Learning Elasticsearch

Managing time-based indices

Time series data is a very common Elasticsearch usage pattern. The data is time sensitive; we can only care about the recent data. For example, application logs, the logs beyond a certain time are not relevant anymore. The most recent logs are what we want to query. Since the old logs are not important anymore, to save the disk space and avoid unnecessary compute cycles, we have to remove the old data. The easiest way to deal with time series data is to group the indices based on a time interval so that the older indices can be deleted easily.

In this section, we will summarize the various APIs we discussed so far to manage time-based indexes. For example, we want to store application logs in day-based indices. ...