Enterprise replication on large tables
Posted in 2012
Topics: High Availability & Replication, Platform-Specific Issues
I'm in the process of replicating some large tables and I'm noticing that the table on the slave, or destination server, has twice the amount of posts as the original, or master server. Is this as planned? Will ER clean up upon completion? Or is something going wrong here? We're using update anywhere so we have a back out plan in case something doesn't work. ER master: HP-UX Risc 11.11 11.70 xC4 ER slave: Solaris 10 Intel 11.70 xC1
The table on the slave now is going removing posts, so I assume it's getting back in sync, so then what's the recommendation on running Enterprise Replication for the first time. Start from a clean server, no data, or import the data via a cold restore on the slave and then peform the sync/check? I got the sneaking suspicion I chose cold restore and that was wrong, as my tables double in size and then slowly posts get removed during the "check --repair". As opposed to just inserting it all from the get go.
I don't believe it's normal, in my case it haven't happened in this way, have you set crcols for using update anywhere? You should first configure scope by row at first sync avoiding long transaction, after you should recreate the replicate changing the scope by transaction. --scope=row Celso Cabral Coimbra Administrador de Banco de Dados ClearTech Ltda "Trust at the heart of Communications" Tel. (11) 3576-4509 -----Mensagem original----- De: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] Em nome de LUKE SIMMONS Enviada em: domingo, 10 de junho de 2012 08:16 Para: ids@iiug.org Assunto: Enterprise replication on large tables [27344] I'm in the process of replicating some large tables and I'm noticing that the table on the slave, or destination server, has twice the amount of posts as the original, or master server. Is this as planned? Will ER clean up upon completion? Or is something going wrong here? We're using update anywhere so we have a back out plan in case something doesn't work. ER master: HP-UX Risc 11.11 11.70 xC4 ER slave: Solaris 10 Intel 11.70 xC1 ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.