Patent Grant
11792159
With the recent increase in cloud native applications, today there is more demand than ever for fast deployment of on-demand networking for connecting machines that are deployed in software defined datacenters (SDDC). It is desirable to provide auto-deployed networking solutions as many compute-cluster administrators do not have extensive knowledge of networking. However, for administrators who wish to adjust their system's networking, it is desirable to provide such administrators with the ability to configure and customize their network deployments.
Some embodiments of the invention provide a method for deploying network elements for a set of machines in a set of one or more software defined datacenters (SDDCs). The datacenter set is part of one availability zone in some embodiments. The method receives intent-based API (Application Programming Interface) requests, and parses these API requests to identify a set of network elements to connect and/or to perform services for the set of machines. In some embodiments, the API is a hierarchical document that can specify multiple different compute and/or network elements at different levels of compute and/or network element hierarchy.
The method performs automated processes to define a virtual private cloud (VPC) to connect the set of machines to a logical network that segregates the set of machines from other machines in the datacenter set. In some embodiments, the set of machines include virtual machines and container Pods, the VPC is defined with a supervisor cluster namespace, and the API requests are provided as YAML files.
The automated processes in some embodiments use templates or preconfigured rules to identify and deploy the network elements that implement the logical network without an administrator performing any action to direct the identification and deployment of the network elements after an API request is received. In some embodiments, the deployed network elements include a gateway router for the VPC (called VPC gateway router) to connect the VPC to a network of the datacenter set or to a network external to the datacenter set.
This gateway router in some embodiments is implemented by one physical router. In other embodiments, the VPC gateway router is a logical gateway router that is implemented by more than one physical router. For instance, in some embodiments, the logical router is implemented with two physical routers in active/active or active/standby configurations. Also, in some embodiments, the logical router includes (1) a distributed router that is implemented by several router instances on host computers and edge appliances, and (2) one or more central routers that are each implemented by an edge appliance.
The VPC gateway router is configured to communicate with a datacenter gateway router to connect the VPC gateway to another VPC gateway of another VPC in order to connect the two VPCs to each other. In some embodiments, the VPC routers of the two VPCs can be configured to communicate with each other directly. In some embodiments, the VPC gateway router is configured to perform source network address translation (SNAT) operation to translate internal network addresses used within the VPC to a set of one or more external source network addresses. The external source network addresses in some embodiments are addresses within the datacenter set, while in other embodiments, the external source network addresses are addresses to a network outside of the datacenter set. In some embodiments, the VPC gateway router does not perform SNAT operations for traffic exchanged between the VPC and another VPC that is deployed in the set of datacenters, while in other embodiments it performs such SNAT operations.
The VPC gateway is configured to perform load balancing operations, or to work with one or more load balancers to perform load balancing operations, on ingress and/or egress traffic entering and/or exiting the VPC. The load balancing operations in some embodiments are Layer 4 (L4) and/or Layer 7 (L7) load balancing operations. In some embodiments, at least a subset of the deployed machines is deployed through Kubernetes, and the L4/L7 load balancing operations implement the load balancing and ingress services of Kubernetes.
To deploy the network elements, the method of some embodiments uses one or more Custom Resource Definitions (CRDs) to define attributes of custom-specified network resources that are referred to by the received API requests. When these API requests are Kubernetes APIs, the CRDs define extensions to the Kubernetes networking requirements. In addition to the Kubernetes-based container Pods, the method of some embodiments deploys network elements to connect non-Kubernetes Pods and/or virtual machines (VMs). For the non-Kubernetes Pods and for VMs, the method of some embodiments uses virtual network interfaces (VIF) CRDs to specify virtual interfaces for connecting the non-Kubernetes Pods and the VMs to software forwarding elements (e.g., software switches) executing on host computers on which the non-Kubernetes Pods and VMs execute.
The method of some embodiments configures the logical network for the VPC to connect the deployed set of machines to each other. For instance, in some embodiments, the logical network includes one or more logical forwarding elements, such as logical switches, routers, gateways, etc. In some embodiments, the method defines a logical forwarding element (LFE) by configuring several physical forwarding elements (PFEs), some or all of which execute on host computers along with the deployed machines (e.g., VMs and Pods). The method in some embodiments configures the PFEs to implement two or more LFEs to connect two or more different subsets of deployed machines.
In some embodiments, the method configures the logical network by configuring two or more sub-networks for the logical networks. In some embodiments, each sub-network has one or more segments (with each segment implemented by a logical switch), connects a different subset of deployed machines, and provides a set of network elements that satisfy a unique set of connectivity requirements for that subset of machines. For instance, in some embodiments, a first sub-network (e.g., a first logical switch) connects the Kubernetes Pods, while a second sub-network (e.g., a second logical switch) connects VMs and/or non-Kubernetes Pods. Another example is having one sub-network for machines (e.g., VMs, Pods, etc.) that need high-bandwidth, and another sub-network for machines that can tolerate less bandwidth.
To deploy some or all of the unique sub-networks, the method of some embodiments uses CRDs to define the attributes of the sub-networks, so that these sub-networks can be referred to by the API requests processed by the method. These CRDs are referred to below as virtual network CRDs. An API that refers to a virtual-network CRD in some embodiments includes a network type value that can be used to define different types of virtual networks.
Some embodiments have the following four network types (1) a default type comprising one or more network segments for the logical sub-network with each network segment connecting to the VPC gateway router, (2) a public type comprising one or more network segments connecting to a separate gateway router for the logical sub-network that connects to the VPC gateway router and supports separate routing table and services for the logical network, (3) a high-bandwidth type comprising one or more segments that connect to a separate gateway router to directly connect to a router of the datacenter set, and (4) a private type comprising one or more segments for the logical sub-network that are isolated inside the VPC.
As mentioned above, the method of some embodiments uses VIF CRDs to define and deploy VIFs for non-Kubernetes Pods and for VMs. The VPC in some embodiments operates within a single routing domain that has an associated IP CIDR (Classless Inter-Domain Routing). For a VIF that belongs to a network segment, the method in some embodiments automatically allocates an IP address from an IP subnet of the VPC IP CIDR that is automatically allocated for the segment, unless the IP addresses from the currently allocated IP subnet are all used, in which case it automatically allocates a new IP subnet from the VPC IP CIDR and automatically allocates an IP address from the newly allocated IP subnet.
In some embodiments, the PFE ports to which the VIFs connects (i.e., with which the VIFs are associated) are configured with one or more hooks, i.e., with one or more functional calls, to one or more service engines that perform service operations on data messages (e.g., packets) passing through the ports. One such service engine in some embodiments is a load balancer that distributes the data traffic load among several sets of endpoint machines that are associated with a network address or a set of network addresses (e.g., a VIP address, or a set of associated IP addresses).
This load balancer distributes the data messages sent from a VIF's associated VM or Pod among the machines in the set of endpoint machines. Also, in some embodiments, this load balancer implements a distributed load balancer (i.e., a logical load balancer) with several other load balancers operating on several different host computers. This distributed load balancer in some embodiments implements one or more ClusterIP services that are specified by the Kubernetes API calls.
Some embodiments utilize an Endpoint Group CRD to define groupings of homogeneous or heterogenous sets of machines as one or more endpoint groups. In some embodiments, members of an endpoint group are specified by one or more selectors, which use different types of network constructs to identify the members of an endpoint group. The selectors in some embodiments include VIF selectors, Pod selectors, Namespace selectors, and Service selectors. Some embodiments use the endpoint group and VIF CRDs to define an endpoint group of the same- or different-type machines to provide a ClusterIP service that is specified by the Kubernetes API calls.
Some embodiments utilize security CRDs to define additional ways to specify security policies for the VPC. For instance, the method of some embodiments uses Security Policy CRD to define firewall rules that control traffic between VPC network endpoints, which can be defined by APIs referencing Endpoint Group CRDs. The endpoint groups in some embodiments can be specified as either source or destination for ingress or egress security rules (e.g., firewall rules). Such rules are enforced with firewall engines that have hooks defined in the ports connecting to the interfaces (e.g., VIFs) of VMs and container Pods.
In some embodiments, selectors (such as VIF selectors, Pod selectors, Service selectors, etc.) can be used to specify an AppliedTo field in an API to specify where a security policy defined through the Security Policy CRD should be enforced. Endpoint groups in some embodiments can also be used to specify locations for applying the security policies that are defined through a Security Policy CRD. The Security Policy CRD provides the convenience of dynamic grouping, since existing solutions require manually association of an endpoint instance to a security group. Security Policy CRDs in some embodiments also support membership across different virtual networks (defined though Virtual Network CRDs) and across different VPCs. Hence, administrators in some embodiments can use Security Policy CRDs to define ingress/egress rules for traffic between two VPCs.
Some embodiments use Admin Policy CRDs to define another type of security policies. These CRDs in some embodiments allow higher priority security policies (e.g., higher priority than the security policies defined through Security Policy CRDs) to be defined for the entire VPC. In some embodiments, Admin Policy CRDs control the north/south traffic between the VPC and an external network (e.g., from another VPC, from an external IP block, or from outside of the datacenter set in which the VPC is deployed). When a sub-network of a logical network has its own gateway, some embodiments use Admin Policy CRDs to define security policies for the gateway to enforce.
In some embodiments, the policies defined through the Admin Policy CRD are implemented through firewall rules that are enforced by the VPC gateway, or one or more firewalls used by the VPC gateway. This is in contrast in some embodiments to the security policies that are defined through the Security Policy CRDs, which are enforced by distributed firewalls executing on the host computers with the source or destination VMs and Pods.
Some embodiments use a virtual service object (VSO) CRD to expose a service (e.g., a middlebox service or an application tier, such as Webserver, AppServer, database server) provided inside of the VPC to machines outside of the VPC or to machines inside of the VPC. In some embodiments, a VSO API based on the VSO CRD maps a set of one or more L4 ports and a protocol to an endpoint group of machines for providing the service. The VSO API in some embodiments also specifies an internal scope and/or external scope for the virtual service, with the internal scope specifying that the virtual service is available to machines inside of the VPC, while the external scope specifies that the virtual service is available to machines outside of the VPC.
In some embodiments, the method automatically allocates an internal VIP address for the virtual service when the internal scope is specified, while automatically allocating an external VIP address for the virtual service when the external scope is scope. The method also automatically generates one or more load balancing rules for one or more load balancers to implement the virtual service. Each load balancing rule includes (1) a match criteria set that comprises the VIP, the port set, and protocol and (2) an action criteria set that comprises a set of one or more identifiers that identify machines in the endpoint group specified for the virtual service deployed with the VSO CRD.
For the external scope, the load balancing rules in some embodiments are implemented by the VPC gateway router, or a set of load balancers used by the VPC gateway router. For the internal scope, the load balancing rules in some embodiments are implemented by the VPC gateway route or a set of load balancers used by the VPC gateway router, and/or by a distributed load balancer implemented by load balancing engines executing on host computers along with the client VMs and Pods that use the virtual service. In some embodiments, the VSO API specifies one or more different types of load balancers for performing load balancing operations with respect to the virtual service for one or more different types of data message traffic. In some embodiments that use Kubernetes APIs, the VSO API can be used to define ClusterIP and Load Balancer service types as the designated load balancing type(s).
The preceding Summary is intended to serve as a brief introduction to some embodiments of the invention. It is not meant to be an introduction or overview of all inventive subject matter disclosed in this document. The Detailed Description that follows and the Drawings that are referred to in the Detailed Description will further describe the embodiments described in the Summary as well as other embodiments. Accordingly, to understand all the embodiments described by this document, a full review of the Summary, the Detailed Description, the Drawings and the Claims is needed. Moreover, the claimed subject matters are not to be limited by the illustrative details in the Summary, the Detailed Description and the Drawing.
