Merge pull request #15271 from quinton-hoole/2015-10-07-kubemark

cluster/kubemark/config-default.sh

@@ -22,7 +22,7 @@
 GCLOUD=gcloud
 ZONE=${KUBE_GCE_ZONE:-us-central1-b}
 MASTER_SIZE=${MASTER_SIZE:-n1-standard-4}
-NUM_MINIONS=${NUM_MINIONS:-100}
+NUM_MINIONS=${NUM_MINIONS:-4000}
 MASTER_DISK_TYPE=pd-ssd
 MASTER_DISK_SIZE=${MASTER_DISK_SIZE:-20GB}
 REGISTER_MASTER_KUBELET=${REGISTER_MASTER:-false}

hack/jenkins/e2e.sh

@@ -645,6 +645,21 @@ case ${JOB_NAME} in
     : ${USE_KUBEMARK:="true"}
     : ${NUM_MINIONS:="10"}
     ;;
+
+  kubernetes-kubemark-gce-scale)
+    : ${E2E_CLUSTER_NAME:="kubernetes-kubemark-scale"}
+    : ${E2E_NETWORK:="kubernetes-kubemark-scale"}
+    : ${PROJECT:="kubernetes-scale"}
+    : ${E2E_UP:="true"}
+    : ${E2E_DOWN:="true"}
+    : ${E2E_TEST:="false"}
+    : ${USE_KUBEMARK:="true"}
+    : ${NUM_MINIONS:="40"}
+    : ${MASTER_SIZE:="n1-standard-32"}  # Note: not available in all zones
+    : ${MINION_SIZE="n1-standard-16"}   # Note: can fit about 17 hollow nodes per core
+    #                                           so NUM_MINIONS x cores_per_minion should
+    #                                           be set accordingly.
+    ;;
 esac
 
 # AWS variables