Re: server time reset

"fitzjarrell_at_cox.net" <oratune_at_msn.com> said:
>On Jun 17, 7:52 am, Ben <benal..._at_yahoo.com> wrote:
>> 10.2.0.2 EE aix 5.2
>>
>> we have a server that somehow got behind in time by 5 minutes. I
>> mentioned it to the sysadmin and next thing I knew she had changed the
>> time on the server ahead 5 minutes. First thought was panic and what
>> is going to happen to the database now. Nothing seemed to happen.

Most likely you'll be fine. Backwards 5 min could have been a different case.

>This happens, twice yearly, when Daylight Savings Time commences and
>when Standard Time returns, and this involves an hour difference
>either way, and I've yet to experience any issues with Oracle. Yes,
>during the above-mentioned scenario you can duplicate an hour which
>could make it difficult to perform a PITR based upon clock time, but
>you still have the SCN as a reference point, and that doesn't change
>when the clock does.

I'm pretty certain that Oracle (as Unix at OS level) does store and handle timestamps in some TZ (and DST) independent format. F.ex. the system clock on Unix machines is a linear second counter from 1.1.1970. To achieve DST changes, only the time display offset is changed, not the system clock itself. The difference in the situation described by Ben was that the system clock was abruptly bumped 5 minutes forward.

One recommendation for Ben is to check with the sysadmin that the Oracle server is running NTP time synchronization (and is being served by stable time servers). Most probably it isn't, and taking NTP in use would avoid this kind of problem in the future.

-- 
Wolf  a.k.a.  Juha Laiho     Espoo, Finland
(GC 3.0) GIT d- s+: a C++ ULSH++++$ P++@ L+++ E- W+$@ N++ !K w !O !M V
         PS(+) PE Y+ PGP(+) t- 5 !X R !tv b+ !DI D G e+ h---- r+++ y++++
"...cancel my subscription to the resurrection!" (Jim Morrison)