Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Migrated to Confluence 5.3

General Information

  • Start Time: 10/23/2007 4:26:35 PM
  • Users: 100
  • Test Duration: 01:16:05

App server config

  • maxThreads=250
  • minSpareThreads=25
  • maxSpareThreads=75

Database Notes

  • Load Average: 0.07 to 6.21
  • Sar data for time in question(recorded on bricker)
    Code Block
    
    

...

  • 04:50:04 PM   runq-sz  plist-sz   ldavg-1   ldavg-5  ldavg-15
    04:20:01 PM         0       386      0.23      0.54      1.17
    04:30:02 PM         5       388      1.73      0.80      0.89
    04:40:01 PM         0       392      2.02      2.60      2.01
    04:50:04 PM         2       491      0.76      1.22      1.62
    05:00:01 PM         1       490      0.80      0.90      1.29
    05:10:01 PM         0       487      0.58      0.90      1.12
    05:20:01 PM         0       489      0.51      0.74      0.93
    05:30:02 PM         2       489      0.16      0.31      0.60
    05:40:01 PM         0       488      0.17      0.36      0.51
    05:50:01 PM         0       487      0.00      0.07      0.27
    

Silk Data

Note: This table is the concatenation of two tables from silk performer.  Therefore blank columns only mean the column didn't exist in the original table.

Stat

Count

Per Second

Per Minute

Per Hour

Average

Min

Max

Count

StdDev

Transactions

429 844

0.10 19

5 11.72 25

343 675

 

 

 

 

 

Requests sent

52 116,948 565

11 25.77 90

705 1,554.97 20

42 93,358 252

 

 

 

 

 

Overall Response Time - Page Time[s]

 

 

 

 

14 20.907 296

0.020 018

588 670.675 437

7 17,924 601

56 64.253 826

Conclusions

We conclude that load test and system configuration simulated the poor performance we saw in production with the views. TBD