DOYENSYS Knowledge Portal

We Welcome you to visit our DOYENSYS KNOWLEDGE PORTAL : Doyensys Knowledge Portal

Tuesday, June 14, 2016

Performing a healthcheck of a live GoldenGate configuration

How to do it...

The state of a GoldenGate replication can be checked by performing the following steps on
both source and target systems:

1. Check the GoldenGate event logger file ggserr.log for any error or warning
messages. You can find these messages using the egrep command as shown
in the following code:

     cd /u01/app/ggate
     egrep –i 'WARNING|ERROR' ggserr.log | more

2. Using the GoldenGate command line interface, check the status of the processes
as follows:

     GGSCI (prim1-ol6-112.localdomain) 1> status *

The status of all processes should be RUNNING.

3. If the status of any of the processes is not RUNNING, check the Process Report
files under $GG_HOME/dirrpt/. For example, for the PGGTEST1 process that has
an ABENDED status you should look for the error in the /u01/app/ggate/dirrpt/
PGGTEST1.rpt report file.

4. If the process is abended due to any data issue, you will see an ORA-error in
the Process Report file. GoldenGate writes failed record information to the
discard file.

5. If all the processes are running, check the lag between the systems as shown in the
following code:

     GGSCI (prim1-ol6-112.localdomain) 1> lag *
     Sending GETLAG request to EXTRACT EGGTEST1 ...
     Last record lag: 5 seconds.
     At EOF, no more records to process.
     Sending GETLAG request to EXTRACT PGGTEST1 ...
     Last record lag: 6 seconds.
     At EOF, no more records to process.

6. Sometimes there can be an issue with the underlying database due to which the
GoldenGate processes can abend. The state of the database can be checked from
the database alert log.

No comments: