• Keynote KITE 3.0 adds Flash Performance Monitoring Capabilities and Third Party Content Filtering

Keynote KITE 3.0 adds Flash Performance Monitoring Capabilities and Third Party Content Filtering

Keynote Systems Inc, recently announced the release of their latest Keynote Internet Testing Environment with version 3.0. There are many new features in the latest KITE release including the ability to separate third party content using virtual pages, the ability to monitor the performance of interactive Flash and Flex applications, and page snapshot which includes a thumbnail-preview of the page being tested in the recording script.

Virtual Pages

Virtual Pages is a new feature in KITE 3.0 that allows for separating out content based on a string in the URL (i.e. domain), content type, element size and/or element error.  The main purpose for Virtual Pages is to isolate third party content performance when measuring or monitoring a Web site. Third party content often skews performance results while doing Web site performance analysis, and by separating this content out, it gives a better perspective on how the core Web site is performing in comparison to third party ad serving, tracking, and video streaming objects that may be affecting page performance overall; Virtual Pages is an effective way to do that.

To create a Virtual, Page, begin by recording a standard KITE script. In this script, I have recorded a single-URL script and inserted a virtual action to separate third party content using a pattern match.

Creating a Virtual Page in KITE 3

Step 1) Record a Script and save it
Step 2) Right click the step in the script where a virtual page should be inserted and choose “Insert Virtual Action” (“action” is often synonymous with “page” in KITE).

Step 2)

Figure 1. Insert Virtual Action

Step 3) Click on the “Match Page Elements” label and set the “Filters Logic” property to “AND”

Setting Filter Logic

Figure 2. Setting Filter Logic

Step 4) Right click the “Match Page Elements” label and click on the “Add URL Match” menu option

Figure 3. Adding an URL Match for the Virtual Page Filter

Figure 3. Adding an URL Match for the Virtual Page Filter

Step 5) In the Script Properties Editor, click the “Invert Filter” checkbox to match everything other than the main Web site domain

Figure 4. Inverting the Filter in the Script Properties Editor

Figure 4. Inverting the Filter in the Script Properties Editor

Step 6) In the Script Properties Editor, enter the regular expression for the domain or domains of the main Web site, this will allow grouping of everything other than the main Web site domain(s) into this separate Virtual Page.
[ Figure 4.]

Step 7) Save the script for playback

Now we have modified the script to support Virtual Pages to filter out the main web site’s content and thus report only on third party content. In the next step, we will see the filtering take place upon playback of a KITE script.

Viewing  Third Party Content with Virtual Pages

Step 1) Click the “Play” button to play back the script with the Virtual Page in it sorting out third party content.  The content that has been filtered out of the Virtual Page (the main web site’s content in this case) now appears with a grey background so we can confirm that it was filtered out.  These grey rows do not impact any statistics and do not appear at all in test results shown in the MyKeynote portal.

Step 2) We can get a better view of the data if we use KITE’s multiple level grouping feature.  Right Click the Transaction Performance details menu and click the “Toggle Group-by Area” menu option, a new Grouping menu should now appear under Transaction Performance Details

Figure 5. Toggling the Group-by Area

Figure 5. Toggling the Group-by Area

Step 3) Click and Drag a property for Grouping to the Group-by area [ Figure 6.1]. In this case, Action and Domain have been grouped together.

Figure 6. Grouping by Action and Domain

Figure 6. Grouping by Action and Domain

Figure 6.1 The Group-By Area

Figure 6.1 The Group-By Area

Step 4) Now even though there is a single page load, KITE creates a second, Virtual Page using the filter we have defined. This Virtual Page contains only the  download sizes, network times, and other information specifically related to third party content.

Figure 7. Viewing Virtual Page Results

Figure 7. Viewing Virtual Page Results

Download the KITE script used in this example here: bitscript.krs

Flash Web Site Performance Monitoring

KITE is unique in that it allows for the recording of interactions with Adobe Flash and Flex based Web sites. Most performance testing tools only download a Flash object, but the KITE 3.0 utility is unique in that it can monitor the performance of full-fledged Flash applications and Flash-based Web sites.

One example is the Mini.com Web site that allows someone interested in buying a Mini Cooper to build their own custom configured vehicle using their Web-based Flash application.

Figure 8. Recording interactions with a Flash Site

Figure 8. Recording interactions with a Flash Site

As the user goes through the workflow of building and customizing a vehicle in the Flash-based vehicle builder Web application, the different Flash interactions are recorded in KITE step-by-step.

Figure 9. Detail of Recording a Flash Site

Figure 9. Detail of Recording a Flash Site in KITE

The steps to record a Flash script are as follows:

Step 1) Open KITE and Click the “Record” Button

Step 2) After the KITE Browser Launches, Click the “Dynamic New Page” dropdown and select “No”. This will ensure that network interactions mid-page do not get separated into separate pages.  Turning off the prompting will also prevent any conflict as KITE captures the screen coordinates of your interaction.

Figure 10. Turn off Prompt

Figure 10. Turn off Prompt

Step 3) Click on the workflow of the Flash-based site to record a user’s typical interaction relative to your Flash or Flex application, such as building a vehicle, or filing a support ticket in a Flash-based help desk system.  KITE will capture most steps automatically, with the exception of a “hover” event which requires the pressing of F9 to capture.  You’ll want to use the F9 to record hover events if you have cascading Flash menus, for example.

Step 4) After recording the entire interaction, follow the KITE help file’s instructions to add timings between pages or between steps within a page as appropriate.  These timers are needed so KITE does not interact with the Flash prior to Flash initialization or animations.

When you play back the Flash script, the KITE Events tab and Performance tab show the performance of each interaction the Flash makes with the network.  It is these “mid Flash” network interactions that are missed by earlier technologies.

Page Snapshot

The page snapshot feature is unique in that it allows for viewing of different points in time of the recorded testing script.  Once a script has been recorded and played back, the page snapshot feature will show in the bottom left hand corner (by default) of the KITE Console.  You can also drag the Page Snapshot tab to the larger area on the right if you prefer a larger image.  As you iterate through the various steps of the recorded script, a thumbnail-sized image will show in the Page Snapshot tab of the KITE Console allowing the user to quickly locate steps in the performance testing script visually allowing for the quick identification of pages, as well as page rendering problems during the recording of the script.

Figure 11. Page Snapshot Feature

Figure 11. Page Snapshot Feature

Learn More

Whats New in KITE 3.0

Testing and Measuring Web Application Performance with Keynote KITE

VIDEO: How to Add Virtual Pages to a KITE Script

Page Snapshot Feature

Share and Enjoy:
  • del.icio.us
  • Facebook
  • Google Bookmarks
  • Current
  • DZone
  • E-mail this story to a friend!
  • FriendFeed
  • HackerNews
  • Identi.ca
  • LinkedIn
  • Reddit
  • Slashdot
  • StumbleUpon
  • Suggest to Techmeme via Twitter
  • Technorati
  • Twitter

This website uses IntenseDebate comments, but they are not currently loaded because either your browser doesn't support JavaScript, or they didn't load fast enough.

Leave a Reply

You must be logged in to post a comment.