Creating and setting up data tiering
A data tier consists of several Elasticsearch nodes that have the same data role and usually run on similar hardware. Often, different hardware is configured for each tier; for example, the hot tier might use the most powerful and expensive hardware, while the cold or frozen tiers could utilize less expensive, storage-oriented hardware. Using data tiers is an efficient strategy for reducing hardware requirements in an Elasticsearch cluster while maintaining access to data and the ability to search through it. To illustrate, a single frozen node can keep up to 100 TB of data compared to 2 TB of data for a hot node.
However, there is a caveat: as data moves to colder tiers, query performance can decrease. This is expected since the data is less frequently queried.
Figure 1.16 – Elasticsearch data tiering
As we can see in Figure 1.16, there are four data tiers provided by Elasticsearch:
- Hot tier: This...
