Network Function Instantiation Fails due to Null Termination Character in the Database.
book
Article ID: 325398
calendar_today
Updated On:
Products
VMwareVMware Telco Cloud Automation
Issue/Introduction
Symptoms: In environments that have been upgraded from 2.2.0 to 2.3.0 or 2.3.0.1, network functions may fail during workflow instantiation. This can be confirmed by checking the upgrade log located in /opt/vmware/upgrade.log and searching for the following SQL exception: ERROR: unsupported Unicode escape sequence Detail: \u0000 cannot be converted to text.
This happens if data in MongoDB prior to the TCA upgrade contains a null termination character that is not allowed in PostgreSQL. In Unicode this character is represented as \u0000.
Resolution
Resolved in TCA 2.3.1.
Workaround:
Procedure For Environments Migrated to TCA 2.3
1. Download the recover_recipe_data.pyscript. 2. SSH as admin into the TCA Manager appliance observing the issue and switch user to root. 3. Copy the recover_recipe_data.py script into the /home/admin directory of the TCA Manager. 4. Restart the postgress-tca-config service to re-populate the built-inside recipes/ingredients. systemctl restart postgres-tca-config 5. Run the script to repopulate the recipe from the CSARs in objectstore.files collection: python3 recover_recipe_data.py
Procedure For TCA 2.2 Environments
1. Download the clean_mongo.py script. 2. SSH as admin into the TCA Manager appliance observing the issue. 3. Copy the clean_mongo.py script into the /home/admin directory of the TCA Manager. 4. Run the script to clean the mongo collection “SiteLevelManagementRecipes”: python3 clean_mongo.py 5. SSH as admin into the TCA-CP appliance observing the issue. 6. Copy the clean_mongo.py script into the /home/admin directory of the TCA-CP. 7. Run the script to clean the mongo collection “SiteLevelManagementRecipes”: python3 clean_mongo.py
Additional Information
Impact/Risks: TCA 2.2.0, 2.3.0 and 2.3.0.1. Does not impact 3.0.