Merge pull request #5298 from aserdean/ovn_ic_dpu_bash

ovnkube.sh: Add new overwriting options for the gateway options and kubernetes node name

Author: trozet

dist/images/ovnkube.sh

@@ -324,6 +324,16 @@ ovn_nohostsubnet_label=${OVN_NOHOSTSUBNET_LABEL:-""}
 # should be set to true when dpu nodes are in the cluster
 ovn_disable_requestedchassis=${OVN_DISABLE_REQUESTEDCHASSIS:-false}
 
+# external_ids:host-k8s-nodename is set on an Open_vSwitch enabled system if the ovnkube stack
+# should function on behalf of a different host than external_ids:hostname. This includes
+# all the components that belond in an ovnkube stack (i.e. NB DB, SB DB, ovnkube etc)
+# overwrite the K8S_NODE env var with the one found within the OVS metadata in this case
+ovn_k8s_node=$(ovs-vsctl --if-exists get Open_vSwitch . external_ids:host-k8s-nodename | tr -d '\"')
+if [[ ! -z $ovn_k8s_node ]]; then
+  echo "host-k8s-nodename is set, overriding K8S_NODE with $ovn_k8s_node"
+  K8S_NODE=$ovn_k8s_node
+fi
+
 # Determine the ovn rundir.
 if [[ -f /usr/bin/ovn-appctl ]]; then
   # ovn-appctl is present. Use new ovn run dir path.
@@ -1356,6 +1366,7 @@ ovn-master() {
     ${network_qos_enabled_flag} \
     ${ovn_enable_dnsnameresolver_flag} \
     ${nohostsubnet_label_option} \
+    ${ovn_stateless_netpol_enable_flag} \
     ${ovn_disable_requestedchassis_flag} \
     --cluster-subnets ${net_cidr} --k8s-service-cidr=${svc_cidr} \
     --gateway-mode=${ovn_gateway_mode} ${ovn_gateway_opts} \
@@ -1626,6 +1637,13 @@ ovnkube-controller() {
   fi
   echo "ovn_observ_enable_flag=${ovn_observ_enable_flag}"
 
+
+  ovn_stateless_netpol_enable_flag=
+  if [[ ${ovn_stateless_netpol_enable} == "true" ]]; then
+          ovn_stateless_netpol_enable_flag="--enable-stateless-netpol"
+  fi
+  echo "ovn_stateless_netpol_enable_flag: ${ovn_stateless_netpol_enable_flag}"
+
   echo "=============== ovnkube-controller ========== MASTER ONLY"
   /usr/bin/ovnkube --init-ovnkube-controller ${K8S_NODE} \
     ${anp_enabled_flag} \
@@ -2054,6 +2072,11 @@ ovnkube-controller-with-node() {
   fi
   echo "ovn_observ_enable_flag=${ovn_observ_enable_flag}"
 
+  ovn_stateless_netpol_enable_flag=
+  if [[ ${ovn_stateless_netpol_enable} == "true" ]]; then
+          ovn_stateless_netpol_enable_flag="--enable-stateless-netpol"
+  fi
+
   echo "=============== ovnkube-controller-with-node --init-ovnkube-controller-with-node=========="
   /usr/bin/ovnkube --init-ovnkube-controller ${K8S_NODE} --init-node ${K8S_NODE} \
     ${anp_enabled_flag} \
@@ -2399,8 +2422,15 @@ ovn-node() {
     wait_for_event ovs_ready
   fi
 
-  echo "=============== ovn-node - (wait for ready_to_start_node)"
-  wait_for_event ready_to_start_node
+  if [[ ${ovnkube_node_mode} == "dpu-host" ]] && [[ ${ovn_enable_interconnect} == "true" ]]; then
+    # ready_to_start_node checks for the NB/SB readiness state.
+    # This is not available on the DPU host when interconnect is enabled,
+    # because the DBs will run locally on the DPU
+    echo "skipping ready_to_start_node on DPU Host and when interconnect is true"
+  else
+    echo "=============== ovn-node - (wait for ready_to_start_node)"
+    wait_for_event ready_to_start_node
+  fi
 
   echo "ovn_nbdb ${ovn_nbdb}   ovn_sbdb ${ovn_sbdb}  ovn_nbdb_conn ${ovn_nbdb_conn}"
 
@@ -2578,12 +2608,6 @@ ovn-node() {
   fi
 
   if [[ ${ovnkube_node_mode} == "dpu" ]]; then
-    # in the case of dpu mode we want the host K8s Node Name and not the DPU K8s Node Name
-    K8S_NODE=$(ovs-vsctl --if-exists get Open_vSwitch . external_ids:host-k8s-nodename | tr -d '\"')
-    if [[ ${K8S_NODE} == "" ]]; then
-      echo "Couldn't get the required Host K8s Nodename. Exiting..."
-      exit 1
-    fi
     if [[ ${ovn_gateway_opts} == "" ]]; then
       # get the gateway interface
       gw_iface=$(ovs-vsctl --if-exists get Open_vSwitch . external_ids:ovn-gw-interface | tr -d \")

docs/features/hardware-offload/dpu-support.md

@@ -17,3 +17,39 @@ on the embedded CPU.
 Any vendor that manufactures a DPU which supports the above model should work with current design.
 
 Design document can be found [here](https://docs.google.com/document/d/11IoMKiohK7hIyIE36FJmwJv46DEBx52a4fqvrpCBBcg/edit?usp=sharing).
+
+## OVN Kubernetes in a DPU-Accelerated Environment
+
+The **ovn-kubernetes** deployment will have two parts one on the host and another on the DPU side.
+
+
+These aforementioned parts are expected to be deployed also on two different Kubernetes clusters, one for the host and another for the DPUs.
+
+
+### Host Cluster
+---
+
+#### OVN Kubernetes control plane related component
+- ovn-cluster-manager
+
+#### OVN Kubernetes components on a Standard Host (Non-DPU)
+- local-nb-ovsdb
+- local-sb-ovsdb
+- run-ovn-northd
+- ovnkube-controller-with-node
+- ovn-controller
+- ovs-metrics
+
+#### OVN Kubernetes component on a DPU-Enabled Host
+- ovn-node
+
+### DPU Cluster
+---
+
+#### OVN Kubernetes components
+- local-nb-ovsdb 
+- local-sb-ovsdb
+- run-ovn-northd
+- ovnkube-controller-with-node
+- ovn-controller
+- ovs-metrics