest growing technology skills, including Puppet, NoSQL, Hadoop, and Python.
How does Docker compare to these?
Battle of the hotness
Using Indeed.com job data on absolute number of jobs requiring a given technology, it's clear that the mainstream enterprise simply hasn't started using Docker in earnest quite yet (Figure C):
Indeed.com job data on absolute number of jobs.
Note that I dropped Python, because it skewed the results too much (its orders of magnitude are more popular as a job skill than any of the other technologies.) Also, I used MongoDB (disclosure: I work for MongoDB) and Apache Cassandra as proxies for NoSQL, because they're the two most popular NoSQL databases according to the DB-Engines popularity ranking.
Hadoop, despite (or perhaps because of) the hype, leads in terms of absolute jobs, while Docker remains relatively small. (Docker wasn't released as open source until 2013, so the earlier jobs aren't related to the technology and must instead be related to dock workers.)
Even if we look at relative job growth, Hadoop dominates (Figure D):
Relative job growth.
Nor does this change much if we remove Hadoop.
Embrace the hockey stick
This is not, of course, to suggest that Docker isn't hot. To paraphrase William Gibson, "the future of Docker is here... it's just not evenly distributed yet." Early adopter, web-scale companies like eBay have already embraced it, and with 85% of enterprises wanting to build and deploy applications more quickly, according to a Progress Software survey, Docker's future is very bright.
A better analysis would have been to compare Docker against technologies of a similar vintage (i.e., released as open source in 2013). I chose not to, however, because Docker's imprint on the industry already promises to make it more than the equal of established technologies.
In other words, it's really, really hot.
Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.