Hi all,

I have to start new ER replicates with tables up to 400 Millions rows. I
want to avoid to unload/load the tables because I have problems with piping
the data with HPL. Furthermore I had to stop all transactions to ensure that
the data the same on both servers before starting the replicate.

Is it possible to use the command "cdr start replicate xxx_repl targetServer
--syncdatasource=sourceServer" with such a high amount of rows? Remark:
Transactions on the sourcetable can't be stopped meanwhile.

How would you start a replicate with many rows? Is there a good strategy or
a best pratice?

TIA,
Reinhard.

-----Original Message-----
From: informix-list-bounces at (no spam) iiug.org
[mailto:informix-list-bounces at (no spam) iiug.org]On Behalf Of theBP
Sent: Wednesday, October 14, 2009 12:14 PM
To: informix-list at (no spam) iiug.org
Subject: Re: Start Replicate for huge tables


Habichtsberg, Reinhard wrote:
Hi all,

I have to start new ER replicates with tables up to 400
Millions rows. I
want to avoid to unload/load the tables because I have
problems with piping
the data with HPL. Furthermore I had to stop all
transactions to ensure that
the data the same on both servers before starting the replicate.

Is it possible to use the command "cdr start replicate
xxx_repl targetServer
--syncdatasource=sourceServer" with such a high amount of
rows? Remark:
Transactions on the sourcetable can't be stopped meanwhile.

How would you start a replicate with many rows? Is there a
good strategy or
a best practice?

TIA,
Reinhard.

What version?

Answer is "yes", but you could consider ifx_replcheck :

http://publib.boulder.ibm.com/infocenter/idshelp/v115/index.js
p?tab=search&searchWord=ifx_replcheck+