Bump up refresh_interval when we're indexing a large amount of data
Quick idea for fixing/improving our indexing speed: bump up refresh_interval
from 1s
to something more reasonable. We don’t need our updates to reflect in 1 second, but changing that value should improve our cluster indexing speed. We currently even have a suggestion in our documentation to set it to -1
(disable) when indexing a large instance.
Also, we may want to put this setting in our admin panel. Bumping up this value could help to achieve stability when cluster is struggling with writes.