Purpose

The purpose of this article is to document the steps sometimes requested by Customer Support to troubleshoot issues with the Database Connector.

High Level Overview

The Database Harvester interacts with the source database via the built-in Analyze "Database Metadata" node. It is possible to generate an extract containing information needed to replicate the issue. Analyze admins can perform these steps by default.

ℹ Note: non-admins can be given access to perform these steps via Settings > Privacy and Security and checking the boxes "Allow non-admin users to download the following logging information: Node Logs, Data.

Pre-Requisites

• If the Database Connector data flow executed via a schedule, schedule retention settings must not have already removed or cleaned temporary data for the execution
• If the Database Connector data flow executed while editing the data flow, run retention settings for interactive runs must not have already removed temporary data

Steps to Produce Comprehensive Logs

1. If executed via a schedule, locate the run under Runs. In the right-side information panel, click the hyperlink next to "Data Flow:" If the execution was interactive, open the data flow and move to step 2.
2. Double-click on the node, "Database Harvester nnn" where nnn represents the version number of the harvester
3. Double-click the "Capture" composite
4. Click on the Database Metadata node
5. In the right-side information panel, click "View Node Log"
6. In the popup window, select "Download"
7. Check all boxes (e.g. 'Diagnostics', 'Input/Output Metadata', 'Input/Output Data', 'Data Flow with Dependencies') and click Download
8. The file generated may be large. Work with your support resource to get the file transferred to Precisely for review. Note: only metadata is extracted by this node; no database records will be included. Passwords stored in the dataflow will not be included, either.

Steps to Provide DDL Object Definitions Only

In the event that the full extract is too large to generate or send, it's possible to troubleshoot some issues with only a subset of the comprehensive extract. This subset includes DDL only. 

1. If executed via a schedule, locate the run under Runs. In the right-side information panel, click the hyperlink next to "Data Flow:" If the execution was interactive, open the data flow and move to step 2.
2. Double-click on the node, "Database Harvester nnn" where nnn represents the version number of the harvester
3. Double-click the "Capture" composite
4. Click on the Database Metadata node pin labeled, "Object Definitions"
5. In the Data Viewer which appears at the bottom of the screen, click the popup menu and select, "Export to CSV".
6. The file generated may be large. Work with your support resource to get the file transferred to Precisely for review. Note: only metadata is extracted by this node.