This topic summarizes the FAQ about data import failure /timeout
Troubleshooting Manual
Complete self-examination
Error message: The xxx table contains no data or the xxx partition contains no data. Check the data source and try again.
Potential causes:
Data is not produced on the data computing platform or is written abnormally.
Troubleshooting Path:
Log on to the data source and query whether the table or partition has data. Check whether there is data during QA scheduling
Solution:
Upstream system supplementary data, rescheduling
If the upstream production data time is uncertain, QA is configured to update regularly. You can use trigger scheduling to connect the upstream system with QA.
Error message: Special field types do not support 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 indicate which fields are of special column types, such as BOOLEAN, ARRAY, and MAP.
MC syntax: DESC table name;
GUASSDB Syntax: SELECT pg_get_tabledef ('Table nay');
Solution:
If this field is not required, you can choose to disable this column during import
If the column must be required, it can be converted into a text type field column for storage.
Self-check + Troubleshooting
Error message: SQL parsing exception Parse exception-invalid token 'FROM'
Potential causes:
An SQL syntax parsing error occurred.
Troubleshooting Path:
On the [Quick Audience-Configuration Management-Data Import-Table Structure Configuration] page, check whether the original list in the table structure configuration contains the special character list, such as /,-, and other symbols.
Solution:
Go to the data source and change the column name to a column name that meets the database specifications (for example, a letter that starts with a letter and contains only alphanumeric_)
If you do not have the preceding issues, contact Quick Audience on duty for assistance.
The error message returned because the xxx table does not exist. The xxx table already exists.
Potential causes:
The table fails to be created or already exists.
Troubleshooting Path:
ADB:
Log on to AnalyticDB for PostDB and query whether the CREATE, DELETE, DROP, INSERT, SELECT, UPDATE, and ALTER operations are performed.
Log on to the ADB console and check whether ADB is backing up snapshots during scheduling. During the backup snapshot phase, DDL statements are not allowed.
HOLO:
Log on to the HOO database and check whether the value of the show hg_experimental_force_sync_replay is off.
Solution:
ADB:
If you do not have permissions, log on to the DMS platform and grant the corresponding permissions to the account.
After the ADB backup is completed, run the Quick Audience scheduling task (Note: If the backup time is too long, please submit an ADB ticket to help handle the backup problem at the time).
HOLO:
by default, holo clusters are synchronized asynchronously. you can execute the: alter database dbname set hg_experimental_force_sync_replay=on; statement in the holo console to forcibly wait for meta synchronization. however, the performance of holo is compromised.
After execution, restart a holo console and use SQL command: show hg_experimental_force_sync_replay; To check whether it has taken effect.
If you do not have the preceding problem, contact the on-duty colleague of Quick Audience for assistance in troubleshooting.
Error message: Quick Audience cancels SQL request canceling statement due to user request
Potential causes:
High load on computing or analysis sources
If you execute SQL statements for a long period of time due to a large amount of data, Quick Audience disconnects the connection. The synchronous execution timeout period is 3 hours.
Troubleshooting Path:
Log on to the computing source to check whether the backend is overloaded.
the background query is executing sql to check whether the data volume in the table is as expected.
Solution:
Ensure adequate resources: space tasks do not affect each other and computing resources can be procured separately. Reduce mutual dependencies to avoid resource preemption and slow running. With the increase of data volume and space, you need to check whether the computing source and analysis source resources are sufficient.
Staggered up time: Multiple spaces use the same computing source and analysis source resources. QA scheduling tasks for each space up time be staggered as much as possible to avoid resource preemption and slow running. If the external system of the customer uses the computing source and analysis source, the usage time must be staggered.
If you do not have the preceding problems, contact Quick Audience on-duty students for help.
Error message: The scheduling task timed out
Potential causes:
High load on computing or analysis sources
If a large amount of data is stored and the SQL statement is executed for a long time, the result is not calculated. The timeout period of the scheduling task is 6 hours.
Troubleshooting Path:
Log on to the computing source to check whether the backend is overloaded.
the background query is executing sql to check whether the data volume in the table is as expected.
Solution:
Ensure adequate resources: space tasks do not affect each other and computing resources can be procured separately. Reduce mutual dependencies to avoid resource preemption and slow running. With the increase of data volume and space, you need to check whether the computing source and analysis source resources are sufficient.
Staggering up time: Multiple spaces use the same computing source and analysis source resources. The scheduling tasks of QA for each space up time be staggered as much as possible to avoid resource preemption and slow running. If the external system of the customer uses the computing source and analysis source, the usage time must be staggered.
If you do not have the preceding problems, contact Quick Audience on-duty students for help.
Solution to Slow Task Running
Ensure adequate resources: space tasks do not affect each other and computing resources can be procured separately. Reduce mutual dependencies to avoid resource preemption and slow running. With the increase of data volume and space, you need to check whether the computing source and analysis source resources are sufficient.
Staggering up time: Multiple spaces use the same computing source and analysis source resources. The scheduling tasks of QA for each space up time be staggered as much as possible to avoid resource preemption and slow running. If the external system of the customer uses the computing source and analysis source, the usage time must be staggered.
Batch scheduling, multi-table import: In the scheduling task, select the table to select multiple tables to configure a scheduling task to reduce the number of idmaping times and. Accelerate multi-table import
Real-time detection configuration of QA scheduling tasks
If you want to know the status of scheduling tasks at a timely time, you can click Quick Audience-Configuration Management-Advanced Settings-Real-time Detection and Alerting to configure real-time DingTalk detection and alerting. This allows you to detect the timeout and failure causes of the tasks in real time. For example, you must create a DingTalk alert group and configure a webhook chatbot.