All Products
Search

This Product

    Quick Audience:Data Import Troubleshooting

    Document Center

    Quick Audience:Data Import Troubleshooting

    Last Updated:Jun 22, 2026

    Common issues that cause data import failures and timeouts, with causes and recommended solutions.

    Self-service troubleshooting guide

    Independent troubleshooting

    Error message: No data in table xxx or partition xxx. Check the data source and try again.

    Potential cause:

    Data has not been generated on the data computing platform, or a write error occurred.

    Troubleshooting path:

    Log on to the data source and query the table or partition to check for data. Verify that data exists when the QA task is scheduled to run.

    Recommended solution:

    1. Add data to the upstream system and reschedule the task.

    2. If the upstream data generation time is irregular and QA is configured for scheduled updates, use triggered scheduling to connect the upstream system with QA.

    Error message: Special field type not supported. Ilegal data type - column 'xxx_column_name' type 'xxx' is not allowed.

    Troubleshooting path:

    Log on to the data source and query the table to identify fields with special column types, such as BOOLEAN, ARRAY, or MAP.

    MaxCompute syntax: `DESC table_name;`

    GaussDB syntax: `SELECT pg_get_tabledef('table_name');`

    Recommended solution:

    1. If the field is not needed, disable the column during import.

    2. If the column is required, convert it to a text-type field for storage.

    Troubleshooting with assistance

    Error message: SQL parsing exception. Parse exception - invalid token 'FROM'

    Potential cause:

    An SQL syntax parsing error occurred.

    Troubleshooting path:

    In Quick Audience, go to Configuration Management > Data Import > Table Structure Configuration. Check whether the original column list contains special characters, such as `/` or `-`.

    Recommended solution:

    1. Go to the data source and rename the column to follow standard database naming conventions. For example, a name should start with a letter and contain only letters, numbers, and underscores (_).

    If your issue is not described above, contact Quick Audience support.

    Error message: ① Table xxx does not exist. ② Table xxx already exists.

    Potential cause:

    The table could not be created, or it already exists.

    Troubleshooting path:

    Android Debug Bridge (ADB):

    1. Log on to ADB and query for CREATE, DELETE, DROP, INSERT, SELECT, UPDATE, or ALTER statements.

    2. Log on to the ADB console and check whether ADB is creating a backup snapshot during the scheduling process. Data Definition Language (DDL) statements are not allowed during the backup snapshot phase.

    For Hologres:

    1. Log on to the Hologres database and check whether the value of `show hg_experimental_force_sync_replay` is `off`.

    Recommended solution:

    For ADB:

    1. If you lack the required permissions, log on to the Data Management Service (DMS) platform to grant them to the account.

    2. Wait for the ADB backup to complete, and then run the QA scheduling task. Note: If the backup takes an excessive amount of time, submit a ticket to ADB support for assistance.

    For Hologres:

    1. By default, Hologres clusters perform asynchronous synchronization. To force synchronous metadata replication, run `alter database dbname set hg_experimental_force_sync_replay = on;` in the Hologres console. However, this may degrade Hologres performance.

    After running the command, open a new Hologres console and run `show hg_experimental_force_sync_replay;` to verify the change.

    For issues not described above, contact the Quick Audience support team.

    Error message: Quick Audience canceled the SQL request. canceling statement due to user request

    Potential causes:

    1. The load on the compute resource or analysis source is high.

    2. The SQL query did not return a result after a long execution time because of a large data volume, and QA actively closed the connection. The synchronous execution timeout is 3 hours.

    Troubleshooting path:

    1. Log on to the compute resource backend to check whether the load is too high.

    2. Check the executing SQL statement and verify that the data volume in the table is as expected.

    Recommended solutions:

    1. Ensure sufficient resources: To prevent tasks in different spaces from affecting each other, purchase separate compute resources for each space. This reduces dependencies, avoids resource contention, and prevents slow performance. As data volume and the number of spaces grow, regularly verify that compute and analysis resources are sufficient.

    2. Stagger execution times: If multiple spaces share the same compute and analysis resources, stagger the execution times of QA scheduling tasks in each space to avoid resource contention. If external systems also use these resources, stagger their usage times as well.

    If the issue persists, contact the Quick Audience support team for assistance.

    Error message: Scheduling task timed out

    Potential causes:

    1. The load on the compute resource or analysis source is high.

    2. The SQL query did not return a result after a long execution time because of a large data volume. The scheduling task timeout is 6 hours.

    Troubleshooting path:

    1. Log on to the compute resource backend to check whether the load is too high.

    2. Check the executing SQL statement and verify that the data volume in the table is as expected.

    Recommended solutions:

    1. Ensure sufficient resources: To prevent tasks in different spaces from affecting each other, purchase separate compute resources for each space. This reduces dependencies, avoids resource contention, and prevents slow performance. As data volume and the number of spaces grow, regularly verify that compute and analysis resources are sufficient.

    2. Stagger execution times: If multiple spaces share the same compute and analysis resources, stagger the execution times of QA scheduling tasks in each space to avoid resource contention. If external systems also use these resources, stagger their usage times as well.

    If your issue is not addressed above, contact Quick Audience support.

    Solutions for slow task execution

    1. Ensure sufficient resources: To prevent tasks in different spaces from affecting each other, purchase separate compute resources for each space. This reduces dependencies, avoids resource contention, and prevents slow performance. As data volume and the number of spaces grow, regularly verify that compute and analysis resources are sufficient.

    2. Stagger execution times: If multiple spaces share the same compute and analysis resources, stagger the execution times of QA scheduling tasks in each space to avoid resource contention. If external systems also use these resources, stagger their usage times as well.

    3. Use batch scheduling for multi-table import: Select multiple tables and configure them in a single scheduling task. This reduces the number of ID mapping operations and speeds up multi-table imports.

    Configure real-time monitoring for QA scheduling tasks

    • To monitor scheduling task status in real time, go to the Real-time Monitoring and Alerting module in Quick Audience at Configuration Management > Advanced Settings. You can configure real-time DingTalk alerts to detect task timeouts and failure reasons as they occur. To receive alerts for tasks that do not complete within the expected time, create a DingTalk alert group and configure a webhook bot.