Hosting is Dead?

Okay so they have a good reason to say that. But it is still worth a read.

Some points I have experienced in the past

Server and Site Updates

"The person tasked with updating security patches had other priorities, and couldn’t stay on top of it. Or, the new developers you just brought in for a site
refresh are constantly requesting SSH keys and passwords. You’ve lost track of who had access to what in the past. Now a robot has hijacked or defaced
the site. "

I think this is also a matter of budget and expectations. The client never realized that their choice of Drupal or Wordpress or Joomla is a long term expense. Server updates, if they are not on a shared host, and Site Updates. By removing from the budget the need for spending time on server updates can be a huge help for small business owners, small tech shops and freelancers. Even if you like being a system admin I would still have to wonder is the expense worth it any more?

Workflow tools

Professional Drupal developers need to follow best-practices like version control, staging servers, deployment workflows, and
automated backups. Without the proper tooling in place, the development process takes more time and introduces significant risk. Even simple bug fixes
and updates are cumbersome and can take down the live site.

Another great point. To have good workflows is a costly investment for an individual or shop. But it is key to working on a project, as a team and even onboarding new people per job. Sure you can spend a ton of time setting up Vagrant, Chef scripts, the perfect server install and distro but can you really ever recover that cost?

Scaling on the fly

This is another one I have thought allot about. It seems that Heroko, Pantheon and the APIs Amazon offers are good example of systems I can scale up and down on the fly without a reboot. But others like Rackspace, Linode etc all seem to have downtime to scale the server? This seems so key to a successful site and budget.


Backups are another huge one. Providers will offer it but is it good enough? Daily? How far back? How quickly can your recover? And are those rdiff-backup or rsync script I made really still running? ( -:

Points of failure

Some of these providers have redundancy built in. That I have not researched enough to know for sure how good it is but this would be a killer feature to have that to build on my own. And again be even harder to recoup the cost from my average client. Even over time cause in time you keep making your system "better" or change them to the latest thing.


Finally costs. They have some charts in the doc about costs. Some of them are clearly one time costs that a shop could then roll out again and again for less hoping to recap the amount needed to pay for the work they did. But overall I would say that the cost to manage a server monthly are at least 2 hours a month (1 minimum). If I am billing a decent rate of $85 to $125 then already the client is paying to much for me to make sure the server (not even the site updates just server) is up to date and secure. Not to mention update but I am expected to be there to fix the server if anything happens 24/7. Honestly I think the client and I are both the losers when those expectations are on the table. As my wife when I got a phone call while getting dressed for my wedding! ( -:

Honestly I think the client and I are both the losers when those expectations are on the table.

Anyways, it is worth reading but more importantly the subject is worth go back and forth on in discussion #hostingisdead to write your thoughts on twitter or link to a post of your own.


this site is on A2hosting and I will soon move it to heroku to further test out these ideas.