Fix bug with kubernetes executor (#470)

Author: Kevin O'Riordan

config/airflow.cfg

@@ -782,7 +782,7 @@ tolerations =
 #   https://raw.githubusercontent.com/kubernetes-client/python/master/kubernetes/client/apis/core_v1_api.py
 # Note that if no _request_timeout is specified, the kubernetes client will wait indefinitely for kubernetes
 # api responses, which will cause the scheduler to hang. The timeout is specified as [connect timeout, read timeout]
-kube_client_request_args = {{"_request_timeout" : [60,60] }}
+kube_client_request_args = {"_request_timeout" : [60,60] }
 
 # Worker pods security context options
 # See: