How to Improve Performance When Using DQC Unification Process
Reported for version 8
1. Do not allocate too much of memory as the maximum heap memory size.This tip is applicable for any DQC processes computing a lot of data. The heap memory size is the allocated space in the system memory that is used by java to save the objects. If the heap size memory is too big, it can decrease the performance. The recommended size of the heap memory is two times of the minimum memory required to run a particular DQC plan.
2. Use different disks for the temporary folder, the repository files, and for the input and output data while using the file repository.
3. Use the Matching Measures option in Unification step always when it is possible. If there are a lot of unification rules, it is always recommended to use the DQC feature called Matching Measures in the Unification step. It can be used for approximative matching using functions like Levenshtein, editDistance etc.
4. Change the runtime.properties settings. The runtime.properties
file is in the [DQC_HOME]/runtime/
folder. The file is usually called via starting script using the runtime.properties parameter. The runtime.properties
file is well described by comments inside the file itself. It is possible to increase the performance by modifying this file. It is recommended to backup the runtime.properties
file before making any changes.
Related articles