The novel features of the invention are set forth in the appended claims. However, for purposes of explanation, several embodiments of the invention are set forth in the following figures.
In the following detailed description of the invention, numerous details, examples, and embodiments of the invention are set forth and described. However, it will be clear and apparent to one skilled in the art that the invention is not limited to the embodiments set forth and that the invention may be practiced without some of the specific details and examples discussed.
Some embodiments of the invention provide a network control system for deploying and managing network elements for a set of related machines in a set of one or more software defined datacenters (SDDCs). The datacenter set is part of one availability zone in some embodiments. The network control system receives intent-based API (Application Programming Interface) requests, and parses these API requests (API calls) to identify network elements to deploy for the set of machines. In some embodiments, the API is a hierarchical document that can specify multiple different compute and/or network elements at different levels of compute and/or network element hierarchy.
The network control system performs automated processes to define a virtual private cloud (VPC) to connect the set of machines to a logical network that segregates these machines from other machines in the datacenter set. In some embodiments, the set of machines include virtual machines (VMs) and container Pods, the VPC is defined with a supervisor cluster namespace, and the API requests are provided as YAML files.
To deploy the network elements, the network control system of some embodiments processes one or more Custom Resource Definitions (CRDs) that define attributes of custom-specified network resources that are referred to by the received API requests. When these API requests are Kubernetes API requests, the CRDs define extensions to the Kubernetes networking requirements. Some embodiments use the following CRDs: Virtual Network Interfaces (VIF) CRDs, Virtual Network CRDs, Endpoint Group CRDs, security CRDs, Virtual Service Object (VSO) CRDs, and Load Balancer CRD.
A VIF CRD in some embodiments is used to define a virtual interface to connect a non-Kubernetes container Pod or VM to software forwarding elements (e.g., software switches) executing on host computers on which the non-Kubernetes Pods and VMs execute. A Virtual Network CRD in some embodiments is used to define the attributes of a logical sub-network that is to connect a subset of the deployed machines. An Endpoint Group CRD is used to define attributes for grouping heterogeneous or homogeneous sets of machines (i.e., machines of the same or different types). Endpoint Group CRD provides a simple mechanism for defining a group of machines for accessing a service or compute operation, and/or for providing a service or compute operation.
Security CRDs are used to specify security policies for the VPC. For instance, some embodiments use Security Policy CRD to define security policies for traffic between VPC network endpoints, which can be defined with Endpoint Group CRDs. Another security CRD in some embodiments is an Admin Policy CRD, which can be used to define security policies for north/south traffic between the VPC and an external network (e.g., from another VPC, from an external IP block, or from outside of the datacenter set in which the VPC is deployed).
A VSO CRD is used to expose a service (e.g., a middlebox service or an application tier, such as Webserver, AppServer, database server) provided inside of the VPC to machines outside of the VPC or to machines inside of the VPC. In some embodiments, an API that refers to a VSO CRD map a set of one or more L4 ports and a protocol to an endpoint group of machines for providing the service. Some embodiments use a Load Balancer CRD to define the configuration for a load balancer service. In some embodiments, the API that refers to the VSO CRD also uses the Load Balancer CRD to specify a load balancer service to use for distributing the traffic load among the endpoint group of machines.
Several more detailed examples of some embodiments will now be described. In these examples, several of the deployed logical networks are Kubernetes-based logical networks that define virtual private clouds (VPC) for corporate entities in one or more datacenters. In some embodiments, the VPC is a “supervisor” Kubernetes cluster with a namespace that provides the tenancy boundary for the entity. These embodiments use CRDs to define additional networking constructs and policies that complement the Kubernetes native resources.
One of ordinary skill will realize that other embodiments define other types of networks for other types of entities, such as other business entities, non-profit organizations, educational entities, etc. In some of these other embodiments, neither Kubernetes nor Kubernetes-based Pods are used. For instance, some embodiments are used to deploy networks for only VMs and/or non-Kubernetes containers/Pods.
As used in this document, data messages refer to a collection of bits in a particular format sent across a network. One of ordinary skill in the art will recognize that the term data message is used in this document to refer to various formatted collections of bits that are sent across a network. The formatting of these bits can be specified by standardized protocols or non-standardized protocols. Examples of data messages following standardized protocols include Ethernet frames, IP packets, TCP segments, UDP datagrams, etc. Also, as used in this document, references to L2, L3, L4, and L7 layers (or layer 2, layer 3, layer 4, and layer 7) are references respectively to the second data link layer, the third network layer, the fourth transport layer, and the seventh application layer of the OSI (Open System Interconnection) layer model.
As shown, the control system 100 includes an API processing cluster 105, a software defined network (SDN) manager cluster 110, an SDN controller cluster 115, and compute managers and controllers 117. The API processing cluster 105 includes two or more API processing nodes 135, with each node comprising an API processing server 140 and a network controller plugin (NCP) 145. The API processing server receives intent-based API calls and parses these calls. In some embodiments, the received API calls are in a declarative, hierarchical Kubernetes format, and may contain multiple different requests.
The API processing server 140 parses each received intent-based API request into one or more individual requests. When the requests relate to the deployment of machines, the API server provides these requests directly to compute managers and controllers 117, or indirectly provide these requests to the compute managers and controllers 117 through an agent running on the Kubernetes master node 135. The compute managers and controllers 117 then deploy VMs and/or Pods on host computers in the availability zone.
The API calls can also include requests that require network elements to be deployed. In some embodiments, these requests explicitly identify the network elements to deploy, while in other embodiments the requests can also implicitly identify these network elements by requesting the deployment of compute constructs (e.g., compute clusters, containers, etc.) for which network elements have to be defined by default. As further described below, the control system 100 uses the NCP 145 to identify the network elements that need to be deployed, and to direct the deployment of these network elements.
In some embodiments, the API calls refer to extended resources that are not defined per se by Kubernetes. For these references, the API processing server 140 uses one or more CRDs 120 to interpret the references in the API calls to the extended resources. As mentioned above, the CRDs in some embodiments include the VIF, Virtual Network, Endpoint Group, Security Policy, Admin Policy, and Load Balancer and VSO CRDs. In some embodiments, the CRDs are provided to the API processing server in one stream with the API calls.
NCP 145 is the interface between the API server 140 and the SDN manager cluster 110 that manages the network elements that serve as the forwarding elements (e.g., switches, routers, bridges, etc.) and service elements (e.g., firewalls, load balancers, etc.) in an availability zone. The SDN manager cluster 110 directs the SDN controller cluster 115 to configure the network elements to implement the desired forwarding elements and/or service elements (e.g., logical forwarding elements and logical service elements) of one or more logical networks. As further described below, the SDN controller cluster interacts with local controllers on host computers and edge gateways to configure the network elements in some embodiments.
In some embodiments, NCP 145 registers for event notifications with the API server 140, e.g., sets up a long-pull session with the API server to receive all CRUD (Create, Read, Update and Delete) events for various CRDs that are defined for networking. In some embodiments, the API server 140 is a Kubernetes master VM, and the NCP 145 runs in this VM as a Pod. NCP 145 in some embodiments collects realization data from the SDN resources for the CRDs and provide this realization data as it relates to the CRD status.
In some embodiments, NCP 145 processes the parsed API requests relating to VIFs, virtual networks, load balancers, endpoint groups, security policies, and VSOs, to direct the SDN manager cluster 110 to implement (1) the VIFs needed to connect VMs and Pods to forwarding elements on host computers, (2) virtual networks to implement different segments of a logical network of the VPC, (3) load balancers to distribute the traffic load to endpoint machines, (4) firewalls to implement security and admin policies, and (5) exposed ports to access services provided by a set of machines in the VPC to machines outside and inside of the VPC.
The API server provides the CRDs that have been defined for these extended network constructs to the NCP for it to process the APIs that refer to the corresponding network constructs. The API server also provides configuration data from the configuration storage 125 to the NCP 145. The configuration data in some embodiments include parameters that adjust the pre-defined template rules that the NCP follows to perform its automated processes. The NCP performs these automated processes to execute the received API requests in order to direct the SDN manager cluster 110 to deploy the network elements for the VPC. For a received API, the control system 100 performs one or more automated processes to identify and deploy one or more network elements that are used to implement the logical network for a VPC. The control system performs these automated processes without an administrator performing any action to direct the identification and deployment of the network elements after an API request is received.
The SDN managers 110 and controllers 115 can be any SDN managers and controllers available today. In some embodiments, these managers and controllers are the NSX-T managers and controllers licensed by VMware Inc. In such embodiments, NCP 145 detects network events by processing the data supplied by its corresponding API server 140, and uses NSX-T APIs to direct the NSX-T manager 110 to deploy and/or modify NSX-T network constructs needed to implement the network state expressed by the API calls. The communication between the NCP and NST-T manager 110 is asynchronous communication, in which NCP provides the desired state to NSX-T managers, which then relay the desired state to the NSX-T controllers to compute and disseminate the state asynchronously to the host computer, forwarding elements and service nodes in the availability zone (i.e., to the SDDC set controlled by the controllers 115).
After receiving the APIs from the NCPs 145, the SDN managers 110 in some embodiments direct the SDN controllers 115 to configure the network elements to implement the network state expressed by the API calls. In some embodiments, the SDN controllers serve as the central control plane (CCP) of the control system 100.
Based on the received configuration data, the LCP agents 220 on the host computers 205 configure one or more software switches 250 and software routers 255 to implement distributed logical switches, routers, bridges and/or service nodes (e.g., service VMs or hypervisor service engines) of one or more logical networks with the corresponding switches and routers on other host computers 205, edge appliances 210, and TOR switches 215. On the edge appliances, the LCP agents 225 configure packet processing stages 270 of these appliance to implement the logical switches, routers, bridges and/or service nodes of one or more logical networks along with the corresponding switches and routers on other host computers 205, edge appliances 210, and TOR switches 215.
For the TORs 215, the TOR agents 230 configure one or more configuration tables 275 of TOR switches 215 through an OVSdb server 240. The data in the configuration tables then is used to configure the hardware ASIC packet-processing pipelines 280 to perform the desired forwarding operations to implement the desired logical switching, routing, bridging and service operations. U.S. Pat. Nos. 10,554,484, 10,250,553, 9,847,938, and 9,178,833 describe CCPs, LCPs and TOR agents in more detail, and are incorporated herein by reference.
After the host computers 205 are configured along with the edge appliances 210 and/or TOR switches 215, they can implement one or more logical networks, with each logical network segregating the machines and network traffic of the entity for which it is deployed from the machines and network traffic of other entities in the same availability zone.
As shown, the logical network 295 includes multiple logical switches 284 with each logical switch connecting different sets of machines and serving as a different network segment. Each logical switch has a port 252 that connects with (i.e., is associated with) a virtual interface 265 of a machine 260. The machines 265 in some embodiments include VMs and Pods, with each Pod having one or more containers.
The logical network 295 also includes a logical router 282 that connects the different network segments defined by the different logical switches 284. In some embodiments, the logical router 282 serves as a gateway for the deployed VPC in
In some embodiments, the centralized and distributed routing components connect through a logical switch 294 defined on the host computers 205 and the edge appliances 210. Also, in some embodiments, the logical router is implemented by a pair of logical nodes 299, with each node having centralized and distributed components. The pair of nodes can be configured to perform in active/active or active/standby modes in some embodiments. U.S. Pat. No. 9,787,605 describes the gateway implementation of some embodiments in more detail and are incorporated herein by reference.
