Help Center/
GaussDB/
Developer Guide(Distributed_V2.0-3.x)/
Development and Design Proposal/
Database Programming Specifications/
DDL
Updated on 2025-03-13 GMT+08:00
DDL
- [Rule] Do not perform DDL operations during peak hours. If DDL operations must be performed, control the DDL operation execution frequency (no more than one DDL operation per second).
- DDL locks are managed using a global regular lock table. DDL operations compete for access to the regular lock table, resulting in a large number of threads being blocked in the LockMgrLock wait event and a risk of thread pool resource exhaustion. This may trigger a switchover or the thread pool's anti-overload mechanism, impacting the service success rate.
- DDL operations require the system catalog cache for invalid objects, enabling other concurrent threads to detect changes in DDL objects. Moreover, other threads need to handle invalid messages, which increases thread processing pressure and poses a risk of increased CPU usage.
- The replay speed of DDL logs is significantly slower than that of DML logs. Frequent DDL operations may increase latency for both primary and standby nodes, triggering flow control. Therefore, it is recommended to control the DDL operation frequency during peak hours to avoid triggering flow control.
- [Rule] DDL and other service operations cannot be performed in the same transaction; DDL operations must be executed in an independent transaction.
Parent topic: Database Programming Specifications
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
The system is busy. Please try again later.
For any further questions, feel free to contact us through the chatbot.
Chatbot