Do You Need A Subscription For Olanzapine : Buy Cheap 100 Mg Doxycycline From The Arkansas (ar) : Amlodipine 30 Pills For Sale Cheap Paypal : Finasteride 180 Pills Drugstore Latvia : No Prescription Generic Lasix : Protonix Without Script : Buy Retin-a 20 20 Gm Montana Drugstore : Abilify Safer Alternatives : Pharmacy No Prescription Shipping From Europe Indocin 60 Pills : Canadian Online Pharmacy No Prescription 156.25 Mg Amoxicillin : Where To Buy Generic 40 Mg Strattera 90 Pills In Canada : 5 Mg Compazine 90 Pills Sales In Vermont : Anafranil Research : Buy Cefixime 10 100 Mg 64.07$ Online Canada : Prochlorperazine 90 Pills 5 Mg Other Day : Did The Price Of Augmentin 120 Pills 635 Mg 298.83$ Go Up? : How To Purchase Olmesartan 10 By Phone : How Do You Take Citalopram 60 Pills 40 Mg : Where To Buy 25 Mg Hydroxyzine Online Overnight : Sildenafil No Prescription

First Byte Response Benchmarking

I was helping @nicoalaryjr this week, trying to minimize his First Byte Response time for his WordPress site, sexlab.tv First Byte Response (FBR) is a metric used to describe the delay between when a users browser sends a request, and when the server fulfills the first byte of that request. It’s a great metric for most WordPress installs, as the first byte is typically indicative of completion of server-side processing. It’s a great way to test to see the performance impact of a particular plugin, as well as the overall responsiveness of your site. I used the open source httperf tool set to run the test I’m showing today, it can be installed on ubuntu by simply using this command:

sudo apt-get install httperf

Benchmark Comparing Hosting Provider Performance

@nicoalaryjr wanted to know what the performance benefits of moving from his current hosting provider to Dreamhost, the hosting provider I use for this blog. The results were partly expected, and partly surprising, with the one anomaly being that Dreamhost seems to have one connection out of a thousand that it just drops the ball on. The real performance metric to look at here is the median response time, which, with 50 requests per second (RS), Dreamhost was 30X faster than his existing hosting! Even at 10 RS, Dreamhost was 7.5X faster.

The really surprising results were the Dreamhost VPS tests. With the default RAM value of 2300MB, at 50 RS, it was nearly 50X SLOWER than shared hosting. Even at 10 RS, it was 14X slower. With the minimum RAM of 150MB, performance is completely unnacceptable. What’s going on here? I can only theorize, but I would like to know the true reason. If anyone has some insight here please comment and let me know. The only theory I can come up with is that CPU throttling is dynamic on the shared hosting, as opposed to being given a static slice of CPU with VPS, therefore, with the short duration of the tests, throttling did not kick in on the shared account.

Let me know what you think!
BTW favorite sexlab.tv video? The Spiderman, it’s hilarious!!

Here is the link to the actual test results: Hosting Provider First Byte Response Benchmarking Results

Reblog this post [with Zemanta]
  • Share/Bookmark
This entry was posted in IT. Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.
blog comments powered by Disqus