We are using time series data that comes into @ 10Gbytes / day rate. Our query response times increase dramatically if user wants to increase the the historical perspective they want to achieve. What are some of the strategies to limit the queries that make processor cores peaking trying to query all shards in the system?
If you know which reports you need, you can pre-aggregate your data. The problem you are seeing is caused by there being too many docs that you are aggregating on the fly, at read-time.
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.