Getting error in the PM-OI connector log

Products

• DX Operational Intelligence

Issue/Introduction

You are using DX Oi 24.1 on premise and integrating PM-OI connector.

Getting error in the PM-OI connector log.

[ERROR] 2026-02-04 06:41:20.890 [pool-6-thread-26] RemoteDataConnectionImpl - Error posting documents to Jarvis index(itoa_groups_capm): 503
Service Unavailable
[ERROR] 2026-02-04 06:41:20.890 [pool-6-thread-26] GroupTaskImpl - Unable to push 3794 groups to data sink.

[ERROR] 2026-02-04 09:03:51.021 [TASIngestionBatcher-2] InventoryTaskImpl - Failed ingesting inventory (iteration 2 of 17) to TAS for CAPC tenant id : default Error: 500: {"error":{"code":301,"message":"DATABASE_ERROR","traceId":"<ID>"}}

CAPC tenant id : default Error: 500: {"error":{"code":301,"message":"DATABASE_ERROR","traceId":"<trace ID>"}}

Environment

DX Oi 24.1 on premise

Resolution

Recycle of the on premise DX Oi helped resolve this issue.

Additional Information

Troubleshooting steps that can be taken:

Netops PM team suggest the following configuration adjustments:

You could decrease the size of the inventory batch in the config.xml file of the OI Connector and restart it. 
Out of the box the inventoryBatchSize is 1000 vertices, try reducing it to something small and restart the connector to see if it lets inventory payloads ingest. 

        <bean id="tasInventoryTask"
                class="com.ca.im.oinet.connector.task.inventory.TASInventoryTaskImpl">
                <!-- Data Source Inventory Update Interval (minutes) -->
                <constructor-arg index="0" value="60" />
                <!-- TAS inventory enable flag-->
                <constructor-arg index="1" value="true" />
                <property name="inventoryBatchSize">
                        <value>1000</value>
                </property>
                <property name="healthStatisticsReporter" ref="tasInventoryHealthReporter"/>
        </bean>
        
        
try 500 first. stop OI Connector, Clear all logs, set the value, start the OI Connector

 

OI Engineering team suggests the following from the OI side:

From the logs, we can see metric ingestion looks fine so there should not be any connectivity issues.
However, inventory data is failing. It looks like TAS is having some issues.

Check the TAS pod logs to see whether there are any issues. 

Also other issue that we see with index, connect to any OI pod and run following command and review the output. 

curl -X GET "http://jarvis-apis:8080/health" -H "accept: application/json;charset=UTF-8"