External Application Load Balancer overview

This document introduces the concepts that you need to understand how to configure an external Application Load Balancer.

An external Application Load Balancer is a proxy-based Layer 7 load balancer that enables you to run and scale your services behind a single external IP address. The external Application Load Balancer distributes HTTP and HTTPS traffic to backends hosted on a variety of Google Cloud platforms (such as Compute Engine, Google Kubernetes Engine (GKE), Cloud Storage, and so on), as well as external backends connected over the internet or via hybrid connectivity. For details, see Application Load Balancer overview: Use cases.

Modes of operation

You can configure an external Application Load Balancer in the following modes:

• Global external Application Load Balancer. This is a global load balancer that is implemented as a managed service on Google Front Ends (GFEs). It uses the open-source Envoy proxy to support advanced traffic management capabilities such as traffic mirroring, weight-based traffic splitting, request/response-based header transformations, and more.
• Classic Application Load Balancer. This is the classic external Application Load Balancer that is global in Premium Tier but can be configured to be regional in Standard Tier. This load balancer is implemented on Google Front Ends (GFEs). GFEs are distributed globally and operate together using Google's global network and control plane.
• Regional external Application Load Balancer. This is a regional load balancer that is implemented as a managed service on the open-source Envoy proxy. It includes advanced traffic management capabilities such as traffic mirroring, weight-based traffic splitting, request/response-based header transformations, and more.

Identify the mode

   gcloud compute forwarding-rules describe FORWARDING_RULE_NAME
   

Architecture

The following resources are required for an external Application Load Balancer deployment:

• For regional external Application Load Balancers only, a proxy-only subnet is used to send connections from the load balancer to the backends.

• An external forwarding rule specifies an external IP address, port, and target HTTP(S) proxy. Clients use the IP address and port to connect to the load balancer.

• A target HTTP(S) proxy receives a request from the client. The HTTP(S) proxy evaluates the request by using the URL map to make traffic routing decisions. The proxy can also authenticate communications by using SSL certificates.

  • For HTTPS load balancing, the target HTTPS proxy uses SSL certificates to prove its identity to clients. A target HTTPS proxy supports up to the documented number of SSL certificates.

• The HTTP(S) proxy uses a URL map to make a routing determination based on HTTP attributes (such as the request path, cookies, or headers). Based on the routing decision, the proxy forwards client requests to specific backend services or backend buckets. The URL map can specify additional actions, such as sending redirects to clients.

• A backend service distributes requests to healthy backends. The global external Application Load Balancers also support backend buckets. One or more backends must be connected to the backend service or backend bucket.

• A health check periodically monitors the readiness of your backends. This reduces the risk that requests might be sent to backends that can't service the request.

• Firewall rules for your backends to accept health check probes. Regional external Application Load Balancers require an additional firewall rule to allow traffic from the proxy-only subnet to reach the backends.

Global

This diagram shows the components of a global external Application Load Balancer deployment. This architecture applies to both, the global external Application Load Balancer, and the classic Application Load Balancer in Premium Tier.

Regional

This diagram shows the components of a regional external Application Load Balancer deployment.

Proxy-only subnet

Proxy-only subnets are only required for regional external Application Load Balancers.

The proxy-only subnet provides a set of IP addresses that Google uses to run Envoy proxies on your behalf. You must create one proxy-only subnet in each region of a VPC network where you use regional external Application Load Balancers. The --purpose flag for this proxy-only subnet is set to REGIONAL_MANAGED_PROXY. All regional Envoy-based load balancers in the same region and VPC network share a pool of Envoy proxies from the same proxy-only subnet. Further:

• Proxy-only subnets are only used for Envoy proxies, not your backends.
• Backend VMs or endpoints of all regional external Application Load Balancers in a region and VPC network receive connections from the proxy-only subnet.
• The IP address of the regional external Application Load Balancer is not located in the proxy-only subnet. The load balancer's IP address is defined by its external managed forwarding rule, which is described below.

If you previously created a proxy-only subnet with --purpose=INTERNAL_HTTPS_LOAD_BALANCER, you need to migrate the subnet's purpose to REGIONAL_MANAGED_PROXY before you can create other Envoy-based load balancers in the same region of the VPC network.

Forwarding rules and IP addresses

Forwarding rules route traffic by IP address, port, and protocol to a load balancing configuration consisting of a target proxy, URL map, and one or more backend services.

IP address specification. Each forwarding rule provides a single IP address that can be used in DNS records for your application. No DNS-based load balancing is required. You can either specify the IP address to be used or let Cloud Load Balancing assign one for you.

Port specification. Each forwarding rule for an Application Load Balancer can reference a single port from 1-65535. To support multiple ports, you must configure multiple forwarding rules. You can configure multiple forwarding rules to use the same external IP address (VIP) and to reference the same target HTTP(S) proxy as long as the overall combination of IP address, port, and protocol is unique for each forwarding rule. This way, you can use a single load balancer with a shared URL map as a proxy for multiple applications.

The type of forwarding rule, IP address, and load balancing scheme used by external Application Load Balancers depends on the mode of the load balancer and which Network Service Tier the load balancer is in.

For the complete list of protocols supported by external Application Load Balancer forwarding rules in each mode, see Load balancer features.

Forwarding rules and VPC networks

This section describes how forwarding rules used by external Application Load Balancers are associated with VPC networks.

Target proxies

Target proxies terminate HTTP(S) connections from clients. One or more forwarding rules direct traffic to the target proxy, and the target proxy consults the URL map to determine how to route traffic to backends.

Do not rely on the proxy to preserve the case of request or response header names. For example, a Server: Apache/1.0 response header might appear at the client as server: Apache/1.0.

The following table specifies the type of target proxy required by external Application Load Balancers.

In addition to headers added by the target proxy, the load balancer adjusts other HTTP headers in the following ways:

• For the global external Application Load Balancer, both request and response headers might be converted to lowercase.

  The only exception to this is when you use global internet NEG backends with HTTP/1.1. For details about how HTTP/1.1 headers are processed with global internet NEGs, see the Internet NEGs overview.

• For the classic Application Load Balancer, request and response headers are converted to lowercase except when you use HTTP/1.1. With HTTP/1.1, headers are proper-cased instead. The first letter of the header's key and every letter following a hyphen (-) is capitalized to preserve compatibility with HTTP/1.1 clients. For example, user-agent is changed to User-Agent, and content-encoding is changed to Content-Encoding.

• Some headers are coalesced. When there are multiple instances of the same header key (for example, Via), the load balancer combines their values into a single comma-separated list for a single header key. Only the headers whose values can be represented as a comma-separated list are coalesced. Other headers, such as Set-Cookie, are never coalesced.

Host header

When the load balancer makes the HTTP request, the load balancer preserves the Host header of the original request.

X-Forwarded-For header

The load balancer appends two IP addresses separated by a single comma to the X-Forwarded-For header in the following order:

• The IP address of the client that connects to the load balancer
• The IP address of the load balancer's forwarding rule

If there is no X-Forwarded-For header on the incoming request, these two IP addresses are the entire header value:

X-Forwarded-For: <client-ip>,<load-balancer-ip>

If the request includes an X-Forwarded-For header, the load balancer preserves the supplied value before the <client-ip>,<load-balancer-ip>:

X-Forwarded-For: <supplied-value>,<client-ip>,<load-balancer-ip>

When running HTTP reverse proxy software on the load balancer's backends, the software might append one or both of the following IP addresses to the end of the X-Forwarded-For header:

• The IP address of the Google Front End (GFE) that connected to the backend. These IP addresses are in the 130.211.0.0/22 and 35.191.0.0/16 ranges.

• The IP address of the backend system itself.

Thus, an upstream process after your load balancer's backend might receive an X-Forwarded-For header of the form:

<existing-values>,<client-ip>,<load-balancer-ip>,<GFE-IP>,<backend-IP>

Cloud Trace support

Trace is not supported with Application Load Balancers. The global and classic Application Load Balancers add the X-Cloud-Trace-Context header if it is not present. The regional external Application Load Balancer does not add this header. If the X-Cloud-Trace-Context header is already present, it passes through the load balancers unmodified. However, no traces or spans are exported by the load balancer.

URL maps

URL maps define matching patterns for URL-based routing of requests to the appropriate backend services. The URL map allows you to divide your traffic by examining the URL components to send requests to different sets of backends. A default service is defined to handle any requests that do not match a specified host rule or path matching rule.

In some situations, such as the multi-region load balancing example, you might not define any URL rules and rely only on the default service.

URL maps support several advanced traffic management features such as header-based traffic steering, weight-based traffic splitting, and request mirroring. For more information, see the following:

• Traffic management overview for global external Application Load Balancer

• Traffic management overview for regional external Application Load Balancer

The following table specifies the type of URL map required by external Application Load Balancers in each mode.

SSL certificates

External Application Load Balancers using target HTTPS proxies require private keys and SSL certificates as part of the load balancer configuration.

• Google Cloud provides two configuration methods for assigning private keys and SSL certificates to target HTTPS proxies: Compute Engine SSL certificates and Certificate Manager. For a description of each configuration, see Certificate configuration methods in the SSL certificates overview.

• Google Cloud provides two certificate types: Self-managed and Google-managed. For a description of each type, see Certificate types in the SSL certificates overview.

The type of external Application Load Balancer (global, regional, or classic) determines which configuration methods and certificate types are supported. For details, see Certificates and Google Cloud load balancers in the SSL certificates overview.

SSL policies

SSL policies specify the set of SSL features that Google Cloud load balancers use when negotiating SSL with clients.

By default, HTTPS Load Balancing uses a set of SSL features that provides good security and wide compatibility. Some applications require more control over which SSL versions and ciphers are used for their HTTPS or SSL connections. You can define an SSL policy to specify the set of SSL features that your load balancer uses when negotiating SSL with clients. In addition, you can apply that SSL policy to your target HTTPS proxy.

The following table specifies the SSL policy support for load balancers in each mode.

Backend services

A backend service provides configuration information to the load balancer so that it can direct requests to its backends—for example, Compute Engine instance groups or network endpoint groups (NEGs). For more information about backend services, see Backend services overview.

For an example showing how to set up a load balancer with a backend service and a Compute Engine backend, see Setting up an external Application Load Balancer with a Compute Engine backend.

Backend service scope

The following table indicates which backend service resource and scope is used by external Application Load Balancers:

Protocol to the backends

Backend services for Application Load Balancers must use one of the following protocols to send requests to backends:

• HTTP, which uses HTTP/1.1 and no TLS
• HTTPS, which uses HTTP/1.1 and TLS
• HTTP/2, which uses HTTP/2 and TLS (HTTP/2 without encryption is not supported.)

The load balancer only uses the backend service protocol that you specify to communicate with its backends. The load balancer doesn't fall back to a different protocol if it is unable to communicate with backends using the specified backend service protocol.

The backend service protocol doesn't need to match the protocol used by clients to communicate with the load balancer. For example, clients can send requests to the load balancer using HTTP/2, but the load balancer can communicate with backends using HTTP/1.1 (HTTP or HTTPS).

Backend buckets

Backend buckets direct incoming traffic to Cloud Storage buckets. For an example showing how to add a bucket to a external Application Load Balancer, see Setting up a load balancer with backend buckets. For general information about Cloud Storage, see What is Cloud Storage?

