Skip to content

CMS performance testing. Part III (Kentico)

February 24, 2011

In this post I’m continue to describe performance of Kentico CMS.

CMS performance testing. Part I

CMS performance testing. Part II (DotNetNuke)

1.1. Blog page

1.1.1. Summary

Load behavior: Start 100 user simultaneously

Maximum running virtual users: 100

1.1.2. Average values

Blog page, 1 Vueser Blog page, 100 Vusers
Response Time 0.5 4.7
Processor Queue Length, MS IIS 0.1 6 Not more than 2 for each processor unit
Processor Time (%),  MS ISS 8 74.8 Not more than 85%
Processor Queue Length, SQL Server 0.3 3.2 Not more than 2 for each processor unit
Processor Time (%), SQL Server 20.1 56 Not more than 85%
Disk Time (%), 

SQL Server

2.6 4.2
Disk Queue Length, SQL Server 0 0.03

1.1.3. Counters comparison graph

1.1.4. Results

In this case we can see that response time is good and the load for MS IIS  is not so high but also can see increased values for SQL server (to compare with Sitefinity).

1.2. Blog post comment (ramp-up)

1.2.1. Summary

Load behavior: Start 25 every 1 min 30 sec

Maximum running virtual users: 100

1.2.2. Counters comparison graph

1.2.3. Results

After this test we figured out the maximum concurrent user amount system can handle. It is not more than 90.

1.3. Blog post comment (flat)

1.3.1. Summary

Load behavior: Start 100 user simultaneously

Maximum running virtual users: 100

1.3.2. Average values

Blog post comments, 1 Vuser Blog post comments, 100 Vusers
Response Time 2.2 Users were dropped by timeout ( > 120 sec )
Processor Queue Length, MS IIS 0.7 Not more than 2 for each processor unit
Processor Time (%),  MS ISS 35.6 Not more than 85%
Processor Queue Length, SQL Server 0.5 Not more than 2 for each processor unit
Processor Time (%), SQL Server 24.3 Not more than 85%
Disk Time (%), 

SQL Server

4.4
Disk Queue Length, SQL Server 0.01

1.3.3. Counters comparison graph

1.3.4. Results

Here one can see actually the same situation as in previous case for Sitefinity CMS. The Kentico CMS is not able to handle such amount of vusers and dropped them.

1.4. Data base trace

1.4.1. Blog page (1 user)

There were 62 queries and 11 inline ones.

SQL Server execution time:  ≈ 2.6 sec!

1.4.2. Blog post comment (1 user)

According to trace table there were 57 queries. Also there 13 inline queries!

SQL Server execution time:  ≈ 1.3 sec!

Advertisements
4 Comments
  1. Dzmitry Kashlach permalink

    Thank you for article! Did you tested this CMS with test-lab or did you used cloud services, such as blazemeter.com?

    • We tested it in in-house test-lab. That time using clouds was not so popular in company I worked for

  2. Omjsr permalink

    what tool you have used for this performance evaluation?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: