Zero Downtime Installation Guide for Oracle

Zero Downtime Installation Guide for Oracle

Configure and Deploy the Messaging Stream

Configure and Deploy the Messaging Stream

Configure and deploy the messaging stream on both systems.
  1. Log in to both the source schema and the target schema.
  2. In both schemas, configure the messaging stream:
    DECLARE out_error_msg VARCHAR2(32000); out_return_code INT; BEGIN CMXZDT.CONFIGURE_GGS_EVENT_REPLICAT(CMXZDT.zdt_local); CMXZDT.CONFIGURE_GGS_EVENT_EXTRACT(CMXZDT.zdt_local); CMXZDT.START_EVENT_QUEUE(out_error_message => out_error_msg, out_return_code => out_return_code); dbms_output.put_line('out_return_code => '||out_return_code||' out_error_msg => '||substr(out_error_msg,1,250)); END; /
  3. On both systems, run the
    info all
    command from the ggsci prompt. Verify that all event queue processes are running.
  4. In the source schema, generate ggsci parameter files and GoldenGate processes:
    BEGIN CMXZDT.CONFIGURE_GGS_EXTRACT(CMXZDT.zdt_local); END; /
  5. In the target schema, generate ggsci parameter files and GoldenGate processes:
    BEGIN CMXZDT.CONFIGURE_GGS_REPLICAT(CMXZDT.zdt_local); END; /
  6. In both schemas, set sequences so that they do not overlap:
    exec cmxzdt.configure_sequences;
    This command sets the source schema to even and the target schema to odd to avoid overlaps.
  7. In the source schema, start the PUMP and EXTRACT processes:
    DECLARE out_error_msg VARCHAR2(32000); out_return_code INT; begin cmxzdt.start_replication_extract(in_env_type => cmxzdt.zdt_local, out_error_message => out_error_msg, out_return_code => out_return_code); dbms_output.put_line('out_return_code => '||out_return_code||' out_error_msg => '||substr(out_error_msg,1,250)); end; /
  8. In the target schema, start the REPLICATE processes:
    DECLARE out_error_msg VARCHAR2(32000); out_return_code INT; begin cmxzdt.start_replication_replay(in_env_type => cmxzdt.zdt_local, out_error_message => out_error_msg, out_return_code => out_return_code); dbms_output.put_line('out_return_code => '||out_return_code||' out_error_msg => '||substr(out_error_msg,1,250)); end; /
  9. On both systems, run the
    info all
    command from the ggsci prompt.
  10. Verify that the
    info all
    summary displays ten processes, each with a status of RUNNING.
    • Messaging stream processes have the prefixes EQ, PQ, and RQ.
    • Data extract processes have the prefixes E_ and P_. These processes extract data from the source system.
    • Data replication processes have the prefix R_. These processes replicate data on the target system.
    The following output shows the processes, with one replication process not running:
    GGSCI (hostname) 13> info all Program Status Group Lag Time Since Chkpt MANAGER RUNNING EXTRACT RUNNING EQENVA 00:00:00 00:00:00 EXTRACT RUNNING EQENVB 00:00:00 00:00:00 EXTRACT RUNNING E_ENVA 00:00:00 00:00:08 EXTRACT RUNNING PQENVA 00:00:00 00:00:09 EXTRACT RUNNING PQENVB 00:00:00 00:00:00 EXTRACT RUNNING P_ENVA 00:00:00 00:00:01 REPLICAT RUNNING RQENVA 00:00:00 00:00:00 REPLICAT RUNNING RQENVB 00:00:00 00:00:08 REPLICAT ABENDED R_ENVB 169:25:21 00:00:02 REPLICAT RUNNING R_ENVBU 00:00:00 00:00:07
If a process is not running, see the log files in the
D:/ggs/dirrpt
directory. There is a separate .dsc and .rpt file for each process. For more information about troubleshooting process errors, see the Oracle GoldenGate documentation for administrators.

0 COMMENTS

We’d like to hear from you!