Last week, I introduced readers to Cloudflare.com, a free service that promises to speed up load time and improve security/risk of service and data loss. Reader Chris Hutchinson of Curb-Tec Europe asked his IT consultant Jan Bartnik (email@example.com) for an assessment.
Jan responded in an email to Chris, forwarded to me:
As far as I can read from it’s spec even the Enterprise option is far from what I expected it to be based on the name. It’s a combination of online support, data filtering (more on that later) and static caching. The static caching is a nice feature but I’m not sure if it’s worth trying (see data filtering) if you have a host provider with 99.9 per cent uptime (and you’re not Facebook) you shouldn’t really care. Last year home.pl got 100 per cent uptime. Always keep in mind caching is only for static content – no forms, no reports, no emails, no dynamic data – just a snapshot of what the page looked like at some point in time.
And now the bast part – data filtering. It sounds like a nice feature. They will monitor the activity that’s between the website and a customer/hacker. How? By analyzing data traffic… the next step is data mining – the same thing that google does. Now you share all your company’s data with a third party – might be your emails, customers database, passwords. Look at https://www.cloudflare.com/security-policy.html and think if it’s worth to share your data with others. Also – think of a leak. That’s a huge database with sensitive informations… perfect for hackers.
Now – my opinion. If you’re a small company with small internet traffic – you shouldn’t really care about all those things they offer. Your site may be down for 6 hours total in a year and nobody will notice. If you’re a big company… you probably have IT guys that will do the same with amazon cloud computing servers
Before posting this email exchange I asked Chris for permission to repost these observations. He responded with this request:
Could I ask though if you do use it could you mention him in a way that could possibly link others to him? I’m asking because he is a young assistant professor at our Polish equivalent of MIT who I sort of pushed into starting a business and would now like to help him if I can. He started doing our IT work for us a few years ago while he was still in school and I am constantly talking with him about the disconnect between the typical IT person and a business leader, and how I wish that IT students were also taught more about business so when they advise companies, they have a better understanding our priorities. This caused him to start his own business and take it into consideration in the classes he teaches. He has 2-3 other customers now but basically I am his main client. So, I like to try to help him out when I can. I remember when we first met he was the stereotypical “techie” and could barely understand how I didn’t care about some of the technical things, only what it did to help my company. He’s come a long way for sure.
Based on these observations, I’ve decided to end the Cloudflare.com experiment. While using Cloudflare I noticed no improvements (nor harm) to my service or blog, other than repeated ISP warning about potential problems with my site’s Domain Name Server (DNS) — the key change you must make to implement Cloudflare. Fortunately, I simply needed to reset the DNS to previous settings to end the experiment.
However, this test also shows the amazing power of the Internet to help develop relationships and opportunities far from home base. Accordingly, I’m happy to pass on the recommendation for Jan Barnik.