Virtual IPs and Service Proxies

Every node in a Kubernetes cluster runs a kube-proxy (unless you have deployed your own alternative component in place of kube-proxy).

The kube-proxy component is responsible for implementing a virtual IP mechanism for Services of type other than ExternalName.

A question that pops up every now and then is why Kubernetes relies on proxying to forward inbound traffic to backends. What about other approaches? For example, would it be possible to configure DNS records that have multiple A values (or AAAA for IPv6), and rely on round-robin name resolution?

There are a few reasons for using proxying for Services:

• There is a long history of DNS implementations not respecting record TTLs, and caching the results of name lookups after they should have expired.
• Some apps do DNS lookups only once and cache the results indefinitely.
• Even if apps and libraries did proper re-resolution, the low or zero TTLs on the DNS records could impose a high load on DNS that then becomes difficult to manage.

Later in this page you can read about how various kube-proxy implementations work. Overall, you should note that, when running kube-proxy, kernel level rules may be modified (for example, iptables rules might get created), which won't get cleaned up, in some cases until you reboot. Thus, running kube-proxy is something that should only be done by an administrator which understands the consequences of having a low level, privileged network proxying service on a computer. Although the kube-proxy executable supports a cleanup function, this function is not an official feature and thus is only available to use as-is.

Some of the details in this reference refer to an example: the backend Pods for a stateless image-processing workloads, running with three replicas. Those replicas are fungible—frontends do not care which backend they use. While the actual Pods that compose the backend set may change, the frontend clients should not need to be aware of that, nor should they need to keep track of the set of backends themselves.

Proxy modes

The kube-proxy starts up in different modes, which are determined by its configuration.

On Linux nodes, the available modes for kube-proxy are:

iptables

A mode where the kube-proxy configures packet forwarding rules using iptables, on Linux.

ipvs

a mode where the kube-proxy configures packet forwarding rules using ipvs.

There is only one mode available for kube-proxy on Windows:

kernelspace

a mode where the kube-proxy configures packet forwarding rules in the Windows kernel

iptables proxy mode

This proxy mode is only available on Linux nodes.

In this mode, kube-proxy watches the Kubernetes control plane for the addition and removal of Service and EndpointSlice objects. For each Service, it installs iptables rules, which capture traffic to the Service's clusterIP and port, and redirect that traffic to one of the Service's backend sets. For each endpoint, it installs iptables rules which select a backend Pod.

By default, kube-proxy in iptables mode chooses a backend at random.

Using iptables to handle traffic has a lower system overhead, because traffic is handled by Linux netfilter without the need to switch between userspace and the kernel space. This approach is also likely to be more reliable.

If kube-proxy is running in iptables mode and the first Pod that's selected does not respond, the connection fails. This is different from the old userspace mode: in that scenario, kube-proxy would detect that the connection to the first Pod had failed and would automatically retry with a different backend Pod.

You can use Pod readiness probes to verify that backend Pods are working OK, so that kube-proxy in iptables mode only sees backends that test out as healthy. Doing this means you avoid having traffic sent via kube-proxy to a Pod that's known to have failed.

Example

As an example, consider the image processing application described earlier in the page. When the backend Service is created, the Kubernetes control plane assigns a virtual IP address, for example 10.0.0.1. For this example, assume that the Service port is 1234. All of the kube-proxy instances in the cluster observe the creation of the new Service.

When kube-proxy on a node sees a new Service, it installs a series of iptables rules which redirect from the virtual IP address to more iptables rules, defined per Service. The per-Service rules link to further rules for each backend endpoint, and the per- endpoint rules redirect traffic (using destination NAT) to the backends.

When a client connects to the Service's virtual IP address the iptables rule kicks in. A backend is chosen (either based on session affinity or randomly) and packets are redirected to the backend without rewriting the client IP address.

This same basic flow executes when traffic comes in through a node-port or through a load-balancer, though in those cases the client IP address does get altered.

Optimizing iptables mode performance

In large clusters (with tens of thousands of Pods and Services), the iptables mode of kube-proxy may take a long time to update the rules in the kernel when Services (or their EndpointSlices) change. You can adjust the syncing behavior of kube-proxy via options in the iptables section of the kube-proxy configuration file (which you specify via kube-proxy --config <path>):

...
iptables:
  minSyncPeriod: 1s
  syncPeriod: 30s
...

Performance optimization for iptables mode

In Kubernetes 1.27 the kube-proxy defaults to a minimal approach to iptables-restore operations, only making updates where Services or EndpointSlices have actually changed. This is a performance optimization. The original implementation updated all the rules for all Services on every sync; this sometimes led to performance issues (update lag) in large clusters.

If you are not running kube-proxy from Kubernetes 1.27, check the behavior and associated advice for the version that you are actually running.

