E.g. You need to send 1000 requests in 10 concurrent threads to localhost:
ab -c 10 -n 1000 http://localhost/
For Debian users: you can find this mighty command in apache2-utils package.
Server Software: Apache-Coyote/1.1
Server Hostname: localhost
Server Port: 80
Document Path: /
Document Length: 405 bytes
Concurrency Level: 10
Time taken for tests: 0.524 seconds
Complete requests: 1000
Failed requests: 0
Write errors: 0
Total transferred: 650000 bytes
HTML transferred: 405000 bytes
Requests per second: 1907.13 [#/sec] (mean)
Time per request: 5.243 [ms] (mean)
Time per request: 0.524 [ms] (mean, across all concurrent requests)
Transfer rate: 1210.58 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 0 2 0.5 2 4
Processing: 0 3 0.5 3 4
Waiting: 0 2 0.7 2 4
Total: 2 5 0.3 5 8
This time I focused on jQuery.Together with Michal Kaukič we started jQuery workshop. Attendees learned basics of jQuery in two hours. At the end they have running application with simple animation in HTML.
I’m building several applications based on newer Flex/Spark architecture. It is good practice to keep application at minimal size. With Spark architecture I hit several problems.
One critical problem is that application with Spark is quite big. I decided to enable support for RSL…
There is problem that RSL is available only for official releases of Flex SDK. You can run aplication on developer’s machine, but user will see some fancy errors.
There is another problem with Adobe’s RSL. Browser will contact Adobe server before application start. You could see 3-5 connections. Flash Player is checking version of RSL. This will slow down start of application. It could take 10 seconds in network with higher latency. That is very bad.
Therefore only viable solution is to turn off RSL and use “Merge into code”.
Folks who advocate for Spark says that it is much more efficient and it has better structure. Sure, it’s true. You can do many things with Spark much easier than with MX.
I was wondering why were my applications with Spark that big. I made small reasearch. You can see comparison between MX, Spark and MX+Spark projects on followich chart. In the first step I was testing only size of application with just one Button in Release mode. Then I’ve added some other components and application starts growing.