Golden Gate functionality during new deployments

Question

Hi can anybody tell me what can be done during deployment activity is performed for Active-Active(bi-directional) Golden Gate configuration?
In our case during new deployments we fail-over complete traffic from once Data centre to another active data centre.
After traffic fail-over we perform deployment work on Database. Which might insert/add columns etc.....
Adding columns and inserting data on once database makes Golden Gate to replicate them to target database which is not desired for a running database.
Anybody suggest what can be done in such scenario?
Also if we try to ignore those tables, we have seen a situation where Golden Gate will fail saying parent keys not found.
Any suggestion on handling such situation will help me a lot.
Thanks in advance.

Answer 1

Hello,
Maybe a late reply, but have you thought of creating an extract process specif to the deployment (active data center)?
You could use the extracted trails and pump them over to the other data center and run the trails/changes as a batch job.
I recently had a similar scenario, and create a specific extract(active center) and batch replicat(passive center). The parameter nodbcheckpoint tell Oracle GG that you are running a batch job.
Here is an example:
GGSCI> ADD REPLICAT LOADDRP, EXTTRAIL ./dirdat/tx nodbcheckpoint
GGSCI> EDIT PARAMS LOADDRP
REPLICAT LOADDRP
SETENV (NLS_LANG = "AMERICAN_AMERICA.WE8MSWIN1252")
EXTTRAIL ./dirdat/tx
HANDLECOLLISIONS
--END RUNTIME
USERID GGATE@MYDB01, PASSWORD *******
ASSUMETARGETDEFS
DISCARDFILE ./dirrpt/loadtemppassive.dsc, purge
STATOPTIONS RESETREPORTSTATS
DDL INCLUDE ALL
DDLERROR DEFAULT IGNORE RETRYOP MAXRETRIES 5
REPORT AT 00:01
REPORTROLLOVER AT 00:01
REPORTCOUNT EVERY 60 SECONDS, RATE
DDLOPTIONS REPORT
MAP I_S_L_2.*, TARGET I_S_L_2.*;
ETC......Hope this gets you on the right track.
Sincerely
Jan S.