I am loosing data when my time server reboots

[upchuckjunk]

My vi runs fine as long as I don't shut down the time server. When I shut down and reboot my time server, I lose about 45 minutes worth of data. My FP-2000's clock can't be that far off. I am running FP-reads every 5 seconds and averaging them for a 5 minute data point. I am almost out of ideas.

[Carl L.]

Rebooting the time server should not have any affect on a vi that is run embedded on the FP-2000. The vi will keep running and the onboard clock will continue to function. For more information the time server and the onboard clock see the following links.
http://digital.ni.com/public.nsf/3efedde4322fef19862567740067f3cc/81585d2a54d1c070862569e70071639a?OpenDocument
http://digital.ni.com/public.nsf/3efedde4322fef19862567740067f3cc/7c6053a96294736286256bd7007406dd?OpenDocument
One of them refers to the FP-1600, but the time server synchronization is the same on a FP-2000.
You can add code to your vi that toggles one of the LEDs on the FP-2000 (channel 5-8), to verify that the vi is running. Does your program depend on the computer that is the time server in so
me other way than just determining the time. Does it save the data to the FP-2000 or to a host computer?

[upchuckjunk]

My VI changes the time on the FP-2000 to MST and
saves data locally. I will check out the info about the time server and onboard clock. I know that VI is running but am unable to track down the fail time. It always loses data when I come back over a wknd and reboot the time server.

[upchuckjunk]

I looked at the time server info and doesn't explain what I am seeing with my VI. I will try the LED flash on one of my AO modules just to see if the VI pauses or stops at some point. I will also try to include a time-out timer for the loop and log that info as well as record any unusual errors. At this point in development it's crunch time, I need a fix for this issue quick.

[upchuckjunk]

Well, I have now tried your suggestions, it seems that I still loose between 30 and 45 minutes of data. I don't know if this is really data lost or just time stamp issues. I have my LED flashing on the AO channel, it doesn't stop running even when the time server is rebooted. Also added a loop counter, seems that the counter is off by 6 minutes instead of 5 when I check the counts. (I am trying a 5 minute ave.)
I am almost out of ideas. The system is autonomous except for the time server. I can let the system run for days, if I don't reboot the time server, and it will be fine.

-- Confused and frustrated!

[Carl L.]

From the information here it is very hard to tell why it is not working. There is absolutely no reason that rebooting a time server should change the behaviour of an application. I am assuming that you are using the built in time server that is set up in the Configure Device option in Remote System Explorer. If you have created a custom time server and you are receiving the values through datasocket, then an explanation could be that datasocket is waiting for an updated value, but doesn't receive on which could then block the rest of the program if it is all in one loop.
Earlier you said that the time server reboots during the weekend. This of course makes it hard to monitor what is happening in the VI. If you can have the time server reboot
while you are there then you can use all the debugging tools from LabVIEW such as probes to see what is happening in the VI while the time server reboots.
The next step would be breaking the program down in to smaller parts and see how they behave while the time server reboots.

[upchuckjunk]

Here is the latest. I have rebooted the time server, no problems. When I shutdown the time server, the FP-2000 runs fine. When I reboot the FP-2000 with the time server down, I get crap for time and date when it reboots. Do I have to activate an onboard time service in the FP-2000? Why doesn't it keep the time based upon the last known update from the time server? I am not mucking with the time local to the FP-2000 other than reseting to MST. (subtract 7 hours)
I don't adjust the date inside my VI at all but the date is reset to 12/01/97 in the absence of the time server. This makes no sense at all. It also throws a huge monkey wrench into my datalogger system.