kind infra, run etcd in memory

Currently we encounter bad performance of KIND cluster on DinD setup, we get 'etcdserver: timeout errors' that causes jobs to fail often. In such cases it is recommanded [1] to use in-memory etcd Running etcd in memory should improve performance and will make sriov provider more stabilized. [1] kubernetes-sigs/kind#845 Signed-off-by: Or Mergi <[email protected]>
ormergi · Nov 15, 2020 · afbc0f7 · afbc0f7
1 parent c2cf403
commit afbc0f7
Showing 1 changed file with 7 additions and 0 deletions.
diff --git a/cluster-up/cluster/kind/common.sh b/cluster-up/cluster/kind/common.sh
@@ -14,6 +14,7 @@ REGISTRY_NAME=${CLUSTER_NAME}-registry
 
 MASTER_NODES_PATTERN="control-plane"
 WORKER_NODES_PATTERN="worker"
+ETCD_DATA_DIR="/mnt/tmpfs/etcd"
 
 function _wait_kind_up {
     echo "Waiting for kind to be ready ..."
@@ -105,6 +106,12 @@ function prepare_workers() {
         "cpu-manager-policy": "static"
         "kube-reserved": "cpu=500m"
         "system-reserved": "cpu=500m"
+    kind: ClusterConfiguration
+    metadata:
+      name: config
+    etcd:
+      local:
+        dataDir: "$ETCD_DATA_DIR"
 EOF
     done
 }