Backends

The following table specifies the backends and related features supported by external Application Load Balancers in each mode.

Backends and VPC networks

The restrictions on where backends can be located depend on the type of load balancer.

For global external Application Load Balancer backends, the following applies:

• Backend instances (for instance group backends) and backend endpoints (for NEG backends) can be located in any VPC network within the same organization. The different VPC networks don't need to be connected using VPC Network Peering because GFEs communicate directly with backends in their respective VPC networks.

• Cloud Storage buckets aren't associated with a VPC network. They can be located in any project within the same organization.

  Global external Application Load Balancers also support Shared VPC environments where you can share VPC networks and their associated resources across projects. However, for global external Application Load Balancers, you don't need to configure Shared VPC to be able to reference backend services or backend buckets from other projects in your organization.

For classic Application Load Balancer backends, the following applies:

• All backend instances from instance group backends and all backend endpoints from NEG backends must be located in the same project. However, an instance group backend or a NEG can use a different VPC network in that project. The different VPC networks don't need to be connected using VPC Network Peering because GFEs communicate directly with backends in their respective VPC networks.

• Cloud Storage buckets are not associated with a VPC network. However, they must be located in the same project as the load balancer.

For regional external Application Load Balancer backends, the following applies:

• For instance groups, zonal NEGs, and hybrid connectivity NEGs, all backends must be located in the same project and region as the backend service. However, a load balancer can reference a backend that uses a different VPC network in the same project as the backend service (this capability is in Preview). Connectivity between the load balancer's VPC network and the backend VPC network can be configured using either VPC Network Peering, Cloud VPN tunnels, Cloud Interconnect VLAN attachments, or a Network Connectivity Center framework.

  Backend network definition

  • For zonal NEGs and hybrid NEGs, you explicitly specify the VPC network when you create the NEG.
  • For managed instance groups, the VPC network is defined in the instance template.
  • For unmanaged instance groups, the instance group's VPC network is set to match the VPC network of the nic0 interface for the first VM added to the instance group.

  Backend network requirements

  Your backend's network must satisfy one of the following network requirements:

  • The backend's VPC network must exactly match the forwarding rule's VPC network.

  • The backend's VPC network must be connected to the forwarding rule's VPC network using VPC Network Peering. You must configure subnet route exchanges to allow communication between the proxy-only subnet in the forwarding rule's VPC network and the subnets used by the backend instances or endpoints.

• Both the backend's VPC network and the forwarding rule's VPC network must be VPC spokes on the same Network Connectivity Center hub. Import and export filters must allow communication between the proxy-only subnet in the forwarding rule's VPC network and the subnets used by backend instances or endpoints.

• For all other backend types, all backends must be located in the same VPC network and region.

  Regional external Application Load Balancers also support Shared VPC environments where you can share VPC networks and their associated resources across projects. If you want the regional external Application Load Balancer's backend service and backends to be in a different project than the forwarding rule, you need to configure the load balancer in a Shared VPC environment with cross-project service referencing.

Backends and network interfaces

If you use instance group backends, packets are always delivered to nic0. If you want to send packets to different NICs, use NEG backends instead.

If you use zonal NEG backends, packets are sent to whatever network interface is represented by the endpoint in the NEG. The NEG endpoints must be in the same VPC network as the NEG's explicitly defined VPC network.

Health checks

Each backend service specifies a health check that periodically monitors the backends' readiness to receive a connection from the load balancer. This reduces the risk that requests might be sent to backends that can't service the request. Health checks do not check if the application itself is working.

For the health check probes, you must create an ingress allow firewall rule that allows health check probes to reach your backend instances. Typically, health check probes originate from Google's centralized health checking mechanism.

Regional external Application Load Balancers that use hybrid NEG backends are an exception to this rule because their health checks originate from the proxy-only subnet instead. For details, see the Hybrid NEGs overview.

Health check protocol

Although it is not required and not always possible, it is a best practice to use a health check whose protocol matches the protocol of the backend service. For example, an HTTP/2 health check most accurately tests HTTP/2 connectivity to backends. In contrast, regional external Application Load Balancers that use hybrid NEG backends do not support gRPC health checks. For the list of supported health check protocols, see Load balancing features.

The following table specifies the scope of health checks supported by external Application Load Balancers in each mode.

For more information about health checks, see the following:

• Health checks overview
• Creating health checks

Firewall rules

The load balancer requires the following firewall rules:

• For the global external Application Load Balancers, an ingress allow rule to permit traffic from Google Front Ends (GFEs) to reach your backends. For the regional external Application Load Balancer, an ingress allow rule to permit traffic from the proxy-only subnet to reach your backends.
• An ingress allow rule to permit traffic from the health check probes ranges. For more information about health check probes and why it's necessary to allow traffic from them, see Probe IP ranges and firewall rules.

Firewall rules are implemented at the VM instance level, not on GFE proxies. You cannot use Google Cloud firewall rules to prevent traffic from reaching the load balancer. For the global external Application Load Balancer and the classic Application Load Balancer, you can use Google Cloud Armor to achieve this.

The ports for these firewall rules must be configured as follows:

• Allow traffic to the destination port for each backend service's health check.

• For instance group backends: Determine the ports to be configured by the mapping between the backend service's named port and the port numbers associated with that named port on each instance group. The port numbers can vary among instance groups assigned to the same backend service.

• For GCE_VM_IP_PORT NEG backends: Allow traffic to the port numbers of the endpoints.

The following table summarizes the required source IP address ranges for the firewall rules:

GKE support

GKE uses external Application Load Balancers in the following ways:

• External Gateways created using the GKE Gateway controller can use any mode of an External Application Load Balancer. You control the load balancer's mode by choosing a GatewayClass. The GKE Gateway controller always uses GCE_VM_IP_PORT zonal NEG backends.

