Due to my time zone being affected by DST, my monthly reports for March and November are skewed by the 23, and 25 hour day. I'm under the impression that the only way to correct this issue is to modify the localhost.tz file which is based on UTC?
You can dump the content of your localhost.tz file via the below method. However, It affects all things related to PI so you may want to be very careful.
PIPC\adm> pidiag.exe -tz -dump
Thank You Jinmo Yi!
Now, the said -dump does not pull the contents correct, it's merely a copy? Then, how does on edit the contents?
Here is an article describing more details on how to edit the file:
Before you make changes, I think you should be careful to make sure that modifying the localhost.tz really is the correct solution.
Can you explain more what your problem is?
Thank You Dan, the problem arises when we pull data via Pi data link to an excel work book for the month of March and November. There are 4 main work books for the year that cover months with a 28, 29, 30, 31 day periods. Although March is a 31 day period, because of DST it has only 23 hours and then the opposite in November (30 day) which has 25 hours, 2 periods hour ending 02:00. November has not been as much of an issue as March because Pi apparently sums the 2, hour ending 02:00 so no data is actually lost. My current solution has been to create near identical work books for each day DST falls on. The difference is, I've removed all 02:00 references, making the work book span 23 hours instead of 24. This effort resolves the lost data errors, but it adds 7 work books to the list to choose from, and if I can made the process easier via the fields in the localhost file, well it's something I'm willing to try because I feel that I have depleted all other known options. I appreciate your feedback.
So you want to ignore DST? That means when a DST transition occurs, your PI Server using server time will appear an hour behind wall-clock time until DST ends. There probably are better solutions to this and maybe some client side manipulations. What type of data queries are you doing?
Dan, I hope you enjoyed your weekend. I work in power generation and report several variations of electrical values to different entities. I've inherited this problem from my manager who recently retired, the sheets are built on a tag reference from our DCS with 5 minute averages for every hour for each day of the month then summarized on a totals sheet for the month. After some in depth reading, I believe I will stick to my current solution regardless of the number of sheets required to accomplish the task.
We've never asked you what you were trying to do in PI DataLink. Could you briefly explain your initial issue? Perhaps there are other options for your problem. What is that you are trying to calculate in excel?
Good morning Jinmo, I'm not sure if you can see the entire thread but if so read my reply to Dan Fishman, if not here is my reply to Dan, I work in power generation and report several variations of electrical values to different entities. I've inherited this problem from my manager who recently retired, the sheets are built on a tag reference from our DCS historian with 5 minute averages for every hour for each day of the month then summarized on a totals sheet for the month. After some in depth reading, I believe I will stick to my current solution regardless of the number of sheets required to accomplish the task.
And BTW, thank you for the instruction on how to read and edit the localhost.tz file.
Unless there is some sort of oddity for the month of November I believe I will call the current issue solved.
Thank you for your assistance with my perceived problem. While I originally was not responsible for the task at hand, I believe the best solution has been implemented. At first it was difficult to believe that my issue had not been addressed, but I guess, the real issue is the way we build our workbooks. I've made the necessary modifications and am happy with the results. Thanks again for your response's!
Retrieving data ...