Annoying TZ Problem
Posted: Thu Jun 13, 2019 11:59 pm
I finally rebuilt my ZM machine after a long time on 1.29.x
Now on 19.04 and 1.32.3 and it MOSTLY works fine, but I have an annoying issue when I click on the console screen, total events for the last hour, I'm apparently 7 or so hours ahead on the timeline. (Nothing displays, if I click TIMELINE, I see an empty timeline in the future.)
If I zoom out and back, events do exist and play at their correct position on the timeline in local time.
Just opening the timeline is in the future for me (apparently UTC, right now I'm UTC -7 so if I do timeline from the last hour [say at 12:00 noon] I'll get a an empty timeline around 7PM. If I arrow back to noon, my events are there at their correct time.)
MY TIMEZONES ARE CORRECT, system time is correct, php.ini is correct, etc. I suspect there is an offset snuck in to the DB itself. (like, I may have initially fired up the DB or copied the DB in while in UTC time.)
I suspect it will work itself out as events get purged, but I have a lot of space With many terabytes of storage it may take weeks for the first purges to hit organically.
Is there a graceful way to iron this out?
I haven't looked deeply into it, but figured I could modify the purge script to kill at 10% full and run it manually and see what happens.
Plan B would be nuke the DB and configure again, (but rather not, already have a lot of tuning done, just ironing out one RTSP H264 smearing problem (getting better).)
Thanks for your ideas!
Now on 19.04 and 1.32.3 and it MOSTLY works fine, but I have an annoying issue when I click on the console screen, total events for the last hour, I'm apparently 7 or so hours ahead on the timeline. (Nothing displays, if I click TIMELINE, I see an empty timeline in the future.)
If I zoom out and back, events do exist and play at their correct position on the timeline in local time.
Just opening the timeline is in the future for me (apparently UTC, right now I'm UTC -7 so if I do timeline from the last hour [say at 12:00 noon] I'll get a an empty timeline around 7PM. If I arrow back to noon, my events are there at their correct time.)
MY TIMEZONES ARE CORRECT, system time is correct, php.ini is correct, etc. I suspect there is an offset snuck in to the DB itself. (like, I may have initially fired up the DB or copied the DB in while in UTC time.)
I suspect it will work itself out as events get purged, but I have a lot of space With many terabytes of storage it may take weeks for the first purges to hit organically.
Is there a graceful way to iron this out?
I haven't looked deeply into it, but figured I could modify the purge script to kill at 10% full and run it manually and see what happens.
Plan B would be nuke the DB and configure again, (but rather not, already have a lot of tuning done, just ironing out one RTSP H264 smearing problem (getting better).)
Thanks for your ideas!