• External Ingresses created using the GKE Ingress controller are always Classic Application Load Balancers. The GKE Ingress controller prefers to use GCE_VM_IP_PORT zonal NEG backends, though instance group backends are also supported.

• You can use GCE_VM_IP_PORT zonal NEG created and managed by GKE Services as backends for any Application Load Balancer or Proxy Network Load Balancer. For more information, see Container-native load balancing through standalone zonal NEGs.

Shared VPC architecture

External Application Load Balancers support networks that use Shared VPC. Shared VPC lets organizations connect resources from multiple projects to a common VPC network so that they can communicate with each other securely and efficiently by using internal IP addresses from that network. If you're not already familiar with Shared VPC, read the Shared VPC overview.

There are many ways to configure an external Application Load Balancer within a Shared VPC network. Regardless of type of deployment, all the components of the load balancer must be in the same organization.

While you can create all the load balancing components and backends in the Shared VPC host project, this type of deployment does not separate network administration and service development responsibilities.

All load balancer components and backends in a service project

The following architecture diagram shows a standard Shared VPC deployment where all load balancer components and backends are in a service project. This deployment type is supported by all Application Load Balancers.

The load balancer components and backends must use the same VPC network.

Serverless backends in a Shared VPC environment

For a load balancer that is using a serverless NEG backend, the backend Cloud Run or Cloud Run functions service must be in the same project as the serverless NEG.

Additionally, for the regional external Application Load Balancer that supports cross-project service referencing, the backend service, serverless NEG, and the Cloud Run service must always be in the same service project.

Cross-project service referencing

Cross-project service referencing is a deployment model where the load balancer's frontend and URL map are in one project and the load balancer's backend service and backends are in a different project.

Cross-project service referencing allows organizations to configure one central load balancer and route traffic to hundreds of services distributed across multiple different projects. You can centrally manage all traffic routing rules and policies in one URL map. You can also associate the load balancer with a single set of hostnames and SSL certificates. You can therefore optimize the number of load balancers needed to deploy your application, and lower manageability, operational costs, and quota requirements.

By having different projects for each of your functional teams, you can also achieve separation of roles within your organization. Service owners can focus on building services in service projects, while network teams can provision and maintain load balancers in another project, and both can be connected by using cross-project service referencing.

Service owners can maintain autonomy over the exposure of their services and control which users can access their services by using the load balancer. This is achieved by a special IAM role called the Compute Load Balancer Services User role (roles/compute.loadBalancerServiceUser).

Cross-project service referencing support differs based on the type of load balancer:

• For global external Application Load Balancers, your load balancer's frontend and URL map can reference backend services or backend buckets from any project within the same organization. No VPC network restrictions apply. While you can use a Shared VPC environment to configure a cross-project deployment as shown in this example, this is not a requirement.

• For regional external Application Load Balancers, you must create the load balancer in a Shared VPC environment. The load balancer's frontend and URL map must be in a host or service project, and the load balancer's backend services and backends can be distributed across host or service projects in the same Shared VPC environment.

To learn how to configure Shared VPC for a regional external Application Load Balancer—with and without cross-project service referencing, see Set up a regional external Application Load Balancer with Shared VPC.

Usage notes for cross-project service referencing

• Cross-project service referencing can be used with instance groups, serverless NEGs, and most other supported backend types. However, the following limitations apply:

  • With global external Application Load Balancers, you can't reference a cross-project backend service if the backend service has serverless NEG backends with App Engine.

  • With regional external Application Load Balancers, you can't reference a cross-project backend service if the backend service has regional internet NEG backends.
• Cross-project service referencing is not supported for the classic Application Load Balancer.
• Google Cloud does not differentiate between resources (for example, backend services) using the same name across multiple projects. Therefore, when you're using cross-project service referencing, we recommend that you use unique backend service names across projects within your organization.
• If you see an error such as "Cross-project references for this resource are not allowed", make sure that you have the permission to use the resource. An administrator of the project that owns the resource must grant you the Compute Load Balancer Services User role (roles/compute.loadBalancerServiceUser). This role can be granted at the project level or at the resource level. For an example, see [Grant permissions to the Compute Load Balancer Admin to use the backend service](/load-balancing/docs/https/set-up-global-ext-https-shared-vpc#grant-bs-user).

Example 1: Load balancer frontend and backend in different service projects

Here is an example of a Shared VPC deployment where the load balancer's frontend and URL map are created in service project A and the URL map references a backend service in service project B.

In this case, Network Admins or Load Balancer Admins in service project A will require access to backend services in service project B. Service project B admins grant the compute.loadBalancerServiceUser IAM role to Load Balancer Admins in service project A who want to reference the backend service in service project B.

Example 2: Load balancer frontend in the host project and backends in service projects

Here is an example of a Shared VPC deployment where the load balancer's frontend and URL map are created in the host project and the backend services (and backends) are created in service projects.

In this case, Network Admins or Load Balancer Admins in the host project will require access to backend services in the service project. Service project admins grant the compute.loadBalancerServiceUser IAM role to to Load Balancer Admins in the host project A who want to reference the backend service in the service project.

Example 3: Load balancer frontend and backends in different projects

Here is an example of a deployment where the global external Application Load Balancer's frontend and URL map are created in a different project from the load balancer's backend service and backends. This type of deployment doesn't use Shared VPC and is supported only for global external Application Load Balancers.

How connections work

Global external Application Load Balancer connections

The global external Application Load Balancers are implemented by many proxies called Google Front Ends (GFEs). There isn't just a single proxy. In Premium Tier, the same global external IP address is advertised from various points of presence, and client requests are directed to the client's nearest GFE.

