jina.serve.runtimes.head.request_handling module#

class jina.serve.runtimes.head.request_handling.HeaderRequestHandler(logger, metrics_registry=None, meter=None, runtime_name=None)[source]#

Bases: MonitoringRequestMixin

Class that handles the requests arriving to the head and the results extracted from the requests future.

Parameters:
  • metrics_registry (Optional[CollectorRegistry]) – optional metrics registry for prometheus. Used if we need to expose metrics from the executor or from the data request handler

  • runtime_name (Optional[str]) – optional runtime_name that will be registered during monitoring

async warmup(connection_pool, stop_event, deployment)[source]#

Executes warmup task against the deployments from the connection pool. :type connection_pool: GrpcConnectionPool :param connection_pool: GrpcConnectionPool that implements the warmup to the connected deployments. :type stop_event: Event :param stop_event: signal to indicate if an early termination of the task is required for graceful teardown. :type deployment: str :param deployment: deployment name that need to be warmed up.