As shown, the process 300 initially allocates (at 305) an IP subnet for the VPC. In some embodiments, the VPC is part of a supervisor cluster that is a single routing domain with a corresponding IP CIDR (Classless Inter-Domain Routing) that specifies a range of IP addresses internal to the availability zone. The allocated IP subnet in some embodiments is a subnet from this IP CIDR. In conjunction with the allocated IP addresses, the process in some embodiments allocates MAC addresses for virtual interfaces of the VPC.
Next, at 310, the process defines a gateway router for the VPC, and associates this gateway router with one or more of the allocated internal IP addresses. These associated addresses are addresses used by VPC switches and routers to reach the gateway.
In some embodiments, the VPC gateway router 282 is configured to connect the VPC with one or more gateway routers 405 of the availability zone (i.e., of the SDDC set that contains the VPC), in order to connect to a network external to the availability zone. Also, in some embodiments, the VPC gateway router 282 is configured to communicate with a datacenter gateway router 405 to connect the VPC gateway 282 to another VPC gateway of another VPC in order to connect the two VPCs to each other. In some embodiments, the VPC gateway router 282 is configured to forward packets directly to the gateway routers (not shown) of the other VPCs.
At 315, the process defines a segment of a logical network that it defines for the VPC and allocates a range of IP addresses to this segment. In some embodiments, this allocated range is a contiguous range, while in other embodiments it is not (i.e., the allocated IP addresses in these embodiments are not necessarily sequential). In some embodiments, the defined logical network segment includes a logical switch that is defined to connect a particular set of machines (e.g., VMs and/or Pods).
As mentioned above, the VPC logical network in some embodiments includes one or more logical forwarding elements, such as logical switches, routers, gateways, etc. In some embodiments, the SDN controller 115 implements the logical network by configuring several physical forwarding elements (such as software and hardware switches, routers, bridges, etc.) on host computers, edge appliances, and TOR switches to implement one or more logical forwarding elements (LFEs).
As further described below, the control system in some embodiments configures the PFEs to implement two or more LFEs to connect two or more different subsets of deployed machines that are in two or more sub-networks of the logical networks. In some embodiments, each sub-network can have one or more segments (with each segment implemented by a logical switch), connects a different subset of deployed machines, and provides a set of network elements that satisfy a unique set of connectivity requirements for that subset of machines. For instance, in some embodiments, a first sub-network (e.g., a first logical switch) connects the Kubernetes Pods, while a second sub-network (e.g., a second logical switch) connects VMs. In other embodiments, one sub-network is for VMs needing high-bandwidth, while another sub-network is for regular VMs. Additional examples are provided below.
Some sub-networks of a VPC's logical network in some embodiments can have their own sub-network gateway router. If the sub-network for the segment defined at 315 has such a sub-network router, the process 300 defines (at 320) the sub-network router for the logical network segment. As further described below, the sub-network routers in some embodiments can be configured to forward packets to the VPC gateway router (e.g., router 282) or the availability-zone router (e.g., router 405).
At 325, the process 300 configures the VPC gateway to connect to the availability-zone gateway and to perform source network address translation (SNAT) operations. For instance, in some embodiments, the process configures the VPC gateway 282 with forwarding rules for the gateway to use to forward certain data message flows to the availability-zone gateway 405. Also, in some embodiments, the VPC gateway router 282 is configured to perform SNAT operations to translate internal network addresses used within the VPC to a set of one or more external source network addresses, and to perform the reverse SNAT operations. The external source network addresses in some embodiments are addresses within the availability zone and/or are addresses to a network outside of the availability zone. In some embodiments, the VPC gateway router 282 does not perform SNAT operations for traffic exchanged between its VPC and another VPC that is deployed in the same availability zone, while in other embodiments, it performs such SNAT operations for some or all of the other VPCs.
In some embodiments, the VPC gateway 282 is configured to perform other service operations or to use service engines/appliances to perform such other service operations. For such embodiments, the process 300 configures (at 330) the VPC gateway to perform other service operations (e.g., load balancing operations) or to forward data messages to service engines/appliances to perform such other service operations. In some embodiments, the VPC gateway is configured to perform service operations and/or forward data messages to service engines/appliances to perform such service operations, but that this configuration is not part of the process 300 when the VPC gateway is deployed and instead is part of another process that is performed subsequently (e.g., upon deployment of machines in the VPC that perform certain services or applications).
In
The control system 502 deploys the default network 502 whenever it has to deploy Kubernetes Pods. For the default network 502, the control system 502 configures the network elements (e.g., the logical switches and routers that implement the default network, as well as the VPC gateway and the service nodes used by this network) to satisfy the Kubernetes networking requirements. For instance, the control system starts with one subnet from the IP CIDR of the VPC for the default network, assigns to the network interface of each Kubernetes Pod an IP address from this subnet and associates this network interface with a logical switch that is used to implement the default network 502. As more Pods are added, the control system automatically allocates extra subnets from the IP CIDR of the VPC when existing subnets are exhausted, and creates new logical switches to implement new segments of the default network for the new subnets.
Also, beyond the SNAT operations on the VPC gateway 282 described above, the control system defines firewall operations on the VPC gateway 282 to ensure that Kubernetes networking assumptions are satisfied, e.g., traffic across VPC that are within one routing domain are allowed to pass through and are not source network address translated. Also, in some embodiments, the VPC gateway 282 is configured to drop traffic that reaches the gateway from outside of the availability zone in order to isolate the VPC. The VPC administrator can define security rules to change some or all of these default settings in some embodiments.
For the Kubernetes load balancer type and Ingress type services, the control system configures the load balancer 415 of the VPC gateway 282 to provide these services for the default network 502 as well as other sub-networks. Also, in some embodiments, the control system 100 configures distributed load balancing engines on host computers to implement Kubernetes specified ClusterIP service, and configures distributed firewall engines on the host computers to implement Kubernetes specified Network Policies, again for the default network 502 as well as other sub-networks.
The control system 100 deploys the sub-networks 504-510 based four virtual network types that are defined through Virtual Network CRDs. To deploy some or all of the unique sub-networks, the network control system of some embodiments receives and processes APIs that refer to Virtual Network CRDs to define the attributes of such sub-networks. As mentioned above, the virtual-network CRDs in some embodiments include a network type value that defines a network type for the virtual networks deployed using these CRDs.
The high-bandwidth type virtual network 508 is the third virtual network in
The last virtual network is the private type virtual network 510. It includes one or more segments of the logical network that are isolated inside the VPC and connected to their own gateway router 530. Each of the sub-networks 502-510 in
As apparent from the four examples 504-510, each Virtual Network CRD in some embodiments defines the network and connectivity for a group of workloads sharing the same networking and connectivity requirements. Virtual network is advantageous because the workloads deployed in a VPC may have different needs for networking configuration. Front tier services, such as websites and streaming services, are always public facing. However, confidential services (e.g., certain database) are always isolated from the Internet. Hence, for such differing types of workloads, it is desirable to segregate the workloads based on different networking connectivity requirements. To provide VPCs with such subnetting functionality, some embodiments use the new VirtualNetwork (VNet) abstraction.
A sample API that refers to a virtual network CRD is as follows:
apiVersion: vmware.com/v1
Kind: VirtualNetwork
Metadata:
Spec:
Upon VNet creation event, NCP will realize the corresponding network resources and routing configuration based on the connectivity field in VirtualNetwork specification, and direct the SDN managers/controllers to deploy such a virtual network with the desired network elements (e.g., switches, routers, and/or gateways). As mentioned above by reference to
In some embodiments, an administrator can place one or more VMs onto a deployed virtual network by specifying the name of the virtual network in API call that request the deployment of the VM(s). The underlying network resources for a virtual network in some embodiments include one or more NSX-T segments, connected to the VPC gateway router (which in NSX-T is a Tier1 router) or a separate virtual network router (e.g., a separate NSX-T Tier1 router) that is deployed for the virtual network (like the gateway routers 520, 525 and 530). The workloads (e.g., VMs, Pods, etc.) being placed on the deployed virtual network, map to the ports on the segments. Each segment is allocated with a subnet from the private IP block. An individual virtual network in some embodiments is not restricted to a fixed CIDR. More segments are auto-created in some embodiments when the current ones are exhausted.
The virtual network design of some embodiments has the following benefits over existing public cloud solutions. They have simplified routing configuration, because unlike cloud VPC, VPC of some embodiments do not require users to dive into the complexity of routing configuration. The NCP 145 also auto-configures the routing tables to satisfy the basic connectivity requirements. The virtual network design of some embodiments also flexibly and easily supports various networking topologies. Apart from the most common subnets' mode, public and private, this virtual network design enables users to choose different underlying resources layout based on performance considerations, as mentioned above.
This design also offers a better user experience and efficiency in subnet usage. Common cloud VPCs are criticized for the inconvenience in subnet planning. For example, in AWS, the subnet must be created with a fixed CIDR, and cannot be expanded when applications scale up later. The VPC design of some embodiments has a subnet auto-scaling mechanism, which can achieve the most efficiency in IPAM utilization, by robustly redistributing the IPAM resources across different virtual networks.
Some embodiments use virtual network CRDs to not only specify virtual networks for VPCs, but also for guest clusters working conjunctively with the VPC. In some embodiments, one or more guest clusters are defined within the same namespace as the VPC. Such guest clusters (GCs) in some embodiments have several host computers on which workload Kubernetes Pods execute, while the VPC has several host computers on which workload non-Kubernetes Pods and VM execute. Defining guest clusters and virtual networks for VPCs and guest clusters are further described in concurrently filed U.S. Patent Publication 2021/0314239. This concurrently filed application is incorporated herein by reference.
The network control system of some embodiments is part of a workload control plane (WCP) that uses NSX-T as the network and security solution for Kubernetes Pods. Kubernetes requires Node network and Pod network to be able to reach each other. Hence, in some embodiments, the master VMs in the supervisor cluster need to be attached to NSX network. At the same time, guest cluster nodes are also attached to the same NSX network, so that supervisor cluster can manage guest cluster nodes as well.
In some embodiments, the NCP 145 that is deployed on master VM provides NSX networking for WCP. In some embodiments, NCP 145 listens to WCP API server events and then directs the NSX-T managers/controllers to create the NSX resources. In some embodiments, the master VM in supervisor cluster and node VM in guest cluster do not have existing object definitions in Kubernetes. Hence, to allow the NCP 145 to create network resources for these VMs, and to provide WCP information regarding the network status regarding the configured VM network, the network control system uses Kubernetes resource CRDs to allow different administrators for different deployments define their own resources. In some embodiments, the network control system defines a set of virtual network CRDs for WCP networking for VM and non-Kubernetes Pod.
On supervisor cluster, WCP in some embodiments creates VIF object for each master VM. For guest cluster creation, WCP in some embodiments creates VNet per cluster and creates VIF object per node VM. NCP 145 watches these virtual network events, and allocates NSX network resources for VIF objects, then reports status back regarding the VIF to the network control system. WCP is then notified about the networking realization result by a CRD status update.
To use virtual network CRD to define a virtual network for a guest cluster, the API that uses the Vnet CRD for a guest cluster includes a namespace scope. For each virtual network CRD, the NCP 145 has the SDN manager/controller deploy a GC gateway router (e.g., a T1 Router), a global SNAT IP for the guest cluster, and allocate a subnet under the deployed GC gateway router. NCP also provides the SNAT IP back as part of the CRD status. In some embodiments, the master VM in supervisor cluster uses VIFs to attach to NSX network, while the guest cluster uses the virtual network to define a cluster router and subnet. In some embodiments, nodes in guest cluster use VIF CRD to create a network resource that attaches to the virtual network.
The CRD 605 defines the name, group and version attributes for the virtual network in a specification section. It also defines a name attribute in a metadata section. The virtual network API 610 defines the virtual network name for a guest cluster in a metadata section. It also specifies the guest cluster default SNAT IP attribute and specifies several condition attributes in a status section. The default SNAT IP specifies that one IP address has to be allocated for the guest cluster. The SNAT operation in some embodiments is performed by a gateway that is deployed for a guest cluster as further described below.
As mentioned above, the network control system of some embodiments uses VIF CRDs to define and deploy VIFs for non-Kubernetes Pods and for VMs. The VPC in some embodiments operates within a single routing domain that has an associated IP CIDR. Separate from the private IP CIDR of the supervisor cluster, the VPC in some embodiments owns a public IP address range. In some embodiments, the control system 100 deploys Pods using Kubernetes Pod manifest, and VMs using a VM CRDs. For a Pod workload, a private IP will be allocated to the Pod. An administrator may also request the public IP address by specifying it in the Pod's or VM's specification.
For a VM, the administrator in some embodiments can create and specify N virtual interface in the VM CRD specification. The VIF CRD defines the virtual interface, which in some embodiments will be realized as an NSX-T logical port connected to the VPC segment. NSX-T in some embodiments also realizes a VNIC for the VM. An example VIF API is shown below. The type field defines the IP addresses allocated for the interface. If the type is set as private, it will only be assigned with a private IP. If the type is set to public, in addition to the private address, a public IP address will also be assigned to this interface.
The virtual network field in this API associates the VIF being deployed to a particular virtual network. In other words, the virtual network field in the VIF API reference a particular virtual network with which the VIF is associated. When the VIF API specifies no virtual network, the VIF is connected to the default network 504. In some of the embodiments that are used by WCP, the default network is a shared Namespace network.
apiVersion: vmware.com/v1
kind: VirtualInterface
metadata:
spec:
For a VIF that belongs to a network segment, the network control system in some embodiments automatically allocates an IP address from an IP subnet of the VPC IP CIDR that is automatically allocated for the segment, unless the IP addresses from the currently allocated IP subnet are all used, in which case it automatically allocates a new IP subnet from the VPC IP CIDR and automatically allocates an IP address from the newly allocated IP subnet. In addition to allocating IP addresses to the VIFs, the NCP in some embodiments also directs the SDN managers/controllers to allocate MAC addresses to VIFs, as further described below.
As shown, the process 700 initially defines (at 705) a new VIF (e.g., defines a new VIF object). It then determines whether the segment (e.g., of the sub-network of the VPC logical network) to which the VIF belongs has a free IP address to allocate. The VIF API in some embodiments includes a virtual network identifier specifying the Vnet to which the VIF belongs. When the API does not provide this value or it provides a default value, the process assigns the VIF to the default network 504. The virtual network that is identified (at 710) for the VIF is the sub-network to which the VIF belongs in some embodiments. Once the process identifies the sub-network for the VIF, it then determines (at 710) whether this sub-network still has free IP addresses to assign to the VIF from a current IP address range that has been assigned for the current segment of the identified sub-network.
If not, the process transitions to 715, where it defines a new segment for the VIF's virtual network and allocates a new IP range for the new segment. In some embodiments, each new segment for a virtual network is a different logical switch. From 715, the process transitions to 720. The process also transitions to 720 when it determines (at 710) that the current IP range that has been assigned to the VIF's virtual segment has not been exhausted (i.e., still has some unassigned IP addresses).
At 720, the process assigns an IP address to the VIF from the current IP range (when the IP range has not been exhausted) or from the new IP range (when the IP range was exhausted and a new one was defined for the VIF). Next, at 725, the process associates the VIF with the port of a software switch (i.e., creates a logical connection between the VIF and the port) executing on a host computer on which the VIF's associated VM executes. The software switch in some embodiments implements the logical switch that defines the segment of the virtual network to which the VIF belongs. After 720, the process ends.
As mentioned above, the NCP can also request the SDN managers/controllers to assign a public IP address to the VIF when the API that refers to that VIF request such an IP address. In some embodiments, the process 700 allocates a MAC address to the VIF along with the IP address(es) that it allocates to the VIF. Whenever the control system 100 allocates a new segment for a sub-network (e.g., the default network 502 or the Vnets 504-510) of the logical network, the control system not only allocates IP addresses and MAC addresses, but also configures the routing tables of the logical network routers to allow the data messages to be forwarded to and from the segments.
For instance, each time a new segment of a logical sub-network is created after the first segment for this sub-network has been created, the SDN managers/controllers configure the virtual routers 255 on the hosts that have machines belonging to the sub-network with routing records (e.g., next-hop forwarding records) that allow these virtual routers to forward data messages between different segments of one sub-network. These virtual routers perform the distributed router 296 of the gateway for the sub-network.
As mentioned above, some embodiments use the VPC gateway 282 as the gateway of the sub-network while others use their own gateways 520, 525 or 530. In some embodiments, the SDN managers/controllers also configure the centralized and distributed routing components of each sub-network to forward traffic from the sub-network to the gateway and from the gateway to the sub-network, in order to allow the data message flows to be exchanged between the sub-networks and between the sub-networks and other networks outside of the VPC's logical network.
In some of these embodiments, however, the private virtual network 510 is kept secluded from the other sub-networks and from the network outside of the VPC. In sum, each time a new segment is created (e.g., at 715), the SDN managers/controllers configure multiple routing records on multiple hosts, edge devices and/or TOR switches to ensure proper forwarding of data message flows to and from the network segment. The baseline forwarding and security policies that are defined for a VPC and a guest cluster will be further described below.
The NCP 145 in some embodiments can define service rules to perform service operations with the VIF-associated machines (e.g., VIF-associated VMs and/or non-Kube Pods) being at the client-end or the server-end of these service operations and/or server operations. For instance, when the VIF-associated machines are at the client-end of such operations, the NCP 145 in some embodiments defines hooks (i.e., function calls) at the software-switch ports connected to the VIFs so that distributed load balancing operations can be performed by load balancing engines executing on the same host computers as the VIFs, in order to distribute the data message flow egressing from the VIF-associated machines among several service nodes that perform the service operations.
In some embodiments, the NCP 145 defines such hooks to implement ClusterIP services defined in Kubernetes APIs (e.g., to distribute data message flows from one or more VIF-associated client machines among several service engines, appliances and/or machines), as well as other distributed load balancing services described in this document. Also, the NCP 145 in some embodiments defines such hooks to distribute data message flows from one or more VIF-associated client machines (e.g., Pods and/or VMs) among several different compute nodes (e.g., VMs, Pods, etc.) that perform the same distributed compute operations (e.g., execute the same server tier, such as the Webserver tier, appserver tier, database server tier, etc.). Hence, the end nodes among which the hooked-in load balancers distribute the data message flows can be service nodes or compute nodes.
The hooks are configured to direct to their respective load balancers ingress and/or egress traffic entering or exiting (provided by or provided to) the VIF-associated machines. Each load balancer 815 uses a set of load balancing rules (stored in an LB rule storage 820) to identify the set of end nodes 825 that should process data message flows entering or exiting the machines 830. In some embodiments, the load balancer then uses load balancing criteria (e.g., weight values for round robin selection of end nodes) to select an end node for each data message flow, and then forwards one or more data messages of a flow to the end node selected for that flow. As shown, the load balancing rules and criteria are stored in the LB rule storage 820 in some embodiments.
This selection of the load balancer for a data message flow can be stateful in some embodiments so that all the data messages of one flow are sent to different one end node 825. Alternatively, this selection can be stateless in some embodiments so that different data messages of the same flow are sent to different end nodes. Also, in some embodiments, the load balancer can be configured to only send the first data message or first few data messages of a flow to the end node machines.
The end nodes 825 in some embodiments can be service nodes in case of ingress or egress traffic, or destination compute nodes in case of egress traffic. The end nodes can be engines/machines on the same host computer 850 as the client VIF-associated machines and the load balancers, can be engines/machines on different host computers, or can be standalone appliances. In some embodiments, the end nodes are associated with a virtual network address (e.g., a VIP address) or a set of associated network addresses (e.g., a set of associated IP addresses). In some embodiments, the end nodes machines are Pods, VMs, and/or containers executing on Pods/VMs.
When forwarding data messages to end node machines residing on the same host computer, a load balancer 815 forwards the data messages through a software switch 855 on its host computer 850 in some embodiments. Alternatively, when forwarding data messages to end node machines not residing on the same host computer, the load balancer 815 forwards the data messages through its host's software switch 855 and/or software routers (not shown) and intervening network fabric.
The NCP 145 in some embodiments directs the SDN managers/controllers to configure hooks in the VIF-associated ports 810 for other middlebox service operations, such as firewall, intrusion detection, intrusion prevention, deep packet inspection, encryption, etc.
As shown, several firewall engines 915 on several host computers implement a distributed firewall 900. In some embodiments, multiple such firewall engines execute on the same host computer for multiple different VIF-associated machines for which distributed firewall operations are configured. To perform their firewall operations, the firewall engines use firewall rules stored in a firewall rule storage 920. The firewall rules are defined by the SDN managers/controllers at the behest of the NCP 145 in some embodiments. In some embodiments, the NCP 145 defines the distributed firewall operations to implement Network Policies services defined in Kubernetes APIs, as well as other firewall operations described in this document.
As mentioned above, the control system 100 in some embodiments can configure the service and forwarding rules to have the VIF-associated machines perform service- or server-end operations. To facilitate the use of VIF-associated machines as service machines or server machines, some embodiments use Endpoint Group CRDs. To simplify the inclusion of VIFs in endpoint groups, the metadata filed for VIFs in some embodiments includes a label field that can be used to specify one or more labels for the VIF. The label field in some embodiments can include one or more key value pairs, with each pair specifying a label and a value for the label.
To further illustrate such labels,
The API 1005 has a virtual network field that identifies guest-cluster-network1 as the virtual network with which the VIF is associated. This virtual network is the network to which the VIF and its associated machine (e.g., Pod) belong. The virtual network field is optional. When it is not provided, NCP will associate the VIF with the default network 504. For each VIF specified by an API, NCP will create a Port with allocated IP address and an allocated MAC address. NCP will report the realization result back as a CRD status, which the compute manager (e.g., WCP) receives and then configures the associated machine (e.g., VM or Pod) accordingly.
The VIF API has an interfaceID field for defining the VIF identifier (VIF_ID). It also has fields that specify that VIF has associated IP address, gateway, subnet mask, and a MAC address. It further specifies that the VIF has a logical switch with a logical switch ID. In some embodiments, these parameters are provided by the SDN managers/controllers 110/115 at the behest of the NCP 145, as described above. The VIF API has a set of metadata fields, which includes Labels. In the example of
As VIFs can be associated with non-Kubernetes Pods and VMs, labels allow endpoint groups to be defined to include heterogeneous type of groups of machines that can serve as the server- or service-nodes for data messages. Endpoint groups are dynamic grouping objects that can be created in some embodiments by using Endpoint Group CRDs and APIs. In some embodiments, the machines that can be associated through an endpoint group can be of the same type or of different types. For instance, in some embodiments, an endpoint group can be defined to include VIFs and Pods through the labels that are associated with VIFs and Pods. More specifically, in some embodiments, administrators can specify an endpoint group (i.e., can define members of the group) by specifying an IP address CIDR or a list of selectors. The selectors in some embodiments relate to different types of network elements. For instance, the selectors in some embodiments include VIF selectors, Pod selectors, Namespace selectors, and Service selectors.
In some embodiments, a selector can identify a group of network elements (VIFs, Pod, Namespace, Service) in terms of one or more of attributes of the network elements, including any label that has been specified for the network elements. Selectors provide an easy way for forming endpoint groups with dynamic membership for various kinds of workloads because the selectors can be defined in terms of attributes (e.g., labels) that are associated with sets of network elements that can dynamically change.
Such groups can then be used to perform services, security, and compute operations. For instance, as mentioned above, endpoint groups and VIFs can be used to define a set of machines of the same type or different types to provide service or compute operations (e.g., ClusterIP services) specified by the API requests.
Below is a sample API call that refers to an Endpoint Group.
apiVersion: vmware.com
kind: EndpointGroup
metadata:
spec:
At 1220, the definition of the endpoint group (i.e., the list of members of the endpoint group) is distributed to host computers and edge appliances that use this definition to resolve service and forwarding rules defined in terms of the endpoint groups. For endpoint groups that include VIFs and Kubernetes Pods, the endpoint group definition in some embodiments includes the endpoint group identifier plus a list of IP addresses of the VIFs and Pod interfaces that are members of the group. Each time new network elements are added with attributes that match the attributes of the endpoint group's selector(s), an updated list of members is distributed to the host computers and edge appliances.
As further described below, some embodiments define each member of an endpoint group in terms of a port address as well as an IP address. In such embodiments, the endpoint group's associated IP and port addresses can be used to define source and/or destination IP and port values of service rules (e.g., firewall rules or other middlebox service rules) that are processed by middlebox service engines to perform middlebox service operations.
In some embodiments, the endpoint groups illustrated in
When the firewall engine 1320 receives a data message, the firewall engine determines whether the flow identifier of the message matches the match classification attributes of one of its firewall rules, and if so, it performs the action of the highest priority matching rule. When the matching classification of a firewall rule includes an endpoint group identifier, the firewall engine uses a list of endpoint groups 1350 to resolve the endpoint group identifier to a list of member IP addresses, and then determines whether the received message's flow identifier has a header value that falls into this member IP list. As shown, the endpoint group list 1350 includes several endpoint group identifiers 1352, and a set of IP addresses 1354 of the VIFs and Pods that are members of that endpoint group.
When an endpoint group identifier is used to identify a destination IP field of a firewall rule, the firewall engine determines whether the destination IP address in the received message's header (in its flow identifier) is an IP address in the endpoint group's definition in the endpoint group list. If so, and if the rest of the data message's flow identifier matches the remaining match classification attributes of the firewall rule, the firewall engine 1320 then determines that the firewall rule matches the data message, and then performs a firewall operation (e.g., allow, drop, etc.) on the flow based on the matching firewall rule's action attribute.
In some embodiments, the endpoint group is identified by an IP address and a port address. In some such embodiments, the firewall engine 1320 uses an endpoint group list to resolve an endpoint group identifier in a firewall rule's match classification attribute set to IP and port addresses (e.g., destination IP and port addresses). It then compares the resolved IP and port addresses to the corresponding IP and port addresses of the received data message to determine whether the data message matches the firewall rule's match classification attributes.
In
The endpoint group identifier in some embodiments identifies an endpoint group identified in an endpoint group list 1360. As shown, the endpoint group list 1360 has multiple rows, with each row indexed by a different endpoint group identifier, representing a different endpoint group and specifying the attributes of the endpoint machines in the group. In this list, each endpoint machine is associated with a VIF or Pod interface, and is either specified by an IP address and a port value when the VIF is associated with a VM or a non-Kubernetes Pod, or by an IP address and a protocol when the interface is associate with a Kubernetes Pod. For a VM or a non-Kubernetes Pod, the port value in some embodiments includes or is associated with a protocol.
The load balancing engine uses the list of endpoint groups 1360 to resolve the endpoint group identifier to (1) a list of IP addresses of member VIF and Pods that are members of the endpoint group, and (2) a set of load balancing criteria (e.g., weight values) for the IP addresses in the list. For VMs and non-Kubernetes Pods, the endpoint group list in some embodiments provides with the IP address of each VIF, a port to use to access the service or compute provided by the endpoint machine associated with the VIF. For Kubernetes Pods, the endpoint group list 1360 in some embodiments provides the VIF IP along with a protocol (e.g., http) to use to access the service or compute operation provided by the endpoint machine associated with the VIF.
Based on the load balancing criteria, the load balancing engine then selects one of endpoint machine from the list (e.g., one IP address in the list), replaces the destination IP address of the data message with the selected IP address, replaces the destination IP port with the destination port associated with the selected IP address if applicable, and then provides the data message to a forwarding element (e.g., a software switch on its host computer) to forward to the newly specified destination IP address. These are the operations performed for VMs and non-Kubernetes Pods.
For Kubernetes Pods, the load balancing engine does not replace the destination IP port, but rather uses the protocol specified by the endpoint group list for the selected endpoint machine. Also, for VMs and non-Kubernetes Pods, the destination port of the received data message in some embodiments corresponds to the destination port used by all the end point machines for providing the service or compute operation. In such embodiments, the load balancer does not need to replace the destination port address before forwarding the data message to the selected endpoint machine.
For a received data message flow, the DLB 1405 in some embodiments identifies the members of the endpoint group by using endpoint group list. It then uses the load balancing criteria (e.g., based on weight values for round robin selection of member machines) specified in the endpoint group list for the identified endpoint group, in order to select a machine from the endpoint group for the flow. The DLB 1405 then replaces the destination VIP address and destination port address in the received data message flows with the resolved IP and port addresses of the endpoint machine selected for the flow, and then forwards the data message flow to this endpoint machine.
In the examples illustrated in
For Kubernetes deployments, some embodiments use the following workflow to deploy ClusterIP services for endpoint groups. Initially, the administrator creates the endpoint group to describe the selectors.
In some embodiments, an endpoint group is defined not only in terms of VIFs, but also in terms of ports/protocols. This is because one VM or Pod can have multiple containers and the port/protocol attributes are needed to differentiate the different containers (e.g., the different services or compute) operations performed on the VM or Pod. Ports are also needed to perform L4 load balancing on the endpoint groups. When the endpoint group definition does not provide a port, a default port is assigned to the endpoint group in some embodiments. In
The creation of the endpoint group triggers the workload controller to create a Kubernetes ClusterIP Service without selector.
After creating the ClusterIP Service without the selector, the workload controller then fills in the related Kubernetes endpoint information according to the selectors in the endpoint group.
This workflow can be used with other Kubernetes services, such as Ingress, IngressRoute, OpenShift Route, Istio VirtualService and other load balancing solutions. Most third-party solutions allow the use of a Service as destination and they can understand Kubernetes Endpoints. However, except the workload controller, other existing service deploying modules do not know the selector, and there is no means to obtain the selector criteria via Kubernetes API. This problem is solved by initially defining the endpoint group. This approach also provides a chance for NSX-T to expose better load balancing experience through virtual service objects and also allows WCP to integrate with third-party load balancers.
VIFs provide one way of exposing a single VM or non-Kubernetes Pod to an external or internal network. Virtual service objects expose multiple workloads with load redistribution to machines inside of the VPC network or outside of the VPC network. Specifically, some embodiments use a virtual service object (VSO) CRD to expose a load-balanced service or other workload (e.g., a middlebox service or an application tier, such as Webserver, AppServer, database server) provided by two or more machines inside of the VPC to machines outside of the VPC or to machines inside of the VPC. As endpoint groups can have different types of machines, VSOs are used to expose in a load-balanced manner different types of machines that perform the same service or compute operations.
An example of the mapping specified by a VSO API is provided in the following YAML code.
apiVersion: vmware.com/v1
kind: VirtualServiceObject
metadata:
spec:
A VSO in some embodiments is associated with one or more load balancers. In some embodiments, the VSO API can specify the set of load balancers associated with the VSO, but when the VSO API does not specify any load balancer, the control system 100 designates the load balancer(s) 415 of the VPC gateway 282 as the VSO's load balancers when the VSO is specified for an endpoint group in the VPC, or the load balancer(s) of the guest cluster gateway as the VSO' load balancer when the VSO is specified for an endpoint group in the guest cluster. In
The NCP 145 creates the load balancing object 1610 to represent the load balancing elements 1620, which can be a set of one or more centralized load balancers (such as the load balancers 415 of the VPC gateway) or a set of one or more distributed load balancing engine (such as load balancing engines 815). The NCP also creates an association (e.g., a reference) between the VSO 1605 and the load balancing object 1610. It also creates an endpoint group object 1615 when the group of endpoint machines 1630 is defined, and creates an association (e.g., a reference) between the VSO 1605 and the endpoint group object 1615 upon creating the VSO 1605.
The VSO API in some embodiments also specifies an internal scope and/or external scope for the virtual service. The internal scope specifying that the virtual service is available to machines inside of the VPC, while the external scope specifies that the virtual service is available to machines outside of the VPC. In some embodiments, the network control system automatically allocates an internal VIP address for the virtual service when the internal scope is specified, while automatically allocating an external VIP address for the virtual service when the external scope is defined as the VSO's scope. As indicated in an exemplary API discussed below by reference to
As shown in
For the external scope, the load balancing rules in some embodiments are implemented by the gateway router (e.g., VPC or GC gateway router), or a set of load balancers used by the gateway router. For the internal scope, the load balancing rules in some embodiments can be implemented by the gateway router (e.g., VPC or GC gateway router), or a set of load balancers used by the gateway router, and/or by a distributed load balancer implemented by load balancing engines executing on host computers along with the client VMs and Pods that use the virtual service. In some embodiments, the VSO API specifies one or more different types of load balancers for performing load balancing operations with respect to the virtual service for one or more different types of data message traffic. In some embodiments that use Kubernetes APIs, the VSO API can use ClusterIP and Load Balancer types as its designated load balancing type(s), or a loadbalancer defined through the load balancing CRDs.
At 1715, the process creates an association between the VSO and the endpoint group exposed by the VSO. This endpoint group is specified in the received API in some embodiments. In some embodiments, the NCP 145 creates an object to represent the endpoint group when it receives an API that defines the endpoint group. As endpoint groups can have different types of machines, VSOs can be used in some embodiments to expose in a load-balanced manner different types of machines that perform the same service or compute operations. To create the association between the VSO and the endpoint group, the process 1700 in some embodiments defines a reference in the VSO to the endpoint group.
Next, at 1720, the process 1700 identifies a set of one or more load balancers associated with the VSO. As mentioned above, the VSO API in some embodiments specifies the set of load balancers associated with the VSO, but when the VSO API does not specify any load balancer, the NCP designates gateway load balancer(s) (e.g., VPC or GC gateway load balancers) as the VSO's load balancers. At 1725, the process creates an association between the VSO object with one or more objects that the NCP has previously created to represent one or more load balancers that it identified as being related to the VSO.
At 1730, the process 1700 allocates an internal VIP address and/or an external VIP address depending on whether the VSO has an internal scope, an external scope, or both. The scope of the VSO is specified by the VSO API in some embodiments. An allocated VIP address is provided to the administrators and/or automated processes of the VPC for distribution to client-side machines that will use the VIP address to access the service or compute operations provided by the endpoint group exposed by the VSO.
