Test #1  


(A) Test Profile  

- Always maintains 30 simultanteous users for 5 hrs:  28 Reporters (with 27 Reporters with no file attachment & 1 Reporter with a file attachment removed), 2 Reviewers
- Approximately 39 report submits/hr, with 1 file attachment upload per hr omitted, ~ 1.2 Full Searches/hr
- Arrival Rate of 1 new user every 20 secs.
- User 'think time' was increased via load tool global parameter change to slow the rate of submitted reports to 39/hr versus previous test run at 66/hr.
- Omitted: Size of uploaded file = 895 KB, filename = "ResearchDocument<1-9999>"
- Total users processed is greater than the number of simultanteous users active at any time during the test run.

 [note: since there was an issue with the load script with uploads, it was removed for this test run]


(B) Test Servers

- Web/App Srv (cf086.serverfarm.cornell.edu)
- DBSvrs (laurel1.cit.cornell.edu and laurel2.cit.cornell.edu)
- Fedora Server



(C) Comparison Criteria Results

Criteria

Sub Detail

Value

Comments

Avg Response Time (select pages of interest)

Reporter: (02) Initial Access

1.7 secs

 

 

Reporter: (16) Complete Entity Report

4.0 secs

 

 

Reporter: (22) Upload Attachments

N/A

No uploads in this run.

 

Reporter: (24) Submit Report

0.4 secs

 

 

Reviewer: (51) Initial Access

2.0 secs

 

 

Reviewer: (52) Submit Report Status

1.7 secs

 

Resp Time Requirement

95% of all Pages

< 10 secs

 

Page Validation

 

100 %

 

Reporter: Total # of Reporters (Processed to End)

 

N/A

 

Reporter: Num of Reports Submitted

 

196

 

Reporter: Num of Attachments Uploaded

 

N/A

 

Reviewer: Num of Report Review Attempts

 

89

 

Searches: Full

 

6

 


Note 1: Data is from Load Tool Report tabs 'Performance' and 'Application - Custom'.


(D) Comments:

- Passing test run, no significant issues seen in test results.
- There were 8 HTTP 500 error counts recorded but did not affect the run at 100% validation.  Suspect a HTTP page element request needs to be updated in the load script.
- Since there was an issue with the load script with pdf uploads, it was removed for this test run. Not considered serious omission due to very low number of uploads in production.


(E) Load Tool Report

http://testtoolsvm1.cit.cornell.edu/RPT_Runs/COI/03-28-13_5hr/Sched3_COI_Reporters_Reviewers_v2-Run30_5hr-NoUpoads_Mar-28-13_6-17PM.html


(F) Attachment - Profile data from COI production activity (background info).

 Load Stats - sorted by day.docx


 

  • No labels