Goldengate database migration

cptkirksh

We are getting a new exadata and are looking to migrate our DBs from Oracle 18c to 19c. Some of the DBs on our 18c environment have just a few schemas while one in particular has 300. I have read about how after GG 12.2 I can utilize ADD SCHEMATRANDATA "Schema" on source and DBOPTIONS ENABLE_INSTANTIATION_FILTERING on the replicat to really integrate Oracle Data Pump in the whole migration process. This sounds like it would work great for those DBs that only have a small number of schemas. Is it best practice in a DB with a large number of schemas to skip the ADD SCHEMATRANDATA and the DBOPTIONS ENABLE_INSTANTIATION_FILTERING and just capture the SCN after export and start the replicat with that SCN? Thanks in advance.

ORASCN

Hi ,

We are not going to replicat the all schemas present in the database. And, the below commands,

add schematrandata schema preparecsn

add trandata schema.table preparecsn

Here comes the trick. We have a internal table where in the SCN information is maintained. When executing the above commands, the SCN will be automatically prepared for the tables / schema we mention. So, other tables or schemas which are not mentioned will not be affected.

Hope I explained it. Please let me know if you need further clarifications.

Regards,

Veera

ORASCN

Hi,

I have written an article on this. Please check the below link,

https://www.oracle-scn.com/oracle-goldengate-integration-with-datapump-dboptions-enable_instantiation_filtering/

Regards,

Veera

cptkirksh

Veera, Thanks for the link. I had read this before i posted my question and I guess my question still remains. How does one use GG when there are a possibility of hundreds of schemas inside a DB? Especially when in a large enterprise each of those schemas could belong to just as many users and applications.

ORASCN

Hi ,

We are not going to replicat the all schemas present in the database. And, the below commands,

add schematrandata schema preparecsn

add trandata schema.table preparecsn

Here comes the trick. We have a internal table where in the SCN information is maintained. When executing the above commands, the SCN will be automatically prepared for the tables / schema we mention. So, other tables or schemas which are not mentioned will not be affected.

Hope I explained it. Please let me know if you need further clarifications.

Regards,

Veera