Troubleshooting Sharding Deployment

After all the prerequisites and configuration of the Sharding, during the deployment, it will create the shards on the specified targets based on the number of shards configured. Now the question is, how to troubleshoot if the sharding deployment failed?

We will see an example of sharding deployment output. In this configuration, we are deploying 2 shards for primary and 2 shards for standby. While deploying shard sh3 the deploy command terminated.

GDSCTL>deploy
deploy: examining configuration...
deploy: deploying primary shard 'sh1' ...
deploy: network listener configuration successful at destination 'ora_c1'
deploy: starting DBCA at destination 'ora_c1' to create primary shard 'sh1' ...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: waiting for 1 DBCA primary creation job(s) to complete...
deploy: DBCA primary creation job succeeded at destination 'ora_c1' for shard 'sh1'
deploy: deploying standby shard 'sh41' ...
deploy: network listener configuration successful at destination 'ora_c3'
deploy: starting DBCA at destination 'ora_c3' to create standby shard 'sh41' ...
Remote job failed with error:
EXTERNAL_LOG_ID="job_73616_22",
USERNAME="oracle"
For more details:
 select destination, output from all_scheduler_job_run_details
 where job_name='SHARD_SH41_DBCREATE'
deploy: DBCA standby creation job failed at destination 'ora_c3' for shard 'sh41'
Deployment has terminated due to previous errors.
The operation completed successfully
GDSCTL>

If we review the deployment log carefully and the answer is within the log. Sharding provided comprehensive and easy troubleshooting to identify possible reasons. It provides us clear reasons by following the more details as below.

SQL> select destination, output from all_scheduler_job_run_details
where job_name='SHARD_SH41_DBCREATE' 2 ;

DESTINATION
--------------------------------------------------------------------------------
OUTPUT
--------------------------------------------------------------------------------
ORA_C3
[FATAL] [DBT-11214] Automatic Memory Management is not a feasible option on the
system.
CAUSE: There is not enough free space on volume /dev/shm to allocate 1,744MB.

From the query, it provides us the output as the shared memory volume was not configured as per the requirement and that lead’s to sharding deployment failure.

Please note that this is one example, similarly, the sharding deployment provides us a detailed log which eases our job. 🙂

Review and Inputs By – Syed Jaffar Hussain