• New Feature
  • Status: Open
  • 2 Major
  • Resolution:
  • prodmgmt
  • Reporter: teck
  • September 15, 2008
  • 0
  • Watchers: 0
  • March 19, 2010


I’ve debugged this a few times now in various customer contexts that it seems we could do something a little more proactive about it at some point. When the timezone is different between two L1s, it can make for an immediate time jump that instantly causes sessions to become invalid.

Seems like we could store the local timezone setting along with a roots in this case and at least produce a warning when a discrepency is noticed (ie. the timezone of the node faulting the root is different from the timezone of the VM that first created it). Doing something similar but at the level of each session would work too, but I can’t see any advantage to storing timezone info for every session