I have a DB2 locate in IBM z/OS platform for online transaction, want to use iWay DataMigrater tool to transform data into Linux DB2 and same time using CDC to know any table that is IUD(Insert/Update/Delete).
Current tables in DB2 are around 500, approximately around 1TB of total size.
Here is my questions:
1. For the initial load from source tables(500 tables) do I need to be offline the DB2 from transaction? or I can get every data as current state of data as User's manual mention?
2. Any system/hardware recommendation for CDC usage for 500 tables?
Thanks,This message has been edited. Last edited by: FP Mod Chuck,
1. You should be able to run a flow with initial load to "get every data as current state of data".
2. Could you please open a Hottrack case and provide the details about your environment and requirements? For example, one of the details about the requirements that we'd like to know is: what is the naming convention for the tables and synonyms names on the target system.
Thank you Maria for your reply, about the requirement I need to be more detail information from user, so i can get back to you. For all I know now is to transform DB2's tables from z/OS into Linux OS as DB2 via iWay DataMigrater.
To continue my question:
1. never done this large amount of table before, not sure how long it take for do initial load? and suppose it take 30min. to complete the load, after 30min. will use CDC to track 30min. data, assume the target will get the data after 30min. but using CDC to tack back 30min. might getting few data repeatedly, which is wrong for the table's data. Or should I offline these table till it's complete?
2. CDC is base on collecting DB's log to do IUD into the target, what if DBA decide to use massive IUD or 'TRUNCATE TABLE' to avoid getting logs, from this point should we re-do initial load again? since there is no logs to identify where source table has been change?
|Powered by Social Strata|