db2 8.2.0
windows

how can i acheive restoring and bringing the target DB to be in sync with Sorce DB after having restored the TGT DB with a image of SRC db.
i Dont want my Production DB to Go down
also complete refresh of Tables takes long time as they are quite big tables

Presently we are doing offline backups > restore > manual refresh,
But now we are thinking of changing it to Online to prevent downtime for Production Server

as you say :
take online backup - move backup image and log files to target system
restore db - rollforward db - start capture - manual refresh (without export import) and start apply again
this should do

i m a bit doubtful regarding manual refresh .
even if i have transfered logs to taget db and roll fwd it to end of logs.
but in the mean time wont the live production DB will get ahead to Target DB.
How to bring in sync among them?

manual refresh will just take the current timestamp and go forward from there..(would it solve the diff in rows between source and target)

i solution i can think of is that while starting 1st time , i have a cold start
but while testing it took very long time

to be 100% in synch you do :
- cold start and refresh all
- manual refresh and do export/load yourself

thanks Guy Przytula

it would be very helpfull for me if u can elaborate further on
- manual refresh and do export/load yourself

automatic refresh = refresh done by apply process
you don't have any control and need much log space
with manual refresh : initiated from replication center - right click on subscription - refresh - manual
this will tell you the names of the tables to be exported from source db
load these files to target db with load and nonrecoverable
continue the refresh process, that will reset the pruncntl/subset table
capture will start to capture from this timestamp

i have used asnload for exporting/loading.
quite fast than earlier tests, But still took approx. 3 hrs to cover up.
Do you have any idea how to improve its performance.
Also many of tables went into integrity pending because of Load.
ne idea to prevent it.

regards,

if foreign keys are available - tables are in check pending : this can not be avoided with load
to fasten the load : utilheap - sortheap - SHEAPTHRES - MAX_QUERYDEGREE are important parameters

actually , i have dropped all the references on target DB before starting replication...
what else could be reason for tables to get into check pending
there was no interruption while asnload was running..

thanks ,