I found out that add
elasticsearch.yml is not supported, but
request.type is ok.
uncaught exception in thread [main]
java.lang.IllegalArgumentException: unknown setting [network.breaker.inflight_requests.type] did you mean any of [network.breaker.inflight_requests.limit, network.breaker.inflight_requests.overhead]?
Reading the code found that only type setting of filed data and request are registered in ClusterSettings.
What are the considerations. Or is it just because of negligence?
It does look like an oversight to me. However, AFAIK none of the
network.breaker.*.type settings are documented and I don't think it makes much sense for an end-user to change them.
In fact, I am just confused that one circuit breaker can be configured and the other cannot be configured. Will there be any problems if I register this setting in ClusterSetting?
And, there are currently some settings in the ES source code without corresponding documentation. eg.
indices.id_field_data.enabled. So for what purpose were they developed if they are not encouraged to use?
There are quite a few settings that are deliberately undocumented and only intended for internal use, e.g. in tests, or to switch on some internally-controlled behaviour.
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.