Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Does CloudFlare really speed up your site? Tests say not really.

blog.dh42.com

Post reply on HN
Print view View original

1–10 of 61 posts

Re: Does CloudFlare really speed up your site? Tests say not really.

#4

Post by dorianj »

The unlabeled referral link here at the end praising his host bothers me a bit -- shouting praises for a good service is fine, but at least note that you are getting a kick back.

> Because I know I will get an angry email or two, just let me say this. I know that Cloudflare touts other features such as security, load balancing, and keeping your site up if your server is offline. I did not test or take any of these features into account. Their biggest sector is to people that want to speed their website up, so I took them to the task on just that claim.

If this is true and people really think that CDN = faster site, then it's a misunderstanding, perhaps perpetuated by malicious marketing by the CDNs themselves.

But putting a proxy in front of an unloaded site (hosted on a relatively fast server) is of course unnecessary.

Re-run the test using `ab2' instead of simple 3-hit tests, and that's when the CDN becomes more useful. Or perhaps host the site on an oversold bargain shared server and see how it fares.

Re: Does CloudFlare really speed up your site? Tests say not really.

#5

Post by jws »

This is a small test[1] that shows the absence of miracles. My interpretation is:

• Location, location, location. – Sometimes cloudflare will have better positioning than your server. In this test the Amsterdam clients benefit significantly compared to his Atlanta Georgia US based server.

• It is not magic. – For the content he is using (base installs of popular packages and highly web optimized site) cloudflare's compression and optimizations are not helping much (at all?). My next path for further study would be to see if these base installs are highly optimized already. It seems reasonable that they would be, but it would need to be looked at. Then consider if you are planning to have unoptimized content (3rd party generated), or could benefit from skipping the optimization task and live with what cloudflare does.

• Do you only have one customer? – One of the attractive features for me is I can share my "worst case"[2] customer load with someone who won't even notice the blip. Tests during a deluge would be interesting.

• Does no one hate you? – Haters aren't necessarily sane[3]. Performance tests during a DDOS would be interesting.

I'm evaluating cloudflare for a site. I don't expect it to out optimize me. I do expect it to help European and Asian load times, and I have high hopes that it will kick in during bad times and make them less bad. To that end, does anyone know of a friendly DDOS service? I'd like to be able to schedule an X gbps DDOS for Y seconds with traffic of form Z for testing purposes.

␄

[1] 26 numbers, each made from three samples, five minutes apart, averaged together. No standard deviations.

[2] Which is the best case in the big picture. Just hard for the computers.

[3] I ended up retiring an IP address out of our C block because someone hated it. It was the "friends and family" email and hosting machine back before Facebook and free web mail services. Maybe someone got offended by something, and it got a persistent DDOS attack that would saturated our incoming IP links. Six months later when I tried to reuse the IP the attack came back immediately. I just marked it "unusable" in our DNS files.

Re: Does CloudFlare really speed up your site? Tests say not really.

#7

Post by jerf »

What's the point of this test? Yes, if your webserver's purpose in life is to serve the Wordpress default template 3 times total in its lifespan, neither Cloudflare nor any other such service will be useful. In fact, if anything, I'm a little unclear on what took .5 seconds at all.

I have neither connection to nor interest in Cloudflare, but I would intellectually be interested in a useful comparison test. This looks less like an interesting comparison test and more like somebody constructing a benchmark to say what they wanted it to say in advance.

Post reply on HN