I have a few questions on rollup jobs. Mind that we are new to ES so please bear with me.
1.I wanted to understand the effect of settings certain index fields to index: false when there are certain rollup jobs that aggregate against those fields.
Example: field: cpu_time set to index = false. Can I create a rollup job that aggregates to average cpu_time per hour?
Can I set field names for rollup indices? Like in the example above. Can I set the name of the average cpu_time somehow.
Not indexing a field means it won't be searched against, which won't impact a rollup.
Good question! You could try specifying a mapping (or template) and then applying an alias. I haven't tried it though, so that's more an educated guess.
Thanks a lot for taking the time to answer my questions!
Let me elaborate a bit on my second question and explain why an alias won't work. I have two indices of raw data. I want to aggregate data from these indexes into a roll-up index.
Example:
Index 1: pageviews (website, page, timestamp)
Index 2: purchases (website, product, timestamp)
These above indexes have hundreds of thousands of events a day
I want to create a roll-up with one document per website per dat
The complexity for me comes because the total_pageviews and total_purchases fields are derived using the same formula on the same field but on different indexes (count of timestamp). If I am not able to name them differently like in my example one will end up overwriting the other
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.