Kibana version: 7.3.0 BC5

Elasticsearch version: 7.3.0 BC5

Server OS version: CentOS 7

Browser version: N/A

Browser OS version: N/A

Original install method (e.g. download page, yum, from source, etc.): rpm default dist

Describe the bug: 2 tests in the integration-test repo create an interval watch to email a Kibana dashboard PDF and PNG report (2 separate tests). On 7.3.0 BC4 and earlier both tests pass. But on 7.3.0 BC5 the second test always fails. I can switch the PDF and PNG reports and it's always the second test that fails.

These watches are sending the reports as email attachments but I don't think that is relevant to the problem based on the error message;

An example of the .watcher-history doc for one of these failures;

00:25:12.817 .watcher-history*={"took":67,"timed_out":false,"_shards":{"total":1,"successful":1,"skipped":0,"failed":0},"hits":{"total":{"value":1,"relation":"eq"},"max_score":0.6931472,"hits":[{"_index":".watcher-history-10-2019.07.19","_type":"_doc","_id":"watcher_report1563554002370_44c82301-867c-4638-a051-5c5090211b9c-2019-07-19T16:45:37.861919Z","_score":0.6931472,"_source":{"watch_id":"watcher_report1563554002370","node":"thvOhxZ2SpqAk0pC-aWvdQ","state":"executed","status":{"state":{"active":true,"timestamp":"2019-07-19T16:45:37.861Z"},"last_checked":"2019-07-19T16:45:37.861Z","last_met_condition":"2019-07-19T16:45:37.861Z","actions":{"email_admin":{"ack":{"timestamp":"2019-07-19T16:45:37.861Z","state":"awaits_successful_execution"},"last_execution":{"timestamp":"2019-07-19T16:45:37.861Z","successful":false,"reason":""}}},"execution_state":"executed","version":-1},"trigger_event":{"type":"schedule","triggered_time":"2019-07-19T16:45:37.861Z","schedule":{"scheduled_time":"2019-07-19T16:45:37.542Z"}},"input":{"none":{}},"condition":{"always":{}},"result":{"execution_time":"2019-07-19T16:45:37.861Z","execution_duration":4,"input":{"type":"none","status":"success","payload":{}},"condition":{"type":"always","status":"success","met":true},"actions":[{"id":"email_admin","type":"email","status":"failure","error":{"root_cause":[{"type":"exception","reason":"Watch[watcher_report1563554002370] reporting[test_report.pdf] Error response when trying to trigger reporting generation host[localhost], port[5601] method[POST], path[/api/reporting/generate/printablePdf], status[408]"}],"type":"exception","reason":"Watch[watcher_report1563554002370] reporting[test_report.pdf] Error response when trying to trigger reporting generation host[localhost], port[5601] method[POST], path[/api/reporting/generate/printablePdf], status[408]"}}]},"messages":[]}}]}}

with the important part being;
Error response when trying to trigger reporting generation host[localhost], port[5601] method[POST], path[/api/reporting/generate/printablePdf], status[408]"

On a local install, I kept Elasticsearch 7.3.0 BC5 and changed Kibana back to BC4 and the tests pass. Change it back to BC5 and they fail again so it really seems like this is caused by a change between those 2 BCs.

Steps to reproduce:

1. Starting with Kibana default distribution (mine is installed via rpm packages on CentOS 7)
2. We have Elasticsearch configured with an email address, security enabled, self-signed certificates but I'll try to reproduce without the email part.
3. We use the watcher API to create the 1st watch which has a 5 minute interval. Then we wait 5 minutes and start checking if the watch has been successfully executed. The we delete that watch.
4. We create the next watch with the only difference being the report generation url so we get a PNG instead of a PDF. wait 5 minutes, check until it's successful, and delete the watch.

Expected behavior: Both watches should work as they have before.

Screenshots (if relevant):

Errors in browser console (if relevant):

Provide logs and/or server output (if relevant): I haven't found any errors in either Kibana or Elasticsearch logs related to the timeout triggering report generation.

Any additional context: