Account Options

  1. Sign in
The old Google Groups will be going away soon.
Switch to the new Google Groups.
Google Groups Home
« Groups Home
Message from discussion Next Meeting - Monday, Sept. 21, 6:30 p.m.
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Tim Morgan  
View profile  
 More options Sep 22 2009, 1:40 pm
From: Tim Morgan <seve...@gmail.com>
Date: Tue, 22 Sep 2009 10:40:16 -0700 (PDT)
Local: Tues, Sep 22 2009 1:40 pm
Subject: Re: Next Meeting - Monday, Sept. 21, 6:30 p.m.
Thanks to everyone who made it to the meeting last night, braving the
sideways rain and zero visibility.

Here is my presentation, if you can find any use for its content at a
later date: http://tinyurl.com/l3g63e

For yous who aren't git ninjas, you can easily view on GitHub a commit
listed in the presentation by copying the commit sha1 hash (that
freakishly long set of letters and numbers) from the slides and paste
into a url like this: http://github.com/seven1m/onebody/commit/c1b2c9279495ab8190ee1744f701...

The memory and speed test results in the last few slides were running
my app on Webrick, which I would never do in production (except that
one time I did it in production). Also, that was running in
development mode -- I'm not sure how that would affect the numbers.

So, unhappy with those tests, I thought I would do some more testing
(totally unscientific, remember) comparing Phusion Passenger on Ruby
Enterprise Edition vs. Ruby 1.9 running my app in production mode.
Here are the results...

2 instances on Ruby Enterprise Edition (PassengerMaxPoolSize 2)

 4.2% 87340 ruby            Rails: /home/tim/p/
onebody
 4.1% 85036 ruby            Rails: /home/tim/p/onebody

2 instances on Ruby 1.9 (PassengerMaxPoolSize 2)

 4.2% 88328 ruby1.9         Rails: /home/tim/p/
onebody
 4.2% 88244 ruby1.9         Rails: /home/tim/p/onebody

Yes, that is saying what you think it's saying. Ruby 1.9 is using
slightly *more* memory than Ruby Enterprise Edition. I did this three
times to make sure I wasn't missing something, and these numbers are
typical. REE tends to beat 1.9 by MB or two. I don't understand why.

Now, let's test speed...

Ruby Enterprise Edition:

  Processing PeopleController#show (for 127.0.0.1 at 2009-09-22
12:31:56) [GET]
  Completed in 380ms (View: 317, DB: 5) | 200 OK [http://onebody.local/
people/27942]

  Processing StreamsController#show (for 127.0.0.1 at 2009-09-22
12:31:54) [GET]
  Completed in 466ms (View: 435, DB: 3) | 200 OK [http://onebody.local/
stream]

Ruby 1.9:

  Processing PeopleController#show (for 127.0.0.1 at 2009-09-22
12:29:48) [GET]
  Completed in 176ms (View: 97, DB: 5) | 200 OK [http://onebody.local/
people/27942]

  Processing StreamsController#show (for 127.0.0.1 at 2009-09-22
12:29:45) [GET]
  Completed in 156ms (View: 128, DB: 2) | 200 OK [http://onebody.local/
stream]

Now *that's* significant. I ran these tests 3 times as well, and these
numbers are typical. 1.9 cut the response time in half or better.

So, there you go.


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.