Depending on where your clients are, multiple GFEs can initiate HTTP(S) connections to your backends. Packets sent from GFEs have source IP addresses from the same range used by health check probers: 35.191.0.0/16 and 130.211.0.0/22.

Depending on the backend service configuration, the protocol used by each GFE to connect to your backends can be HTTP, HTTPS, or HTTP/2. For HTTP or HTTPS connections, the HTTP version used is HTTP 1.1.

HTTP keepalive is enabled by default, as specified in the HTTP 1.1 specification. HTTP keepalives attempt to efficiently use the same TCP session; however, there's no guarantee. The GFE uses a client HTTP keepalive timeout of 610 seconds and a default backend keepalive timeout value of 600 seconds. You can update the client HTTP keepalive timeout but the backend keepalive timeout value is fixed. You can configure the request/response timeout by setting the backend service timeout. Though closely related, an HTTP keepalive and a TCP idle timeout are not the same thing. For more information, see timeouts and retries.

To ensure that traffic is load balanced evenly, the load balancer might cleanly close a TCP connection either by sending a FIN ACK packet after completing a response that included a Connection: close header, or it might issue an HTTP/2 GOAWAY frame after completing a response. This behavior does not interfere with any active requests or responses.

The numbers of HTTP connections and TCP sessions vary depending on the number of GFEs connecting, the number of clients connecting to the GFEs, the protocol to the backends, and where backends are deployed.

For more information, see How external Application Load Balancers work in the solutions guide: Application Capacity Optimizations with Global Load Balancing.

Regional external Application Load Balancer connections

The regional external Application Load Balancer is a managed service implemented on the Envoy proxy. The regional external Application Load Balancer uses a shared subnet called a proxy-only subnet to provision a set of IP addresses that Google uses to run Envoy proxies on your behalf. The --purpose flag for this proxy-only subnet is set to REGIONAL_MANAGED_PROXY. All regional Envoy-based load balancers in a particular network and region share this subnet.

Clients use the load balancer's IP address and port to connect to the load balancer. Client requests are directed to the proxy-only subnet in the same region as the client. The load balancer terminates clients requests and then opens new connections from the proxy-only subnet to your backends. Therefore, packets sent from the load balancer have source IP addresses from the proxy-only subnet.

Depending on the backend service configuration, the protocol used by Envoy proxies to connect to your backends can be HTTP, HTTPS, or HTTP/2. If HTTP or HTTPS, the HTTP version is HTTP 1.1. HTTP keepalive is enabled by default, as specified in the HTTP 1.1 specification. The Envoy proxy sets both the client HTTP keepalive timeout and the backend keepalive timeout to a default value of 600 seconds each. You can update the client HTTP keepalive timeout but the backend keepalive timeout value is fixed. You can configure the request/response timeout by setting the backend service timeout. For more information, see timeouts and retries.

Client communications with the load balancer

• Clients can communicate with the load balancer by using the HTTP 1.1 or HTTP/2 protocol.
• When HTTPS is used, modern clients default to HTTP/2. This is controlled on the client, not on the HTTPS load balancer.
• You cannot disable HTTP/2 by making a configuration change on the load balancer. However, you can configure some clients to use HTTP 1.1 instead of HTTP/2. For example, with curl, use the --http1.1 parameter.
• External Application Load Balancers support the HTTP/1.1 100 Continue response.

For the complete list of protocols supported by external Application Load Balancer forwarding rules in each mode, see Load balancer features.

Source IP addresses for client packets

The source IP address for packets, as seen by the backends, is not the Google Cloud external IP address of the load balancer. In other words, there are two TCP connections.

• Connection 1, from original client to the load balancer (GFE):

  • Source IP address: the original client (or external IP address if the client is behind NAT or a forward proxy).
  • Destination IP address: your load balancer's IP address.

• Connection 2, from the load balancer (GFE) to the backend VM or endpoint:

  • Source IP address: an IP address in one of the ranges specified in Firewall rules.

  • Destination IP address: the internal IP address of the backend VM or container in the VPC network.

• Connection 1, from original client to the load balancer (proxy-only subnet):

  • Source IP address: the original client (or external IP address if the client is behind NAT or a forward proxy).
  • Destination IP address: your load balancer's IP address.

• Connection 2, from the load balancer (proxy-only subnet) to the backend VM or endpoint:

  • Source IP address: an IP address in the proxy-only subnet that is shared among all the Envoy-based load balancers deployed in the same region and network as the load balancer.

  • Destination IP address: the internal IP address of the backend VM or container in the VPC network.

Special routing paths

Google Cloud uses special routes not defined in your VPC network to route packets for the following types of traffic:

• For health checks, except distributed Envoy health checks. For more information, see Paths for health checks.

• Between GFEs and backends of global external Application Load Balancers and classic Application Load Balancers. For more information, see Paths between Google Front Ends and backends.

Google Cloud uses subnet routes for proxy-only subnets to route packets for the following types of traffic:

• When using distributed Envoy health checks.

For regional external Application Load Balancers, Google Cloud uses open-source Envoy proxies to terminate client requests to the load balancer. The load balancer terminates the TCP session and opens a new TCP session from the region's proxy- only subnet to your backend. Routes defined within your VPC network facilitate communication from Envoy proxies to your backends and from your backends to the Envoy proxies.

Open ports

GFEs have several open ports to support other Google services that run on the same architecture. When you run a port scan, you might see other open ports for other Google services running on GFEs.

Running a port scan on the IP address of a GFE-based load balancer is not useful from an auditing perspective for the following reasons:

• A port scan (for example, with nmap) generally expects no response packet or a TCP RST packet when performing TCP SYN probing. GFEs will send SYN-ACK packets in response to SYN probes only for ports on which you have configured a forwarding rule. GFEs only send packets to your backends in response to packets sent to your load balancer's IP address and the destination port configured on its forwarding rule. Packets that are sent to a different IP address or port are not sent to your backends.