After the allocation of the VIP address(es), the process 1700 defines (at 1735) a set of one or more load balancing rules for distributing the internal and/or external data message flows that are addressed to the allocated VIP address(es) among the members of the associated endpoint group. In some embodiments, the load balancing rules are generated with a set of load balancing criteria that control how the data message flows are distributed among the endpoint groups. Next, at 1740, the process 1700 distributes the set of load balancing rules to the set of load balancers identified at 1720, and then ends. The SDN controllers in some embodiments periodically collect statistics regarding the load on the endpoint group machines, and based on the collected statistics, modify the load balancing criteria from time to time, and distribute the modified load balancing criteria to the load balancers.
The VSO API also specifies three different set of ports/protocols to expose three different services, called serviceX, serviceY, and serviceZ. All three services will be accessible with the same VIP address, but along different sets of ports/protocols. ServiceX is exposed along TCP port 8080, serviceY is exposed along UDP port 5353, and serviceZ is exposed along TCP ports 80, 443, 8000-10000.
In the VSO API, (1) serviceX is associated with an endpoint group defined by Kubernetes, which is identified by a reference to a Kubernetes defined object, (2) serviceY with an endpoint group called EPA, and (3) serviceZ is associated with an endpoint group called epgB. As noted in the annotations in the code, an endpoint group can be identified in some embodiments by a unique identifier (uid) for the endpoint group; otherwise, it is identified by its name in the namespace specified for the endpoint group. Also, when the namespace is omitted for the endpoint group, the endpoint group's namespace is assumed to be the same namespace as the VSO.
Some embodiments use a Load Balancer CRD to define the configuration for a load balancer service. In some embodiments, settings defined through the Load Balancer CRD apply as default values to the L4/L7 routes bound on a particular load balancer service. A virtual network in some embodiments defines the bindings between the load balancer service and a gateway router (e.g., a VPC gateway router). By default, the load balancer service in some embodiments is placed on the gateway router cluster (e.g., the VPC gateway router cluster or a GC gateway router).
In some embodiments, an API that refers to a Load Balancer CRD to create a load balancer service can also refer to a virtual network CRD to specify that the load balancer is for that virtual network. As discussed above, a virtual service object serves as a linking object between a load balancer object and an endpoint object in some embodiments. The load balancer object in some embodiments can link a virtual network object (e.g., a virtual network) with a virtual service object and an endpoint group object.
Some embodiments utilize security CRDs to provide additional ways to specify security policies for the VPC or GC. For instance, the network control system of some embodiments uses Security Policy CRD to define security rules for filtering traffic between VPC network endpoints, which can be defined with endpoint group created by APIs referencing Endpoint Group CRDs. The endpoint groups in some embodiments can be specified as either source or destination for ingress or egress security rules (e.g., firewall rules). In some embodiments, such rules are enforced by distributed firewall engines executing on host computers (e.g., firewall engine 915) and/or edge appliances.
Endpoint groups in some embodiments can also be used to specify locations for applying the security policies that are defined through a Security Policy CRD.
In some embodiments, the appliedToGroup of the Security Policy API can be defined through the use of selectors. Examples of such selectors in some embodiments include VIF selectors, Pod selectors, Service selectors, etc. The Security Policy CRD provides the convenience of dynamic grouping, since existing solutions require manually association of an endpoint instance to a security group. Security Policy CRDs in some embodiments also support membership across different virtual networks defined through Virtual Network CRDs. Security Policy CRDs in some embodiments also support membership across different VPCs or different GCs. Hence, administrators in some embodiments can use Security Policy CRDs to define ingress/egress rules for traffic between two VPCs or two GCs. In some embodiments, the Security Policy CRDs allow administrators to define rules that specify the type of traffic to allow. In these embodiments, the APIs that refer to the Security Policy CRD are used to only define firewall rules that allow certain kind of traffic, as opposed to defining firewall rules that block other kind of traffic. Other embodiments, however, use the Security Policy CRD to define firewall rules that allow and/or deny traffic. In some of embodiments, the ingress/egress firewall rules created through Security Policy CRDs are firewall rules enforced by the distributed firewall engines, as mentioned above.
If an appliedTo field is specified by the API, the process 2400 identifies (at 2420) a set of enforcement locations for the security policy. Next, at 2425, the process formulates a firewall rule with the source and destination parameters identified at 2410 and 2415. This firewall rule has an appliedTo field set to the identifier(s) of the enforcement location(s) identified at 2420, if any such locations were identified. In the embodiments where the Security Policy CRDs allow administrators to define rules that specify the type of traffic to allow, the firewall rule specified at 2425 is defined with an Allow action parameter. The process 2400 distributes (at 2430) the specified firewall rule to distributed firewall engines that are to enforce this rule. When the appliedTo parameter is specified by the received Security Policy API, the distribution of the firewall rule is to a more limited set of distributed firewall engines that fall within the enforcement locations identified by the appliedTo field. After distributing the firewall rule, the process then ends.
The above-described
Some embodiments provide Admin Policy CRDs as a mechanism for defining another type of security policies. These CRDs in some embodiments allow high priority security policies to be defined for an entire VPC or GC. In some embodiments, the security policies defined through the Admin Policy CRDs control the north/south traffic between the VPC and an external network (e.g., from another VPC, from an external IP block, or from outside of the availability zone in which the VPC is deployed). When a sub-network of a cluster (e.g., a sub-network 504-510) has its own gateway, some embodiments use Admin Policy CRDs to define security policies for the gateway to enforce. Similarly, in some embodiments, Admin Policy APIs control the north/south traffic between the GC and machines outside of the GC.
In some embodiments, the policies defined through Admin Policy CRD are implemented through firewall rules that are enforced by the VPC or GC gateway, or one or more firewalls used by the VPC or GC gateway. This is in contrast in some embodiments to the security policies that are defined through the Security Policy CRDs, which are enforced by distributed firewalls executing on the host computers with the source or destination VMs and Pods.
In some embodiments, security rules defined through Admin Policy APIs have higher priority over the security rules defined through Kubernetes Network Policies and through Security Policy CRD. With Admin Policy CRDs, VPC administrators can explicitly define inbound/outbound traffic for a VPC from/to another VPC or an external IP block, or define inbound/outbound traffic for a GC from/to another GC or an external IP block. The network control system of some embodiments leverages Kubernetes RBAC (role-based authentication controls) to make sure only the VPC administrator users can create, update, or delete Admin Policy CRDs of the VPC, and only the GC administrator users can create, update, or delete Admin Policy CRDs of the GC.
At 2720, the process 2700 determines whether it has processed all security rules defined by the API. As illustrated by the YAML example in
Many of the above-described features and applications are implemented as software processes that are specified as a set of instructions recorded on a computer readable storage medium (also referred to as computer readable medium). When these instructions are executed by one or more processing unit(s) (e.g., one or more processors, cores of processors, or other processing units), they cause the processing unit(s) to perform the actions indicated in the instructions. Examples of computer readable media include, but are not limited to, CD-ROMs, flash drives, RAM chips, hard drives, EPROMs, etc. The computer readable media does not include carrier waves and electronic signals passing wirelessly or over wired connections.
In this specification, the term “software” is meant to include firmware residing in read-only memory or applications stored in magnetic storage, which can be read into memory for processing by a processor. Also, in some embodiments, multiple software inventions can be implemented as sub-parts of a larger program while remaining distinct software inventions. In some embodiments, multiple software inventions can also be implemented as separate programs. Finally, any combination of separate programs that together implement a software invention described here is within the scope of the invention. In some embodiments, the software programs, when installed to operate on one or more electronic systems, define one or more specific machine implementations that execute and perform the operations of the software programs.
The bus 2805 collectively represents all system, peripheral, and chipset buses that communicatively connect the numerous internal devices of the computer system 2800. For instance, the bus 2805 communicatively connects the processing unit(s) 2810 with the read-only memory 2830, the system memory 2825, and the permanent storage device 2835.
From these various memory units, the processing unit(s) 2810 retrieve instructions to execute and data to process in order to execute the processes of the invention. The processing unit(s) may be a single processor or a multi-core processor in different embodiments. The read-only-memory (ROM) 2830 stores static data and instructions that are needed by the processing unit(s) 2810 and other modules of the computer system. The permanent storage device 2835, on the other hand, is a read-and-write memory device. This device is a non-volatile memory unit that stores instructions and data even when the computer system 2800 is off. Some embodiments of the invention use a mass-storage device (such as a magnetic or optical disk and its corresponding disk drive) as the permanent storage device 2835.
Other embodiments use a removable storage device (such as a floppy disk, flash drive, etc.) as the permanent storage device. Like the permanent storage device 2835, the system memory 2825 is a read-and-write memory device. However, unlike storage device 2835, the system memory is a volatile read-and-write memory, such as random access memory. The system memory stores some of the instructions and data that the processor needs at runtime. In some embodiments, the invention's processes are stored in the system memory 2825, the permanent storage device 2835, and/or the read-only memory 2830. From these various memory units, the processing unit(s) 2810 retrieve instructions to execute and data to process in order to execute the processes of some embodiments.
The bus 2805 also connects to the input and output devices 2840 and 2845. The input devices enable the user to communicate information and select requests to the computer system. The input devices 2840 include alphanumeric keyboards and pointing devices (also called “cursor control devices”). The output devices 2845 display images generated by the computer system. The output devices include printers and display devices, such as cathode ray tubes (CRT) or liquid crystal displays (LCD). Some embodiments include devices such as touchscreens that function as both input and output devices.
Finally, as shown in
Some embodiments include electronic components, such as microprocessors, that store computer program instructions in a machine-readable or computer-readable medium (alternatively referred to as computer-readable storage media, machine-readable media, or machine-readable storage media). Some examples of such computer-readable media include RAM, ROM, read-only compact discs (CD-ROM), recordable compact discs (CD-R), rewritable compact discs (CD-RW), read-only digital versatile discs (e.g., DVD-ROM, dual-layer DVD-ROM), a variety of recordable/rewritable DVDs (e.g., DVD-RAM, DVD−RW, DVD+RW, etc.), flash memory (e.g., SD cards, mini-SD cards, micro-SD cards, etc.), magnetic and/or solid state hard drives, read-only and recordable Blu-Ray® discs, ultra-density optical discs, any other optical or magnetic media, and floppy disks. The computer-readable media may store a computer program that is executable by at least one processing unit and includes sets of instructions for performing various operations. Examples of computer programs or computer code include machine code, such as is produced by a compiler, and files including higher-level code that are executed by a computer, an electronic component, or a microprocessor using an interpreter.
While the above discussion primarily refers to microprocessor or multi-core processors that execute software, some embodiments are performed by one or more integrated circuits, such as application specific integrated circuits (ASICs) or field programmable gate arrays (FPGAs). In some embodiments, such integrated circuits execute instructions that are stored on the circuit itself.
As used in this specification, the terms “computer”, “server”, “processor”, and “memory” all refer to electronic or other technological devices. These terms exclude people or groups of people. For the purposes of the specification, the terms “display” or “displaying” mean displaying on an electronic device. As used in this specification, the terms “computer readable medium,” “computer readable media,” and “machine readable medium” are entirely restricted to tangible, physical objects that store information in a form that is readable by a computer. These terms exclude any wireless signals, wired download signals, and any other ephemeral or transitory signals.
