We are expecting GB's of data daily basis in Elasticsearch. This DB's of data cannot be inserted into Oracle database directly because of our infra setup / capability to handle soo much inserts is also doubtful in oracle ... Elasticsearch is fast , as our data does not involve any transactions hence we have Elasticsearch for our application logging.
Any methodology you can suggest to be used to achieve ES - to - Oracle Table , in real time or say delay of 1 hour
You could use an ingest pipeline linked to the indices you want to extract data from to add a timestamp indicating when the data was indexed. You can then write a script/application to periodically extract data for a specific time period and index this into Oracle.
Oracle has our major data of application based on OLTP.
Elasticsearch has our logging data from application / middleware servers ...
what ever logging information success / failure we have is in ES. This data we have to use with user information and map some reports for business purpose. Hence we have this architecture in place.
You can enrich every log with the user information and store that in elasticsearch.
Then do your analysis in Kibana. In real time. Whatever the size of the data is and without needing to read again tons of data and write them elsewhere.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.