Bruce,
Why wait nine hours? I would bring up the system with the correct time,
and run an offline backup of the DB(s), and you are good to go.
Regards,
Rob Arthur
IBM Certified Solutions Expert - DB2 UDB V6.1 Database Administration
Sr. I/T Specialist
IBM Global Services
Global Smart Card Solutions
Bruce Allen <[EMAIL PROTECTED]>@lists1.ba.best.com on 04/02/2001
05:29:35 AM
Please respond to [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
cc:
Subject: DB2EUG: Wrong system date on AIX
DB2 v6.1 on AIX 4.3.3
A couple of our unix boxes have been set up with the wrong time (a couple
are also correct). They present a DB2 Coordinated Universal Time. or CUT,
which is ahead by 9 hours. The timezone is also incorrect, but we plan to
change this environment setting and we expect no impact on the operation of
DB2. However, for the system time, we are not so certain. We have an
opportunity during a major outage this week, so we are evaluating our
options as a matter of urgency.
It seems the critical place of impact is roll forward recovery to a point
in time (specified in CUT). At this stage we have 2 options to change the
time.
1. Change the system time and do not start DB2 for 9 hours, thus waiting
until the clock 'catches up' to CUT so that ther are no overlapping CUT
times to confuse recovery (if required). We expect that this is the safe
option.
2. Change the system time and NOT wait for 9 hours. This means that DB2
will be in operation over the same CUT for a period of up to 9 hours. Is
this a problem? If we need to recover and roll forward, either to a point
in time or even to end of logs, over this time period, should we expect it
to work?
I dont know if there is any impact on the recovery history file, but we can
avoid the items listed in here, except for restore and roll forward (which
we would like to avoid, but cannot guarantee).
Is recovery the only thing in DB2 impacted by a change of the CUT?
BTW, we are also investigating if any applications have been using the SQL
reserved words (current timestamp etc) to determine any imapct on the
stored data.
Has anybody done this before?
Thanks and Regards,
Bruce Allen
=====
To unsubscribe, send 'unsubscribe' to [EMAIL PROTECTED]
For other info (and scripts), see http://people.mn.mediaone.net/scottrmcleod
