Which open source software is better for ETL

1 Depending on the database environment, different tools may be used, but this is not limiting, most tools support different data environments. Sometimes several tools can be used together, combining the main and auxiliary;
odi and GoldenGate are optional under oracle ; sqldeveloper can also be used as a tool for loading data;
Oracle warehouse uses owb;
Optional vw (visual warehouse) under ibm;

Under mysql, kettle and GoldenGate are optional ;
Under mssql, a unique dts is used;

2 If you want to learn, you can familiarize yourself with kettle and GoldenGate first .

 

follow up

Thank you for your answer, I still have a question. The project uses mysql. Now there is a data extraction function to be implemented. Because I have never been exposed to ETL, I don’t know which software is suitable. If kettle is extracting data When it fails (interrupted for some reason, some data may be extracted), what is the processing strategy of kettle for the data that has not been extracted this time? ?
Does kettle have a mechanism for timing extraction? ?

 

follow up

1 Kettle is a lightweight tool that only provides technical fragments or basic components to implement functions. The strategy of each link needs to be designed and combined with each technical fragment to achieve. In other words, it does not provide a complete strategy for direct application.

2 The following approaches can be considered:
  (1) Use the timestamp comparison of the source and target tables, and press the timestamp to start again after an error;
  (2) A flat data file is generated from the daily source data, which is first inserted into the intermediate table (this table is one per day, and only saves the data operated on that day. During this processing period, primary summary, verification, etc. can be performed), and then from this intermediate table Fetch data and load it into the target table (this process can also complete additional functions such as re-summarization); errors can be designed to generate error records, and rollback is processed as the corresponding deletion of the records in the intermediate table of the day in the target table, and then reloaded;