Add organization structure to watcher

[rukas]

Currently there is no way to organize the watcher list into a folder or tree structure. As the list of watches grows it would be nice to have a way to organize the watches so that they're easier to find.

We currently use watches to alert on events from the following scenarios:

• Using metricbeat for monitoring our servers, processes, and Windows services

• Using heartbeat for monitoring server / service / URL up time

• Using filebeat to gather logs and alert on the absence of the logs altogether as well as error level events and keyword events

• Using winlogbeat to gather security and application logs and alert on server / database login events and other application error events

For our team this would allow us to break out the watches into a structure like the following:

• Server Health

• Application Health

• Elasticsearch Cluster Health

• User Specific watches

[elasticmachine]

Pinging @elastic/es-core-infra

[bleskes]

@rukas sounds like a valid feature request to me, but I believe it is better to place it in the https://github.com/elastic/kibana repo where all things UI live. If it ends up requiring something from ES it will bubble up. Can you please open an issue there? Thanks.