I managed to find a way around this in case anyone else is interested. Here's what I had to do:
-
Stop new index auto-creation
- Added
action.auto_create_index: false
to the user override settings for the node
- Added
-
Deleted the
metricbeat-7.4.0
index (I can afford to lose this data) -
Created a new
metricbeat-7.4.0-2019-11-06
index and assign it the alias
PUT /metricbeat-7.4.0-2019-11-06
{
"actions" : [
{ "add" : { "index" : "metricbeat-7.4.0-2019.11.06-000001", "alias" : "metricbeat-7.4.0", "is_write_index" : "true" }
}
]
}
- Re-enable auto-index creation (remove from user config)
The index I created quickly rolled over but I'm all good now. None of the metricbeat clients had to be touched to get back to a working state. Hoorah!
Cheers
Dave