Do I need to worry about Daylight Saving Time or time zones when using RTI Connext?

RTI Connext uses clocks to measure time and generate timestamps. The middleware uses two clocks: an internal clock and an external clock. The internal clock measures time and handles all timing in the middleware. The external clock is used solely to generate timestamps (such as the source timestamp and the reception timestamp), in addition to providing the time given by the DomainParticipant’s get_current_time() operation.

RTI Connext generally uses a mode of time that is not affected by Daylight Saving Time (DST) or time zones. Therefore you do not need to worry about these issues. 

  • On all UNIX and Win32 platforms, the time we use is the number of seconds since 00:00 GMT on January 1st, 1970. This is, of course, independent of the time zone. If you convert it to a time of day, you are using the time functions of your OS (which we hope you have patched independently from RTI.) 
    • A note about UNIX platforms (including Linux): The time obtained from the kernel will always be "UNIX time", which is independent of the time zone. But on PC hardware, it will be stored in the clock hardware in a different format, and that does depend on the time zone. So for example, a Linux system can be configured to store it either as GMT (which is less hassle, since the time base does not need to be updated when DST switches), or as a local time (so you can dual-boot to Windows/DOS, which uses local time in the clock). But what gettimeofday() gives you does not depend on the time zone (assuming your time is set correctly).
  • On Windows CE platforms, we use GetTickCount(), i.e., ticks since boot. This is not time-zone or DST dependent. 
  • On VxWorks platforms, we use tickGet, i.e., ticks since boot. This is also not time-zone or DST dependent. 

In recent version of RTI Connext, we provide two clock implementations (platform dependent): a real-time clock (default) and a monotonic clock. Please refer to the User Manual on Clock Selection. 

Keywords: