When I drill down from the Java requests dashboard to a single trace the time range gets locked.
The problem is that they time range stays locked when I go to another dashboard (hosts, agents, main it doesn't
really matter) so a user that checked a trace has his/her gui stuck on a time frame (in most cases not noticing it
or wanting it).
Flow is simple - open a java system that has traces > drill down to a trace (notice the time range locked)
>navigate to another screen (hosts/agents etc) notice time frame is locked > you can wait a couple of minutes and
navigate again and the display is has the old time range frozen.
There is actually a wcf feature called â€diagnostic workflowwâ€ which freezes the time range when navigating to that dashboard, and unfreezes (or takes the global timerange) when you go back into the dashboard which initiated the call. But if you navigate away totally from these dashboards into another one, then the timerange remains frozen/locked up. This is widely seen with dashboards drilled down from an alarms or alarms dashboard.
This feature was done on purpose, with the following use case: Once the user is in a diagnostic mode (either by manually freezing the timeframe, or by going into a diagnostic dashboard), the user would have the ability to navigate to any dashboard and continue the diagnosis for the same time frame at which the diagnosis started. For eg. while diagnosing a single trace, if they wanted to see how the resources usage was like on the monitored host or any other dashboard of interest, the time range is locked so that the user does not miss the context.