If you were previously overriding minSyncPeriod, you should try removing that override and letting kube-proxy use the default value (1s) or at least a smaller value than you were using before upgrading. You can select the legacy behavior by disabling the MinimizeIPTablesRestore feature gate (you should not need to).

minSyncPeriod

The minSyncPeriod parameter sets the minimum duration between attempts to resynchronize iptables rules with the kernel. If it is 0s, then kube-proxy will always immediately synchronize the rules every time any Service or Endpoint changes. This works fine in very small clusters, but it results in a lot of redundant work when lots of things change in a small time period. For example, if you have a Service backed by a Deployment with 100 pods, and you delete the Deployment, then with minSyncPeriod: 0s, kube-proxy would end up removing the Service's endpoints from the iptables rules one by one, for a total of 100 updates. With a larger minSyncPeriod, multiple Pod deletion events would get aggregated together, so kube-proxy might instead end up making, say, 5 updates, each removing 20 endpoints, which will be much more efficient in terms of CPU, and result in the full set of changes being synchronized faster.

The larger the value of minSyncPeriod, the more work that can be aggregated, but the downside is that each individual change may end up waiting up to the full minSyncPeriod before being processed, meaning that the iptables rules spend more time being out-of-sync with the current API server state.

The default value of 1s should work well in most clusters, but in very large clusters it may be necessary to set it to a larger value. Especially, if kube-proxy's sync_proxy_rules_duration_seconds metric indicates an average time much larger than 1 second, then bumping up minSyncPeriod may make updates more efficient.

syncPeriod

The syncPeriod parameter controls a handful of synchronization operations that are not directly related to changes in individual Services and EndpointSlices. In particular, it controls how quickly kube-proxy notices if an external component has interfered with kube-proxy's iptables rules. In large clusters, kube-proxy also only performs certain cleanup operations once every syncPeriod to avoid unnecessary work.

For the most part, increasing syncPeriod is not expected to have much impact on performance, but in the past, it was sometimes useful to set it to a very large value (eg, 1h). This is no longer recommended, and is likely to hurt functionality more than it improves performance.

IPVS proxy mode

This proxy mode is only available on Linux nodes.

In ipvs mode, kube-proxy watches Kubernetes Services and EndpointSlices, calls netlink interface to create IPVS rules accordingly and synchronizes IPVS rules with Kubernetes Services and EndpointSlices periodically. This control loop ensures that IPVS status matches the desired state. When accessing a Service, IPVS directs traffic to one of the backend Pods.

The IPVS proxy mode is based on netfilter hook function that is similar to iptables mode, but uses a hash table as the underlying data structure and works in the kernel space. That means kube-proxy in IPVS mode redirects traffic with lower latency than kube-proxy in iptables mode, with much better performance when synchronizing proxy rules. Compared to the other proxy modes, IPVS mode also supports a higher throughput of network traffic.

IPVS provides more options for balancing traffic to backend Pods; these are:

• rr: round-robin
• lc: least connection (smallest number of open connections)
• dh: destination hashing
• sh: source hashing
• sed: shortest expected delay
• nq: never queue

kernelspace proxy mode

This proxy mode is only available on Windows nodes.

The kube-proxy configures packet filtering rules in the Windows Virtual Filtering Platform (VFP), an extension to Windows vSwitch. These rules process encapsulated packets within the node-level virtual networks, and rewrite packets so that the destination IP address (and layer 2 information) is correct for getting the packet routed to the correct destination. The Windows VFP is analogous to tools such as Linux nftables or iptables. The Windows VFP extends the Hyper-V Switch, which was initially implemented to support virtual machine networking.

When a Pod on a node sends traffic to a virtual IP address, and the kube-proxy selects a Pod on a different node as the load balancing target, the kernelspace proxy mode rewrites that packet to be destined to the target backend Pod. The Windows Host Networking Service (HNS) ensures that packet rewriting rules are configured so that the return traffic appears to come from the virtual IP address and not the specific backend Pod.

Direct server return for kernelspace mode

As an alternative to the basic operation, a node that hosts the backend Pod for a Service can apply the packet rewriting directly, rather than placing this burden on the node where the client Pod is running. This is called direct server return.

To use this, you must run kube-proxy with the --enable-dsr command line argument and enable the WinDSR feature gate.

Direct server return also optimizes the case for Pod return traffic even when both Pods are running on the same node.

Session affinity

In these proxy models, the traffic bound for the Service's IP:Port is proxied to an appropriate backend without the clients knowing anything about Kubernetes or Services or Pods.

If you want to make sure that connections from a particular client are passed to the same Pod each time, you can select the session affinity based on the client's IP addresses by setting .spec.sessionAffinity to ClientIP for a Service (the default is None).

Session stickiness timeout

You can also set the maximum session sticky time by setting .spec.sessionAffinityConfig.clientIP.timeoutSeconds appropriately for a Service. (the default value is 10800, which works out to be 3 hours).

IP address assignment to Services

Unlike Pod IP addresses, which actually route to a fixed destination, Service IPs are not actually answered by a single host. Instead, kube-proxy uses packet processing logic (such as Linux iptables) to define virtual IP addresses which are transparently redirected as needed.