  GFEs implement security features such as Google Cloud Armor. With Google Cloud Armor Standard, GFEs provide always-on protection from volumetric and protocol-based DDoS attacks and SYN floods. This protection is available even if you haven't explicitly configured Google Cloud Armor. You'll only be charged if you configure security policies, or if you enroll in Managed Protection Plus.

• Packets sent to the IP address of your load balancer could be answered by any GFE in Google's fleet; however, scanning a load balancer IP address and destination port combination only interrogates a single GFE per TCP connection. The IP address of your load balancer is not assigned to a single device or system. Thus, scanning the IP address of a GFE-based load balancer does not scan all the GFEs in Google's fleet.

With that in mind, the following are some more effective ways to audit the security of your backend instances:

• A security auditor should inspect the forwarding rules configuration for the load balancer's configuration. The forwarding rules define the destination port for which your load balancer accepts packets and forwards them to the backends. For GFE-based load balancers, each external forwarding rule can only reference a single destination TCP port. For a load balancer using TCP port 443, UDP port 443 is used when the connection is upgraded to QUIC (HTTP/3).

• A security auditor should inspect the firewall rule configuration applicable to backend VMs. The firewall rules that you set block traffic from the GFEs to the backend VMs, but don't block incoming traffic to the GFEs. For best practices, see the firewall rules section.

TLS termination

The following table summarizes how TLS termination is handled by external Application Load Balancers.

Timeouts and retries

External Application Load Balancers support the following types of timeouts for HTTP/HTTPS traffic:

Backend service timeout

The configurable backend service timeout represents the maximum amount of time that the load balancer waits for your backend to process an HTTP request and return the corresponding HTTP response. Except for serverless NEGs, the default value for the backend service timeout is 30 seconds.

For example, if you want to download a 500-MB file, and the value of the backend service timeout is 90 seconds, the load balancer expects the backend to deliver the entire 500-MB file within 90 seconds. It is possible to configure the backend service timeout to be insufficient for the backend to send its complete HTTP response. In this situation, if the load balancer has at least received HTTP response headers from the backend, the load balancer returns the complete response headers and as much of the response body as it could obtain within the backend service timeout.

You should set the backend service timeout to the longest amount of time that you expect your backend to need in order to process an HTTP response. You should increase the backend service timeout if the software running on your backend needs more time to process an HTTP request and return its entire response. For example, you should increase the timeout if you see HTTP 408 responses with jsonPayload.statusDetail client_timed_out.

The backend service timeout accepts values between 1 and 2,147,483,647 seconds; however, larger values are not practical configuration options. Google Cloud also does not guarantee that an underlying TCP connection can remain open for the entirety of the value of the backend service timeout. In case of the global and classic Application Load Balancers, GFEs impose an effective maximum backend service timeout of 86,400 seconds (1 day). Client systems must implement retry logic instead of relying on a TCP connection to be open for long periods of time.

To configure the backend service timeout, use one of the following methods:

Regional external Application Load Balancers use the configured routeActions.timeout parameter of the URL maps and ignores the backend service timeout. When routeActions.timeout is not configured, the value of the backend service timeout is used. When routeActions.timeout is supplied, the backend service timeout is ignored, and the value of routeActions.timeout is used as the request and response timeout instead.

Client HTTP keepalive timeout

The client HTTP keepalive timeout represents the maximum amount of time that a TCP connection can be idle between the (downstream) client and one of the following types of proxies:

• For a global external Application Load Balancer or a classic Application Load Balancer: a first-layer Google Front End

• For a regional external Application Load Balancer: an Envoy proxy

An HTTP keepalive timeout represents the TCP idle timeout for the underlying TCP connections. The client HTTP keepalive timeout does not apply to websockets.

• For a global external Application Load Balancer, the default value is 610 seconds. You can configure the client HTTP keepalive timeout with a value between 5 and 1200 seconds.
• For a classic Application Load Balancer, the client HTTP keepalive timeout is fixed at 610 seconds.
• For a regional external Application Load Balancer, the default value is 600 seconds. You can configure the client HTTP keepalive timeout with a value between 5 and 600 seconds.

To configure the keepalive timeout parameter, use one of the following methods:

The load balancer's client HTTP keepalive timeout should be greater than the HTTP keepalive (TCP idle) timeout used by downstream clients or proxies. If a downstream client has a greater HTTP keepalive (TCP idle) timeout than the load balancer's client HTTP keepalive timeout, it's possible for a race condition to occur. From the perspective of a downstream client, an established TCP connection is permitted to be idle for longer than permitted by the load balancer. This means that the downstream client can send packets after the load balancer considers the TCP connection to be closed. When that happens, the load balancer responds with a TCP reset (RST) packet.

Backend HTTP keepalive timeout

External Application Load Balancers are proxies that use at least two TCP connections:

• For a global external Application Load Balancer or a classic Application Load Balancer, a first TCP connection exists between the (downstream) client and a first-layer GFE. First-layer GFEs connect to second layer GFEs, and then the second-layer GFEs open a second TCP connection to your backends.

• For a regional external Application Load Balancer, a first TCP connection exists between the (downstream) client and an Envoy proxy. The Envoy proxy then opens a second TCP connection to your backends.

The load balancer's secondary TCP connections might not get closed after each request; they can stay open to handle multiple HTTP requests and responses. The backend HTTP keepalive timeout defines the TCP idle timeout between the load balancer and your backends. The backend HTTP keepalive timeout does not apply to websockets.