While the invention has been described with reference to numerous specific details, one of ordinary skill in the art will recognize that the invention can be embodied in other specific forms without departing from the spirit of the invention. Several embodiments were described above that use certain CRDs. One of ordinary skill will realize that other embodiments use other types of CRDs. For instance, some embodiments use LB monitor CRD so that load balancing monitors can be created through APIs that refer to such a CRD. LB monitors in some embodiments provide statistics to reflect the usage and overall health of the load balancers. Also, while several example above refer to container Pods, other embodiments use containers outside of Pods. Thus, one of ordinary skill in the art would understand that the invention is not to be limited by the foregoing illustrative details, but rather is to be defined by the appended claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| PCT/CN2020/082785 | Apr 2020 | WO | international |
| Number | Name | Date | Kind |
|---|---|---|---|
| 7869439 | Ramberg et al. | Jan 2011 | B1 |
| 7890543 | Hunt et al. | Feb 2011 | B2 |
| 7912955 | Machiraju et al. | Mar 2011 | B1 |
| 8627442 | Ji et al. | Jan 2014 | B2 |
| 8683560 | Brooker et al. | Mar 2014 | B1 |
| 9152803 | Biswas et al. | Oct 2015 | B2 |
| 9225638 | Jain et al. | Dec 2015 | B2 |
| 9258312 | O'Neill et al. | Feb 2016 | B1 |
| 9531590 | Jain et al. | Dec 2016 | B2 |
| 9536077 | Bignon et al. | Jan 2017 | B2 |
| 9590901 | Tubaltsev et al. | Mar 2017 | B2 |
| 9594546 | Todd et al. | Mar 2017 | B1 |
| 9674275 | Engers et al. | Jun 2017 | B1 |
| 9755898 | Jain et al. | Sep 2017 | B2 |
| 9774537 | Jain et al. | Sep 2017 | B2 |
| 9813509 | Visser et al. | Nov 2017 | B1 |
| 9825810 | Jain et al. | Nov 2017 | B2 |
| 9935827 | Jain et al. | Apr 2018 | B2 |
| 9979641 | Jain et al. | May 2018 | B2 |
| 10095669 | Karppanen | Oct 2018 | B1 |
| 10122735 | Wohlgemuth | Nov 2018 | B1 |
| 10129077 | Jain et al. | Nov 2018 | B2 |
| 10135737 | Jain et al. | Nov 2018 | B2 |
| 10193977 | Ke et al. | Jan 2019 | B2 |
| 10205701 | Voss et al. | Feb 2019 | B1 |
| 10225137 | Jain et al. | Mar 2019 | B2 |
| 10257095 | Jain et al. | Apr 2019 | B2 |
| 10270796 | Veeraswamy et al. | Apr 2019 | B1 |
| 10320679 | Jain et al. | Jun 2019 | B2 |
| 10341233 | Jain et al. | Jul 2019 | B2 |
| 10496605 | Melnik et al. | Dec 2019 | B2 |
| 10516568 | Jain et al. | Dec 2019 | B2 |
| 10547521 | Roy et al. | Jan 2020 | B1 |
| 10594743 | Hong et al. | Mar 2020 | B2 |
| 10609091 | Hong et al. | Mar 2020 | B2 |
| 10613888 | Mentz et al. | Apr 2020 | B1 |
| 10628144 | Myneni et al. | Apr 2020 | B2 |
| 10652143 | Ravinoothala et al. | May 2020 | B2 |
| 10693782 | Jain et al. | Jun 2020 | B2 |
| 10708368 | Young et al. | Jul 2020 | B1 |
| 10725836 | Savenkov et al. | Jul 2020 | B2 |
| 10795909 | Bond et al. | Oct 2020 | B1 |
| 10812337 | Vaidya et al. | Oct 2020 | B2 |
| 10841226 | Mariappan et al. | Nov 2020 | B2 |
| 10860444 | Natanzon | Dec 2020 | B2 |
| 10942788 | Palavalli et al. | Mar 2021 | B2 |
| 10944691 | Raut et al. | Mar 2021 | B1 |
| 10951661 | Medan et al. | Mar 2021 | B1 |
| 10972341 | Mudigonda | Apr 2021 | B2 |
| 10972386 | Mackie et al. | Apr 2021 | B2 |
| 11074091 | Nayakbomman et al. | Jul 2021 | B1 |
| 11086700 | Myneni et al. | Aug 2021 | B2 |
| 11159366 | Gawade et al. | Oct 2021 | B1 |
| 11190491 | Kaciulis et al. | Nov 2021 | B1 |
| 11194483 | Dontu et al. | Dec 2021 | B1 |
| 11277309 | Vaidya et al. | Mar 2022 | B2 |
| 11316822 | Gawade et al. | Apr 2022 | B1 |
| 11436057 | Shen et al. | Sep 2022 | B2 |
| 11500688 | Liu et al. | Nov 2022 | B2 |
| 11570146 | Liu et al. | Jan 2023 | B2 |
| 11606254 | Liu et al. | Mar 2023 | B2 |
| 11671400 | Zhou et al. | Jun 2023 | B2 |
| 11671401 | Singh et al. | Jun 2023 | B2 |
| 20040098154 | McCarthy | May 2004 | A1 |
| 20050129019 | Cheriton | Jun 2005 | A1 |
| 20070244962 | Aadan et al. | Oct 2007 | A1 |
| 20100177674 | Aggarwal | Jul 2010 | A1 |
| 20100211815 | Mankovskii et al. | Aug 2010 | A1 |
| 20100246545 | Berzin | Sep 2010 | A1 |
| 20100293378 | Xiao et al. | Nov 2010 | A1 |
| 20110161988 | Kashyap | Jun 2011 | A1 |
| 20110194494 | Aso et al. | Aug 2011 | A1 |
| 20110282936 | Chekhanovskiy et al. | Nov 2011 | A1 |
| 20110289508 | Fell et al. | Nov 2011 | A1 |
| 20120117226 | Tanaka et al. | May 2012 | A1 |
| 20120150912 | Ripberger | Jun 2012 | A1 |
| 20120304275 | Ji et al. | Nov 2012 | A1 |
| 20130018994 | Flavel et al. | Jan 2013 | A1 |
| 20130019314 | Ji et al. | Jan 2013 | A1 |
| 20130125230 | Koponen et al. | May 2013 | A1 |
| 20130174168 | Abuelsaad et al. | Jul 2013 | A1 |
| 20130266019 | Qu et al. | Oct 2013 | A1 |
| 20130283339 | Biswas et al. | Oct 2013 | A1 |
| 20140036730 | Nellikar et al. | Feb 2014 | A1 |
| 20140129690 | Jaisinghani et al. | May 2014 | A1 |
| 20140164897 | Yucel et al. | Jun 2014 | A1 |
| 20140223556 | Bignon et al. | Aug 2014 | A1 |
| 20140237100 | Cohn et al. | Aug 2014 | A1 |
| 20140258479 | Tenginakai et al. | Sep 2014 | A1 |
| 20150063166 | Sif et al. | Mar 2015 | A1 |
| 20150081767 | Evens | Mar 2015 | A1 |
| 20150222598 | Koponen et al. | Aug 2015 | A1 |
| 20150249574 | Zhang | Sep 2015 | A1 |
| 20150263899 | Tubaltsev et al. | Sep 2015 | A1 |
| 20150263946 | Tubaltsev et al. | Sep 2015 | A1 |
| 20150317169 | Sinha et al. | Nov 2015 | A1 |
| 20150348044 | Smith | Dec 2015 | A1 |
| 20150379281 | Feroz et al. | Dec 2015 | A1 |
| 20160036860 | Xing et al. | Feb 2016 | A1 |
| 20160080422 | Belgodere et al. | Mar 2016 | A1 |
| 20160094454 | Jain et al. | Mar 2016 | A1 |
| 20160094457 | Jain et al. | Mar 2016 | A1 |
| 20160094650 | Rio | Mar 2016 | A1 |
| 20160094661 | Jain et al. | Mar 2016 | A1 |
| 20160217301 | Watanabe et al. | Jul 2016 | A1 |
| 20160241436 | Fourie et al. | Aug 2016 | A1 |
| 20160254964 | Benc | Sep 2016 | A1 |
| 20160269318 | Su et al. | Sep 2016 | A1 |
| 20160294612 | Ravinoothala et al. | Oct 2016 | A1 |
| 20160315809 | McMurry et al. | Oct 2016 | A1 |
| 20160335129 | Behera et al. | Nov 2016 | A1 |
| 20160337334 | Murr | Nov 2016 | A1 |
| 20170005923 | Babakian | Jan 2017 | A1 |
| 20170005986 | Bansal et al. | Jan 2017 | A1 |
| 20170031956 | Burk et al. | Feb 2017 | A1 |
| 20170063632 | Goliya et al. | Mar 2017 | A1 |
| 20170063782 | Jain et al. | Mar 2017 | A1 |
| 20170085561 | Han et al. | Mar 2017 | A1 |
| 20170093790 | Banerjee et al. | Mar 2017 | A1 |
| 20170171144 | Sagiraju et al. | Jun 2017 | A1 |
| 20170177394 | Barzik et al. | Jun 2017 | A1 |
| 20170195210 | Jacob et al. | Jul 2017 | A1 |
| 20170206034 | Fetik | Jul 2017 | A1 |
| 20170207963 | Mehta et al. | Jul 2017 | A1 |
| 20170258479 | Hänni et al. | Sep 2017 | A1 |
| 20170286698 | Shetty et al. | Oct 2017 | A1 |
| 20170317954 | Masurekar et al. | Nov 2017 | A1 |
| 20170332307 | Pan | Nov 2017 | A1 |
| 20170353351 | Cheng et al. | Dec 2017 | A1 |
| 20170366416 | Beecham et al. | Dec 2017 | A1 |
| 20170374106 | Hamou et al. | Dec 2017 | A1 |
| 20180063194 | Vaidya et al. | Mar 2018 | A1 |
| 20180083835 | Cole et al. | Mar 2018 | A1 |
| 20180089299 | Collins et al. | Mar 2018 | A1 |
| 20180123943 | Lee et al. | May 2018 | A1 |
| 20180131675 | Sengupta et al. | May 2018 | A1 |
| 20180167453 | Luo | Jun 2018 | A1 |
| 20180167487 | Vyas et al. | Jun 2018 | A1 |
| 20180183757 | Gunda et al. | Jun 2018 | A1 |
| 20180205605 | Mittal et al. | Jul 2018 | A1 |
| 20180234459 | Kung et al. | Aug 2018 | A1 |
| 20180248827 | Scharber et al. | Aug 2018 | A1 |
| 20180262424 | Roeland et al. | Sep 2018 | A1 |
| 20180287996 | Tripathy et al. | Oct 2018 | A1 |
| 20180295036 | Krishnamurthy et al. | Oct 2018 | A1 |
| 20180331885 | Raymond et al. | Nov 2018 | A1 |
| 20180359323 | Madden | Dec 2018 | A1 |
| 20190034237 | Siddappa et al. | Jan 2019 | A1 |
| 20190036868 | Chandrashekhar et al. | Jan 2019 | A1 |
| 20190042518 | Marolia et al. | Feb 2019 | A1 |
| 20190068544 | Hao et al. | Feb 2019 | A1 |
| 20190079751 | Foskett et al. | Mar 2019 | A1 |
| 20190102280 | Caldato et al. | Apr 2019 | A1 |
| 20190103992 | Cidon et al. | Apr 2019 | A1 |
| 20190132220 | Boutros et al. | May 2019 | A1 |
| 20190132221 | Boutros et al. | May 2019 | A1 |
| 20190132283 | Ballard et al. | May 2019 | A1 |
| 20190140895 | Ennis, Jr. et al. | May 2019 | A1 |
| 20190140921 | Xu et al. | May 2019 | A1 |
| 20190149512 | Sevinc et al. | May 2019 | A1 |