When clients connect to the VIP, their traffic is automatically transported to an appropriate endpoint. The environment variables and DNS for Services are actually populated in terms of the Service's virtual IP address (and port).

Avoiding collisions

One of the primary philosophies of Kubernetes is that you should not be exposed to situations that could cause your actions to fail through no fault of your own. For the design of the Service resource, this means not making you choose your own IP address if that choice might collide with someone else's choice. That is an isolation failure.

In order to allow you to choose an IP address for your Services, we must ensure that no two Services can collide. Kubernetes does that by allocating each Service its own IP address from within the service-cluster-ip-range CIDR range that is configured for the API Server.

IP address allocation tracking

To ensure each Service receives a unique IP, an internal allocator atomically updates a global allocation map in etcd prior to creating each Service. The map object must exist in the registry for Services to get IP address assignments, otherwise creations will fail with a message indicating an IP address could not be allocated.

In the control plane, a background controller is responsible for creating that map (needed to support migrating from older versions of Kubernetes that used in-memory locking). Kubernetes also uses controllers to check for invalid assignments (e.g. due to administrator intervention) and for cleaning up allocated IP addresses that are no longer used by any Services.

The background controller is also replaced by a new one to handle the new IPAddress objects and the migration from the old allocator model.

One of the main benefits of the new allocator is that it removes the size limitations for the service-cluster-ip-range, there is no limitations for IPv4 and for IPv6 users can use masks equal or larger than /64 (previously it was /108).

Users now will be able to inspect the IP addresses assigned to their Services, and Kubernetes extensions such as the Gateway API, can use this new IPAddress object kind to enhance the Kubernetes networking capabilities, going beyond the limitations of the built-in Service API.

kubectl get services

NAME         TYPE        CLUSTER-IP        EXTERNAL-IP   PORT(S)   AGE
kubernetes   ClusterIP   2001:db8:1:2::1   <none>        443/TCP   3d1h

kubectl get ipaddresses

NAME              PARENTREF
2001:db8:1:2::1   services/default/kubernetes
2001:db8:1:2::a   services/kube-system/kube-dns

IP address ranges for Service virtual IP addresses

Kubernetes divides the ClusterIP range into two bands, based on the size of the configured service-cluster-ip-range by using the following formula min(max(16, cidrSize / 16), 256). That formula paraphrases as never less than 16 or more than 256, with a graduated step function between them.

Kubernetes prefers to allocate dynamic IP addresses to Services by choosing from the upper band, which means that if you want to assign a specific IP address to a type: ClusterIP Service, you should manually assign an IP address from the lower band. That approach reduces the risk of a conflict over allocation.

If you disable the ServiceIPStaticSubrange feature gate then Kubernetes uses a single shared pool for both manually and dynamically assigned IP addresses, that are used for type: ClusterIP Services.

Traffic policies

You can set the .spec.internalTrafficPolicy and .spec.externalTrafficPolicy fields to control how Kubernetes routes traffic to healthy (“ready”) backends.

Internal traffic policy

You can set the .spec.internalTrafficPolicy field to control how traffic from internal sources is routed. Valid values are Cluster and Local. Set the field to Cluster to route internal traffic to all ready endpoints and Local to only route to ready node-local endpoints. If the traffic policy is Local and there are no node-local endpoints, traffic is dropped by kube-proxy.

External traffic policy

You can set the .spec.externalTrafficPolicy field to control how traffic from external sources is routed. Valid values are Cluster and Local. Set the field to Cluster to route external traffic to all ready endpoints and Local to only route to ready node-local endpoints. If the traffic policy is Local and there are are no node-local endpoints, the kube-proxy does not forward any traffic for the relevant Service.

Traffic to terminating endpoints

If the ProxyTerminatingEndpoints feature gate is enabled in kube-proxy and the traffic policy is Local, that node's kube-proxy uses a more complicated algorithm to select endpoints for a Service. With the feature enabled, kube-proxy checks if the node has local endpoints and whether or not all the local endpoints are marked as terminating. If there are local endpoints and all of them are terminating, then kube-proxy will forward traffic to those terminating endpoints. Otherwise, kube-proxy will always prefer forwarding traffic to endpoints that are not terminating.

This forwarding behavior for terminating endpoints exist to allow NodePort and LoadBalancer Services to gracefully drain connections when using externalTrafficPolicy: Local.

As a deployment goes through a rolling update, nodes backing a load balancer may transition from N to 0 replicas of that deployment. In some cases, external load balancers can send traffic to a node with 0 replicas in between health check probes. Routing traffic to terminating endpoints ensures that Node's that are scaling down Pods can gracefully receive and drain traffic to those terminating Pods. By the time the Pod completes termination, the external load balancer should have seen the node's health check failing and fully removed the node from the backend pool.

What's next

To learn more about Services, read Connecting Applications with Services.

You can also:

• Read about Services as a concept
• Read about Ingresses as a concept
• Read the API reference for the Service API