The backend keepalive timeout is fixed at 10 minutes (600 seconds) and cannot be changed. The load balancer's backend keepalive timeout should be less than the keepalive timeout used by software running on your backends. This avoids a race condition where the operating system of your backends might close TCP connections with a TCP reset (RST). Because the backend keepalive timeout for the load balancer is not configurable, you must configure your backend software so that its HTTP keepalive (TCP idle) timeout value is greater than 600 seconds.

The following table lists the changes necessary to modify keepalive timeout values for common web server software.

QUIC session idle timeout

The QUIC session idle timeout represents the maximum amount of time that a QUIC session can be idle between the client and the GFE of a global external Application Load Balancer or a classic Application Load Balancer.

The QUIC session idle timeout value is defined as the minimum of either the client idle timeout or the GFE idle timeout (300 seconds). The GFE idle timeout is fixed at 300 seconds. The client idle timeout can be configured.

Retries

Support for retry logic depends on the mode of the external Application Load Balancer.

The WebSocket protocol is supported with GKE Ingress.

Illegal request and response handling

The load balancer blocks both client requests and backend responses from reaching the backend or the client, respectively, for a number of reasons. Some reasons are strictly for HTTP/1.1 compliance and others are to avoid unexpected data being passed to or from the backends. None of the checks can be disabled.

The load balancer blocks the following requests for HTTP/1.1 compliance:

• It cannot parse the first line of the request.
• A header is missing the colon (:) delimiter.
• Headers or the first line contain invalid characters.
• The content length is not a valid number, or there are multiple content length headers.
• There are multiple transfer encoding keys, or there are unrecognized transfer encoding values.
• There's a non-chunked body and no content length specified.
• Body chunks are unparseable. This is the only case where some data reaches the backend. The load balancer closes the connections to the client and backend when it receives an unparseable chunk.

Request handling

The load balancer blocks the request if any of the following are true:

• The total size of request headers and the request URL exceeds the limit for the maximum request header size for external Application Load Balancers.
• The request method does not allow a body, but the request has one.
• The request contains an Upgrade header, and the Upgrade header is not used to enable WebSocket connections.
• The HTTP version is unknown.

Response handling

The load balancer blocks the backend's response if any of the following are true:

• The total size of response headers exceeds the limit for maximum response header size for external Application Load Balancers.
• The HTTP version is unknown.

When handling both the request and response, the load balancer might remove or overwrite hop-by-hop headers in HTTP/1.1 before forwarding them to the intended destination.

Traffic distribution

When you add a backend instance group or NEG to a backend service, you specify a balancing mode, which defines a method measuring backend load and a target capacity. External Application Load Balancers support two balancing modes:

• RATE, for instance groups or NEGs, is the target maximum number of requests (queries) per second (RPS, QPS). The target maximum RPS/QPS can be exceeded if all backends are at or above capacity.

• UTILIZATION is the backend utilization of VMs in an instance group.

How traffic is distributed among backends depends on the mode of the load balancer.

Global external Application Load Balancer

Before a Google Front End (GFE) sends requests to backend instances, the GFE estimates which backend instances have capacity to receive requests. This capacity estimation is made proactively, not at the same time as requests are arriving. The GFEs receive periodic information about the available capacity and distribute incoming requests accordingly.

What capacity means depends in part on the balancing mode. For the RATE mode, it is relatively simple: a GFE determines exactly how many requests it can assign per second. UTILIZATION-based load balancing is more complex: the load balancer checks the instances' current utilization and then estimates a query load that each instance can handle. This estimate changes over time as instance utilization and traffic patterns change.

Both factors—the capacity estimation and the proactive assignment—influence the distribution among instances. Thus, Cloud Load Balancing behaves differently from a simple round-robin load balancer that spreads requests exactly 50:50 between two instances. Instead, Google Cloud load balancing attempts to optimize the backend instance selection for each request.

For the global external Application Load Balancer, load balancing is two-tiered. The balancing mode determines the weighting or fraction of traffic that should be sent to each backend (instance group or NEG). Then, the load balancing policy (LocalityLbPolicy) determines how traffic is distributed to instances or endpoints within the group. For more information, see the Load balancing locality policy (backend service API documentation).

For the classic Application Load Balancer, the balancing mode is used to select the most favorable backend (instance group or NEG). Traffic is then distributed in a round robin fashion among instances or endpoints within the backend.

How requests are distributed

GFE-based external Application Load Balancers use the following process to distribute incoming requests:

