Session length data

[Guest]

What determines the "session length" data? Is the time span start and end with the .createMonitor command?

[Guest]

Hi James The Session Length data is obtained from the data reported from the individual monitors. They report their session length starting from the Start()-call and continues until the Stop()-call. If the application is closed (or crashes) without calling the Stop method, we have a standard session length accuracy of 1 minute (since we automatically save the ongoing session length every minute to disk). Hope this helps. If you have any issues with this or need further information I'll be happy to answer any questions you may have Best regards Soren EQATEC Team

[Guest]

Thanks soren. In that case we have a bug of some sorts if I understand the system correctly. We have an unusually high incidence of session lengths <0.5 minutes. To check this I am filtering data to a version only I run. I ran several sessions yesterday and found the session length data didn't reflect my usage. Today i read the stats between 22 & 25th Jan. 9 uses ( 2<0.5, 3<1, 2<5 & 2<10min) I started the program today and left it running, then activiated the "live" stats. It detected the new session but was recorded as a session less than 30 seconds. I have yet to shut the program down and call stop. Any suggestions?

[Guest]

Hi S?ren, waite with answering this. I use some bridging s/w (unprotected host to protected .Net modules). The problem may lie within this DLL. I'll check this first and let you know. regards Jim

[Guest]

Hi James Just a quick point to make. The monitor tracks the session length from Start() to Stop() locally on the client machine. This data is persisted every minute to the local disk but the data is pushed in batches to the servers only "when needed". This means that the monitor pushes data to the server (asynchronously) immediately after Start() and when you call Stop() and in between these data is only pushed to the server when our internal logic triggers such a sending. So dependent upon how you have instrumented your application you will not see the session length increase for a specific session before you call Stop. Perhaps this explains at least the effect you are seeing while debugging. For further troubleshooting and determining what and when data is pushed to the server, we recommend installing Fiddler and look for http traffic going to *.monitor-eqatec.com (see http://support.eqatec.com/entries/20953498-troubleshooting for more) Let me know if I can do some additional debugging in our end. Best regards Soren EQATEC Team

[Guest]

Thanks S?ren, had a notion the <0.5 was incremented until the program was closed after running 2 hours of test. Today the stats in Aqatec match exactly what I performed but have queries over yesterdays. Will keep testing and see if anything abnormal appears. Thanks for your help. Have another query about the FeatureValue function. It works for some categories but not for others. In Aqatec analitics there is no data registered in several categories but is data is sent just as those which have statistics. It displays that no intervals are defined but I can't define them due to lack of test data. Is data necessary to set up the intervals. Regards Jim

[Guest]

Hi James (or Jim?) If you are doing very specific testing and want to see data as quickly as possible, I would recommend the following: *) in your code, on the monitor instance, access monitor.Status.CookieId and copy that value somewhere *) In our web client, go to the 'Sessions' view, paste in this value and select the link that appears after searching This should presented you with a list of each individual session that has occurred on your specific machine and these values are updated within a few minutes of us receiving the data. You can press the reload button on top to check for new sessions. You can also navigate to the individual session to see what data we have received, including session time (as far as I recall). The Cookieid should remain unchanged for your machine (unless you repave your machine) and I think this is a quicker testing turnaround. As for setting up intervals in Feature Values, we require a minimum number of values in order to provide meaningful guidance on some suggested intervals. However, you should be able to setup these intervals even if we do not find enough samples to work from. Hope this helps Best regards Soren EQATEC Team

[Guest]

Hi Soren, thanks for the session tip. The dialog for setting up intervals is disabled due to lack of data. I checked the session details and can see that the featurevalue is received CalibratorUsed.DutLogger = 3 but doesn't seem to be displayed in its respective statisics group. regards jim

[Guest]

Hi Jim I'll check up on that, the feature value configuration dialog should not be disabled, you should still be able to register intervals manually on the lefthand side. There is an edit box that lights up to the left of the plus-image when you hover over it, which we need to make more visible. I can verify (by looking into our data source) that a number of CalibratorUsed.DutLogger = 3 have been tracked as feature values. I just checked and it seems that you have renamed that internal name 'CalibratorUsed.DutLogger' to 'Logging instruments' which may be why you can't seem to find it right away. You can change this name yourself by going into the 'Configure' menu item on the right hand side. Hope this helps EQATEC Team

[Guest]

Thanks S?ren, I found the "hidden" edit box. Entered the intervals and the data appeared. Also have double checked the session lengths again today and they appear correct. Thanks for your help.