| 20190149516 | Rajahalme et al. | May 2019 | A1 |
| 20190149518 | Sevinc et al. | May 2019 | A1 |
| 20190171650 | Botev et al. | Jun 2019 | A1 |
| 20190173780 | Hira et al. | Jun 2019 | A1 |
| 20190229987 | Shelke et al. | Jul 2019 | A1 |
| 20190238363 | Boutros et al. | Aug 2019 | A1 |
| 20190238364 | Boutros et al. | Aug 2019 | A1 |
| 20190245757 | Meyer et al. | Aug 2019 | A1 |
| 20190273683 | Jiang et al. | Sep 2019 | A1 |
| 20190288947 | Jain et al. | Sep 2019 | A1 |
| 20190306036 | Boutros et al. | Oct 2019 | A1 |
| 20190306086 | Boutros et al. | Oct 2019 | A1 |
| 20190356693 | Cahana et al. | Nov 2019 | A1 |
| 20190384645 | Palavalli et al. | Dec 2019 | A1 |
| 20190386877 | Vaidya et al. | Dec 2019 | A1 |
| 20200065080 | Myneni et al. | Feb 2020 | A1 |
| 20200065166 | Myneni et al. | Feb 2020 | A1 |
| 20200073692 | Rao | Mar 2020 | A1 |
| 20200076684 | Naveen et al. | Mar 2020 | A1 |
| 20200076685 | Vaidya | Mar 2020 | A1 |
| 20200076734 | Naveen et al. | Mar 2020 | A1 |
| 20200092275 | Seed et al. | Mar 2020 | A1 |
| 20200112504 | Osman | Apr 2020 | A1 |
| 20200213366 | Hong et al. | Jul 2020 | A1 |
| 20200250009 | Jaeger et al. | Aug 2020 | A1 |
| 20200250074 | Zhang et al. | Aug 2020 | A1 |
| 20200252376 | Feng et al. | Aug 2020 | A1 |
| 20200301801 | Hegde | Sep 2020 | A1 |
| 20200314006 | Mackie et al. | Oct 2020 | A1 |
| 20200314173 | Pahwa et al. | Oct 2020 | A1 |
| 20200344120 | Pianigiani et al. | Oct 2020 | A1 |
| 20200366558 | Vaidya et al. | Nov 2020 | A1 |
| 20200374186 | Scott | Nov 2020 | A1 |
| 20200379812 | Ranjan et al. | Dec 2020 | A1 |
| 20200382556 | Woolward et al. | Dec 2020 | A1 |
| 20200401457 | Singhal et al. | Dec 2020 | A1 |
| 20200403853 | Garipally et al. | Dec 2020 | A1 |
| 20200403860 | Ewis et al. | Dec 2020 | A1 |
| 20200409671 | Mazurskiy | Dec 2020 | A1 |
| 20210004292 | Zlotnick et al. | Jan 2021 | A1 |
| 20210064442 | Alluboyina et al. | Mar 2021 | A1 |
| 20210099335 | Li | Apr 2021 | A1 |
| 20210165695 | Palavalli et al. | Jun 2021 | A1 |
| 20210200814 | Tal et al. | Jul 2021 | A1 |
| 20210243164 | Murray et al. | Aug 2021 | A1 |
| 20210273946 | Iqbal et al. | Sep 2021 | A1 |
| 20210306285 | Hirasawa et al. | Sep 2021 | A1 |
| 20210311803 | Zhou et al. | Oct 2021 | A1 |
| 20210314190 | Liu et al. | Oct 2021 | A1 |
| 20210314239 | Shen et al. | Oct 2021 | A1 |
| 20210314240 | Liu et al. | Oct 2021 | A1 |
| 20210314300 | Shen et al. | Oct 2021 | A1 |
| 20210314361 | Zhou et al. | Oct 2021 | A1 |
| 20210314388 | Zhou et al. | Oct 2021 | A1 |
| 20210328858 | Asveren et al. | Oct 2021 | A1 |
| 20210352044 | Asveren et al. | Nov 2021 | A1 |
| 20210365308 | Myneni et al. | Nov 2021 | A1 |
| 20210397466 | McKee et al. | Dec 2021 | A1 |
| 20210409336 | Talur et al. | Dec 2021 | A1 |
| 20220012045 | Rudraraju et al. | Jan 2022 | A1 |
| 20220035651 | Maurya et al. | Feb 2022 | A1 |
| 20220038311 | Shen et al. | Feb 2022 | A1 |
| 20220070250 | Baid et al. | Mar 2022 | A1 |
| 20220158926 | Wennerström et al. | May 2022 | A1 |
| 20220182439 | Zhou et al. | Jun 2022 | A1 |
| 20220200865 | Vaidya et al. | Jun 2022 | A1 |
| 20220210113 | Pillareddy et al. | Jun 2022 | A1 |
| 20220278926 | Sharma et al. | Sep 2022 | A1 |
| 20220303246 | Miriyala et al. | Sep 2022 | A1 |
| 20220311738 | Singh et al. | Sep 2022 | A1 |
| 20220321495 | Liu et al. | Oct 2022 | A1 |
| 20220400053 | Liu et al. | Dec 2022 | A1 |
| 20230043362 | Kita et al. | Feb 2023 | A1 |
| 20230070224 | Huo et al. | Mar 2023 | A1 |
| 20230104568 | Miriyala et al. | Apr 2023 | A1 |
| 20230179573 | Sosnovich et al. | Jun 2023 | A1 |
| Number | Date | Country |
|---|---|---|
| 2004227600 | May 2009 | AU |
| 106789367 | May 2017 | CN |
| 107947961 | Apr 2018 | CN |
| 108809722 | Nov 2018 | CN |
| 110531987 | Dec 2019 | CN |
| 110611588 | Dec 2019 | CN |
| 111327640 | Jun 2020 | CN |
| 111371627 | Jul 2020 | CN |
| 111865643 | Oct 2020 | CN |
| 113141386 | Jul 2021 | CN |
| 2464151 | Jun 2012 | EP |
| 2464152 | Jun 2012 | EP |
| 3316532 | May 2018 | EP |
| 3617879 | Mar 2020 | EP |
| 2011070707 | Apr 2011 | JP |
| 2012099048 | May 2012 | JP |
| 2018523932 | Aug 2018 | JP |
| 2011159842 | Dec 2011 | WO |
| 2016160523 | Oct 2016 | WO |
| 2018044352 | Mar 2018 | WO |
| 2019241086 | Dec 2019 | WO |
| 2020041073 | Feb 2020 | WO |
| 2021196080 | Oct 2021 | WO |
| 2022026028 | Feb 2022 | WO |
| 2022204941 | Oct 2022 | WO |
| Entry |
|---|
| Author Unknown, “Kubernetes Core Concepts for Azure Kubernetes Service (AKS)”, Jun. 3, 2019, 6 pages, retrieved from https://docs.microsoft.com/en-us/azure/aks/concepts-clusters-workloads. |
| Chawla, Harsh, et al., “Building Microservices Applications on Microsoft Azure: Designing, Developing, Deploying, and Monitoring,” Month Unknown 2019, 271 pages, Harsh Chawla and Hemant Kathuria, India. |
| Sayfan, Gigi, “Mastering Kubernetes: Automating container deployment and management,” May 2017, 426 pages, Packt Publishing, Birmingham, UK. |
| Wodicka, Brent, “A Developer's Guide to Container Orchestration, Kubernetes, & AKS,” Mar. 19, 2019, 5 pages, AIS, Reston, VA, USA. |
| Author Unknown, “E-Security Begins with Sound Security Policies,” Jun. 14, 2001, 23 pages, Symantec Corporation. |
| Darabseh, Ala, et al., “SDDC: A Software Defined Datacenter Experimental Framework,” Proceedings of the 2015 3rd International Conference on Future Internet of Things and Cloud, Aug. 24-26, 2015, 6 pages, IEEE Computer Society, Washington, D.C., USA. |
| Non-published Commonly Owned U.S. Appl. No. 16/897,627, filed Jun. 10, 2020, 85 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 16/897,640, filed Jun. 10, 2020, 84 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 16/897,652, filed Jun. 10, 2020, 84 pages, VMware, Inc. |
| Non-published Commonly Owned Related U.S. Appl. No. 16/897,666 with similar specification, filed Jun. 10, 2020, 78 pages, VMware, Inc. |
| Non-published Commonly Owned Related U.S. Appl. No. 16/897,695 with similar specification, filed Jun. 10, 2020, 77 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 16/897,704, filed Jun. 10, 2020, 85 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 16/897,715, filed Jun. 10, 2020, 70 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 16/945,727, filed Jul. 31, 2020, 41 pages, VMware, Inc. |
| Rouse, Margaret, “What is SDDC (software-defined data center)?—Definition from WhatIs.com,” Mar. 2017, 5 pages, TechTarget.com. |
| Non-Published Commonly Owned Related International Patent Application PCT/CN2020/082785 with similar specification, filed Apr. 1, 2020, 94 pages, VMware, Inc. |
| Non-published Commonly Owned U.S. Appl. No. 17/176,191, filed Feb. 16, 2021, 36 pages, VMware, Inc. |
| PCT International Search Report and Written Opinion of Commonly Owned International Patent Application PCT/CN2020/082785, dated Dec. 31, 2020, 13 pages, International Searching Authority (CN). |
| Abhashkumar, Anubhavnidhi, et al., “Supporting Diverse Dynamic Intent-based Policies Using Janus,” CoNEXT 17, Dec. 12-15, 2017, 14 pages, ACM, Incheon, KR. |
| Abwnawar, Nasser, “A Policy-Based Management Approach to Security in Cloud Systems,” Feb. 2020, 184 pages, De Monfort University, Leicester, UK. |
| Non-Published Commonly Owned U.S. Appl. No. 17/692,634, filed Mar. 11, 2022, 42 pages, VMware, Inc. |
| Author Unknown, “Advanced Networking Features in Kubernetes and Container Bare Metal,” Document 606835-001, Dec. 2018, 42 pages, Intel Corporation. |
| Author Unknown, “Containers and Container Networking for Network Engineers: VMware NSX Container Networking,” Jan. 2018, 58 pages, VMware, Inc. |
| Author Unknown, “OpenShift Container Platform 4.6,” Mar. 3, 2021, 41 pages, Red Hat, Inc. |
| Qi, Shixiong, et al., “Assessing Container Network Interface Plugins: Functionality, Performance, and Scalability,” IEEE Transactions on Network and Service Management, Mar. 2021, 16 pages, vol. 198, No. 1, IEEE. |
| Author Unknown, “NSX vSphere API Guide—NSX 6.2 for vSphere,” Jan. 6, 2017, 400 pages, VMware, Inc. |
| Balla, David, et al., “Adaptive Scaling of Kubernetes Pods,” NOMS 2020—2020 IEEE/IFIP Network Operations and Management Symposium, Apr. 20-24, 2020, 5 pages, IEEE, Budapest, Hungary. |
| Author Unknown, “Chapter 4: Default Security Policy,” IBM Security Access Manager Version 9.0, Oct. 2015, 18 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20210349765 A1 | Nov 2021 | US |
