[Bug] Importing huge data is very CPU and memory intensive

[tinesoft]

Bug Report or Feature Request (mark with an x)

- [x] bug report -> please search issues before submitting
- [ ] feature request

Loading huge data file (200K+ entries) is very memory and CPU intensive (almost 100% of cpu used). Make sure that the streaming of each line/row being read is effective.

Tested on a Docker env, capped to 300M of memory and 1 CPU

Spring Versions?

- `spring boot version`: all
- or `spring`, 'spring-test`, and `spring-data-elasticsearch` versions: all

Elasticsearch server Version?

6.7.0

OS Version?

Linux (CentOS)

Repro steps

Import a huge dataset (200K+ entries), on a docker env, capped to 300M and 1 CPU for example.

The log given by the failure

Desired functionality

Mention any other details that might be useful