robots.txt is served as text/html instead of text/plain

Issue Summary

robots.txt bundled in Jira is served with an incorrect MIME type which may impact the way search engines index and position instance's pages 

Steps to Reproduce

• Visit JIRA_BASE_URL/robots.txt

Expected results

The response's MIME type is text/plain

Actual Results

Note that it's rendering as HTML (MIME type is text/html)

Notes 

This behaviour is a result of changes introduced in Tomcat which was upgraded to version 8.5.40 in Jira 8.2.2. 

Workaround

As robots.txt is a static file, Content-Type header for the response from accessing it can be easily overridden using e.g. a reverse proxy 

Currently no workaround based on altering e.g. web.xml or server.xml (Tomcat config files) is known. If one is discovered, it will be posted here.