Re: NTP on linux box was ahead of 5 hrs

From: Andrew Kerber <andrew.kerber_at_gmail.com>
Date: Mon, 18 Nov 2013 10:54:59 -0600
Message-ID: <CAJvnOJYvh5tKmCFTgPM7n=Rmx6WEQGAoAfoV6+Ad_EQaA5Aftg_at_mail.gmail.com>



You might want to take some notes about when it happened, how long you think it was a problem, etc, so that should you need to run a recovery over that time frame you will have a good idea what happened. Trying to recall all of this a few months down the road, might create even more headaches.

On Mon, Nov 18, 2013 at 10:47 AM, Niall Litchfield < niall.litchfield_at_gmail.com> wrote:

> You likely need to understand how the application works and what you are
> worried about. This doesn't have a general answer.
>
> Making the simplified assumption that everyone is in the same timezone as
> the database server you still need to distinguish between code such as
>
> "insert into time_tab(cdate) values(:dt)" and "insert into time_tab(cdate)
> values(sysdate)" and "insert into time_tab(cdate) values('string from
> appserver')"
>
>
>
> On Mon, Nov 18, 2013 at 4:10 PM, nilesh kumar <nileshkum_at_gmail.com> wrote:
>
>> Thanks Martijn,
>>
>> Suppose this database belongs to a Banking sector, and few users perform
>> the transaction when NTP was 5 hrs ahead. And Suppose if user asks the
>> details of their transaction in that case we are in bit confusion state
>> right.
>>
>> In general I was checking with everyone, if someone has faced this kind
>> of issue and the work around they did to fix this one.
>>
>> Thanks for replying Martijn
>>
>> Thanks
>> Nilesh
>>
>> On Mon, Nov 18, 2013 at 8:10 PM, Martijn Bos <maboc_at_maboc.nl> wrote:
>>
>>> Hi,
>>>
>>> See my ideas inline.
>>>
>>> Best Regards,
>>> Martijn
>>>
>>>
>>> Best Regards,
>>> Martijn
>>>
>>>
>>> On Mon, Nov 18, 2013 at 07:03:32PM +0530, nilesh kumar wrote:
>>> > Hello All,
>>> >
>>> > NTP was running ahead of 5 hrs of normal time on Linux box.
>>> >
>>> > We have oracle database running on that server version 10g.
>>> >
>>> > Network team tried to synchronize the NTP and they synchronized the
>>> NTP on
>>> > the server.
>>> >
>>> > >From oracle database endi have few questions.
>>> >
>>> > I took a levelzero backup and all archive log backup immediately and
>>> asked
>>> > the application team to release all the conections to the database and
>>> > connection refresh so that none of the users are connected to the
>>> > database, and then bounced the db and again took a level zero backup.
>>> >
>>> > Am i missing any thing in this case that need to be taken care.
>>> >
>>> > few more question in mind:-
>>> > ----------------------------------------
>>> >
>>> > Database was running and users were connected to the database during
>>> this
>>> > issue occured, if some users committed the transactions and when NTP
>>> was 5
>>> > hrs ahead and does any inconsistency occur after the NTP was
>>> synchronized.
>>>
>>> The database ticks around scn's so, on a 'physical' level I do not
>>> expect inconsistencys.
>>> On a logical level (data-model), you may expect inconsistencys when
>>> 'sysdate' has been used to fill some rows.
>>> (Also think of RI where sysdate is used in a column)
>>>
>>> >
>>> > ----------------------------------------
>>> > This has not occured but have a question for the below conditon.
>>> > If suppose the DB would have crashed after the NTP synchronization and
>>> > backup taken last sunday level zero, till what time would we able to
>>> > recover the DB, as few archives have generated after when the NTP was
>>> 5 hrs
>>> > ahead in time.
>>> >
>>>
>>> I think that a Point In Time recovery will be a little hazardous.
>>> In this case I would try to restore to a sequence number or even a scn.
>>>
>>> I _think_ that a restore would restore to the time before the
>>> re-syncronaziation (the first time that time x appeared).
>>>
>>> > Please advise
>>> >
>>> > Thanks
>>> > Nilesh
>>> >
>>> >
>>> >
>>> > --
>>> > Thanks & Regards
>>> > Nilesh
>>> > Oracle DBA (sqlplus "/as sysdba" is my world)
>>> >
>>> > "As our circle of knowledge expands, so does the circumference of
>>> darkness
>>> > surrounding it."--Einstein
>>> > "All men die, few men ever truly live" -- Braveheart
>>> > "You can get help from teachers, but you are going to have to learn a
>>> lot
>>> > by yourself, sitting alone in a room." --Dr. Seuss
>>> > "Know me for what I do and for not what I am. Judge me for what I can
>>> do
>>> > and not for what I have done."-Unknown
>>> > "Don't walk in front of me, I may not follow. Don't walk behind me, I
>>> may
>>> > not lead. Just walk beside me and be my friend."- Albert Camus
>>> > "Never argue with an idiot. They drag you down to their level and then
>>> beat
>>> > you with experience." - Unknown
>>> >
>>> >
>>>
>>
>>
>>
>> --
>> Thanks & Regards
>> Nilesh
>> Oracle DBA (sqlplus "/as sysdba" is my world)
>>
>> "As our circle of knowledge expands, so does the circumference of
>> darkness surrounding it."--Einstein
>> "All men die, few men ever truly live" -- Braveheart
>> "You can get help from teachers, but you are going to have to learn a lot
>> by yourself, sitting alone in a room." --Dr. Seuss
>> "Know me for what I do and for not what I am. Judge me for what I can do
>> and not for what I have done."-Unknown
>> "Don't walk in front of me, I may not follow. Don't walk behind me, I may
>> not lead. Just walk beside me and be my friend."- Albert Camus
>> "Never argue with an idiot. They drag you down to their level and then
>> beat you with experience." - Unknown
>>
>
>
>
> --
> Niall Litchfield
> Oracle DBA
> http://www.orawin.info
>

-- 
Andrew W. Kerber

'If at first you dont succeed, dont take up skydiving.'

--
http://www.freelists.org/webpage/oracle-l
Received on Mon Nov 18 2013 - 17:54:59 CET

Original text of this message