Sourcepub async fn submit_request_to_hm<Req, Resp>( &self, endpoint: &str, req_schema: &FastSchema, resp_schema: &FastSchema, request: &Req, ) ->Result<Resp, KineticaError>
pub async fn submit_request_to_hm<Req, Resp>( &self, endpoint: &str, req_schema: &FastSchema, resp_schema: &FastSchema, request: &Req, ) ->Result<Resp, KineticaError>
where Req: KineticaRecord, Resp: DeserializeOwned,
Submits a request to the host manager of the cluster in use: once, with no fail-over.
The generated host manager endpoints (/admin/add/host, /admin/alter/host, /admin/remove/host, /admin/show/alerts and /admin/switchover) come through here. They name hosts and ranks within the cluster they are sent to, so a failure goes back to the caller rather than replaying the operation on another cluster.