Dear colleagues,
We are developing a standalone application that is batch inserting record in CSV tables. It uses HXTT CSV and Hibernate (with the com.hxtt.support.hibernate.HxttTextDialect). The data model is complex with many relationships between entities. We have observed that when the CSV files starts growing (around 10000 lines) the performance decreases a lot.
After 40 minutes of execution, the time taken to perform a flush() is multiplied by 10.
Could you help us with this issue?
Kind regards,
Carmen
|
CSV file need to parse first. You can use delayedClose=120 for connection property.
delayedClose: Indicates the delayed seconds for close transaction. That option is used to avoid frequent close/open table operations for following sqls. Automatic temporary index is disabled when delayedClose<=60s. You can use 0~120 seconds. Default: 3.
|