What, if anything, might fail in these cases:
* Master and Slave are both set to "System" timezone, but the systems
have different timezones.
* Master and Slave are both set to a specific timezone, say PDT, but the
systems have different timezones.
* Master is Pacific time; slave is UTC. Pacific time switched yesterday
due to DST. (Replication hung on duplicate key; trying to trace it back
to timezone issues. PT slaves did not hang.)
* Upgrade machines (Master, Slaves) one at a time -- either changing
system timezone or changing MySQL timezone. Or must they be done
simultaneously.
Correct?
* DATETIME and DATE values in tables will be totally unaffected by
changes to timezone settings in OS or MySQL.
* TIMESTAMP values on disk are _not_ affected by OS/MySQL changes
(assuming one switches correctly).
* INSERT/SELECT of TIMESTAMP columns _will_ return different values
before versus after OS/MySQL change to timezone setting.
| Thread |
|---|
| • changing of timezone | Rick James | 15 Mar |