ER extent sizes
Posted in 2019
Topics: High Availability & Replication, Storage & Space Management, Server Administration, Platform-Specific Issues, Third-Party Tools & Monitoring
Hi all, Replicating: Informix 12.10.FC4 Solaris 10 --> Informix 12.10.FC12 RedHat I am setting up Enterprise Replication using OAT and running into an issue: the new tables created on the secondary server are created using 16/16 as the extent sizes instead of the defined values from the replicated server. I can find no way to override this. When I sync, I am now getting tables with 40 - 80 extents (and VERY LARGE next extent sizes due to the internal doubling. Ugh!). Any ideas? Thanks, Michael Hoffman Lead Production DBA Colorado Secretary of State
other than manually creating the tables on the "secondary" server with the correct sizing you can try exploring storage optimization methods : https://www.ibm.com/support/knowledgecenter/en/SSGU8G_12.1.0/com.ibm.admin.doc/i ds_admin_1421.htm Mark
Thanks, Mark. It looks as if creating the schema prior to syncing may be the best option. ER/Grid Sync (via OAT) also doesn't bring across indexes or triggers either. It does create unique indexes on the secondary, but using Informix' numeric naming scheme. So if I want an exact working duplicate of my Primary server, as well as correct extent sizing, having the schema in place prior to data sync seems to be the best practice. Mike
Hi Michael, I suppose you're implying ER's auto-create feature (--autocreate option)? Yes, this would give you a very rudimentarty table definition only, just enough so ER can replicate it. I think, meanwhile this even is documented. -> Predefine the target schema and all is well. BR, Andreas