Problems with the build-IN SNTP-Client on cRIO9014&cRIO9012

Question

Hi out there,
I have a problem using the build-IN SNTP-Client of a cRIO-9014 (and a cRIO-9012).
It seems that the cRIO sometimes gets another time (e.g. 28.01.1920  12:03:00) other than it should get (e.g. 31.08.2014 09:59:00).
I've found this problem in our measurement-files. They had some unusual timestamps (as showed in the following example).
To prevent the SNTP-Client from doing too many damages to our measurement-files, I built a time-validator (with a range of 5 min), which checks one of my servers (and let it write every time change to a log file).
The Log-File looks like (Where "$Message" is a custom-message. It is unrelated to the problem, but occurs every minute):
31.08.2014          09:58:00 $Message
31.08.2014          09:59:00 $Message
31.08.2014          10:20:00 Setting Time from 28.01.1920 12:14:09 to: 31.08.2014 10:20:00
31.08.2014          10:21:00 $Message
31.08.2014          10:22:00 $Message
31.08.2014          10:54:00 $Message
31.08.2014          10:55:00 $Message
31.08.2014          10:59:00 $Message
31.08.2014          11:00:00 $Message
As you can see, the cRIO loses its time somewhere between 09:59:00 and 10:00:00(UTC) and it has been set back at 10:58:XX (UTC). It is shown with the gap between the messages at 10:55:00 and 10:59:00.
The configuration for the timesync was:
[TIME SYNC]
source.rtc.enable=True
source_priority=sntp;rtc;
source.sntp.enable=true
source.sntp.address=192.53.103.108
source.sntp.interval=60
source.sntp.port=123
source.sntp.verbose=false
source.sntp.log=false
sntp.address is a public  German time-server, hosted by our national institute.
I've set sntp.verbose and sntp.log to TRUE for more information, but have to wait some time for a logfile.
The network has the following scheme:
cRIO -> UMTS-Modem -> ISP-VPN -> Internet
Does anyone have had this problem and/or know, how to fix it?
Thanks for your help!
Cateros

Answer 1

Hi!
The problem happens sporadically, at first, we didn't even notice it, just that there are files missing and we got the strange logs. Now, with the workaround, we can finally trace the problem (and in some cases restore our measurement).
What do you mean with "connectthe cRIO directly to a router?"? Could you please explain it?
In the enclosure are more networkdevices, but only the cRIO uses the connection to the internet.
The "UMTS Modem" has also routing capacities, which we use to remote control both of the RIOs.
Thanks for your reply
Cateros