test report wrong time stamp

[jchurlik4]

I have a couple of test systems running teststand 4.2 in which when the tests are run the report date and time stamp is wrong as compared to the file creation date and time. It can be off as few as a few minuts or as much as a day. Has any one else experienced this ? Any Ideas?

[EEsRULEtheWORLD]

Hi jchurlik4,

How long do your tests run?  It may be that the report's timestamp and file creation date and time are taken at different instances in the run of the test sequence.  This timestamp is pulled from the system clock, so check to see if the time that it takes the test to run corresponds with the difference in the time indicated by the timestamp and the file creation time.

[jchurlik4]

Unfortunately I am dealing with something else. That was my first intial thought. Not the case though as I have run a test timed it and had the timestamp be off as much a day and 3 hours. My next thought is I may have a memory leak causing performance issues. Does anyone know how test stand get the date timestamp string. Is it from a windows API call? The only other thing I have noticed is when a large number of report files are in the report directory test stand starts to perform slower. Any more ideas?

[Ski]

Hello jchurlik4,

I am very new to TS/LV so I hope I don't lead you astray.  I have just started looking into a similar problem with our TS systems.  They were set up by someone else and I am trying to understand how it all goes together and learn TS/LV at the same time.  We also found that our data records were showing up with timestamps that were off by up to a day.  When I told our developer about this he indicated that it was because the files were not actually being created until we started testing the "next" UUT.  It might be a day, (or overnight), before the next UUT was tested and hence, why the files had creation dates that were so far off from the timestamp values inside the files.  

I can't believe this would be standard behavior in the Sequential Model, so I assume this means that our ProcessModel has been updated incorrectly.  There is no reason to wait for the next UUT before writing out the results from the last UUT, but as I said I am just learning how all this works and have not figured out what changes were made to the ProcessModel yet. 

[jchurlik4]

Ski,

Thanks for the reply. You post is correct but not my problem that I am having.

You can look at the model that you are using and follow the report generation steps. If you are new to this; one good way to learn the flow is to single step through the process model as your test is running.

[Ski]

Thanks for the pointer, good luck with your issue.

[dug9000]

The date in the report is the date testing started which it not necessarily the same as the date the report file was created. TestStand uses the Win32 API GetLocalTime() and GetDateFormat(). Have you modified the process model in anyway? Is your machine by any chance going to sleep or hibernating while TestStand is running (This should only matter if you are calling Date() or Time() expression functions with Seconds() as the source for the date, which the default process models do not do)?

-Doug

[Manooch_H]

jchurlike4 -

Could you please provide screenshots that explain the behavior you are experiencing? I think this would give us a better idea of what the problem is.

I look forward to hearing back from you.

[Ski]

I'm not sure if it's fair to come back a year later but what the heck!  I have a tiny bit more data that may help explain what I am seeing and I have been ignoring this for too darn long...

I am working on code that someone else set up, so I apologize if I don't have all the details.  We are using TS 4.2, and using the Sequential Process Model, testUUT's, dumping the results to an XML file.  

When I test the first unit and the XML file is generated, the "StartTime" value inside the XML file is correct. <happy>  After that, each successive test that is run produces an XML file where the "StartTime" element is farther and farther off.  Obviously, it is not being reset or updated for each subsequent test. <unhappy>  

Assuming I have *not* gone in and screwed with the default process model.  How can I configure either the report options or process model to update that value for every unit tested.  Note that when each new instrument is tested, we begin by scanning in a unique barcode for it.  (That would seem like a pretty good spot to insert something that updated the "StartTime" entity.)

I'm just not sure how to do it.

Thanks in advance for any help on this.

Ski

[Manooch_H · Accepted solution]

Ski -

When you say that it is farther and farther off, could you elaborate on this? Are the times the same as the first UUT or are they different from the first UUT but not correct? If the latter, how far off is the time with each successive UUT? Is it the same interval difference or does the interval difference get larger with each successive UUT?

If the times are the same as the first UUT, you are likely running into a known issue documented in the below KB:

Why Does the Time Field of the UUT Report Header Remain Constant for Multiple UUTs when Generating X...

The issue documented in the above mentioned KB is fixed in TestStand 2010. Hope this helps.