1. From client to first-layer GFE. Edge routers advertise the forwarding rule's external IP address at the borders of Google's network. Each advertisement lists a next hop to a Layer 3/4 load balancing system (Maglev). The Maglev systems route traffic to a first-layer Google Front End (GFE).
   • When using Premium Tier, Google advertises your load balancer's IP address from all points of presence, worldwide. Each load balancer IP address is global anycast.
   • When using Standard Tier, Google advertises your load balancer's IP address from points of presence associated with the forwarding rule's region. The load balancer uses a regional external IP address. Using a Standard Tier forwarding rule limits you to instance group and zonal NEG backends in the same region as the load balancer's forwarding rule.
2. From first-layer GFE to second-layer GFE. The first-layer GFE terminates TLS if required and then routes traffic to second-layer GFEs according to the following process:
   • First-layer GFEs parse the URL map and select a backend service or backend bucket.
   • For backend services with internet NEGs, the first layer-GFEs select a second-layer external forwarding gateway colocated with the first-layer GFE. The forwarding gateway sends requests to the internet NEG endpoint. This concludes the request distribution process for internet NEGs.
   • For backend services with serverless NEGs and Private Service Connect (PSC) NEGs, and single-region backend buckets, first-layer GFEs select a second-layer GFE in the region matching the region of the NEG or bucket. For multi-region Cloud Storage buckets, first-layer GFEs select second-layer GFEs either in the region of the bucket, or a region as close as possible to the multi-region bucket (defined by network round trip time).
   • For backend services with instance groups, zonal NEGs with GCE_VM_IP_PORT endpoints, and hybrid NEGs, Google’s capacity management system informs first-layer GFEs about the used and configured capacity for each backend. The configured capacity for a backend is defined by the balancing mode, the target capacity of the balancing mode, and the capacity scaler.
     • Standard Tier: First-layer GFEs select a second layer GFE in the region containing the backends.
     • Premium Tier: First-layer GFEs select second-layer GFEs from a set of applicable regions. Applicable regions are all regions where backends have been configured, excluding those regions with configured backends having zero capacity. First-layer GFEs select the closest second-layer GFE in an applicable region (defined by network round-trip time). If backends are configured in two or more regions, first-layer GFEs can spill requests over to other applicable regions if a first-choice region is full. Spillover to other regions is possible when all backends in the first-choice region are at capacity.
3. Second layer GFEs select backends. Second-layer GFEs are located in zones of a region. They use the following process to select a backend:
   • For backend services with serverless NEGs, Private Service Connect NEGs, and backend buckets, second-layer GFEs forward requests to Google’s production systems. This concludes the request distribution process for these backends.

   • For backend services with instance groups, zonal NEGs with GCE_VM_IP_PORT endpoints, and hybrid NEGs, Google’s health check probe systems inform second-layer GFEs about the health check status of the backend instances or endpoints.

     Premium Tier only: If the second-layer GFE has no healthy backend instances or endpoints in its region, it might send requests to another second-layer GFE in a different applicable region with configured backends. Spillover between second-layer GFEs in different regions does not exhaust all possible region-to-region combinations. If you need to direct traffic away from backends in a particular region, instead of configuring backends to fail health checks, you should set the capacity scaler of the backend to zero so the first-layer GFE excludes the region during the previous step.

   The second-layer GFE then directs requests to backend instances or endpoints in zones within its region as discussed in the next step.

4. Second layer GFE selects a zone. By default, second-layer GFEs use the WATERFALL_BY_REGION algorithm where each second-layer GFE prefers to select backend instances or endpoints in the same zone as the zone that contains the second-layer GFE. Because WATERFALL_BY_REGION minimizes traffic between zones, at low request rates, each second-layer GFE might exclusively send requests to backends in the same zone as the second-layer GFE itself.

   For global external Application Load Balancers only, second-layer GFEs can be configured to use one of the following alternative algorithms by using a serviceLbPolicy:

   • SPRAY_TO_REGION: Second-layer GFEs do not prefer selecting backend instances or endpoints in the same zone as the second-layer GFE. Second-layer GFEs attempt to distribute traffic to all backend instances or endpoints in all zones of the region. This can lead to more even distribution of load at the expense of increased traffic between zones.
   • WATERFALL_BY_ZONE: Second-layer GFEs strongly prefer selecting backend instances or endpoints in the same zone as the second-layer GFE. Second-layer GFEs only direct requests to backends in different zones after all backends in the current zone have reached their configured capacities.
5. Second layer GFE selects instances or endpoints within the zone. By default, a second-layer GFE distributes requests among backends in a round-robin fashion. For global external Application Load Balancers only, you can change this by using a load balancing locality policy (localityLbPolicy). The load balancing locality policy applies only to backends within the selected zone discussed in the previous step.

Regional external Application Load Balancer

For regional external Application Load Balancers, traffic distribution is based on the load balancing mode and the load balancing locality policy.

The balancing mode determines the weight and fraction of traffic that should be sent to each group (instance group or NEG). The load balancing locality policy (LocalityLbPolicy) determines how backends within the group are load balanced.

When a backend service receives traffic, it first directs traffic to a backend (instance group or NEG) according to the backend's balancing mode. After a backend is selected, traffic is then distributed among instances or endpoints in that backend group according to the load balancing locality policy.

For more information, see the following:

• Balancing modes
• Load balancing locality policy (regional backend service API documentation).

Session affinity

Session affinity provides a best-effort attempt to send requests from a particular client to the same backend for as long as the backend is healthy and has the capacity, according to the configured balancing mode.

When you use session affinity, we recommend the RATE balancing mode rather than UTILIZATION. Session affinity works best if you set the balancing mode to requests per second (RPS).

External Application Load Balancers offer the following types of session affinity:

• NONE. Session affinity is not set for the load balancer.
• Client IP affinity
• Generated cookie affinity
• Header field affinity
• HTTP Cookie affinity
• Stateful cookie-based session affinity

The following table summarizes the supported session affinity options supported by external Application Load Balancers:

High availability and failover

High availability and failover in external Application Load Balancers can be configured at the load balancer level. This is handled by creating backup regional external Application Load Balancers in any region where you require backup.

The following table describes the failover behavior.

HTTP/2 support

HTTP/2 is a major revision of the HTTP/1 protocol. HTTP/2 is supported for connections between clients and the external Application Load Balancer, and for connections between the load balancer and its backends.

The load balancer automatically negotiates HTTP/2 with clients as part of the TLS handshake by using the ALPN TLS extension. Even if a load balancer is configured to use HTTPS, modern clients default to HTTP/2. This is controlled on the client side, not on the load balancer.

If a client doesn't support HTTP/2 and the load balancer is configured to use HTTP/2 between the load balancer and the backend instances, the load balancer might still negotiate an HTTPS connection or accept insecure HTTP requests. Those HTTPS or HTTP requests are then transformed by the load balancer to proxy the requests over HTTP/2 to the backend instances.

To use HTTP/2, you must enable TLS on your backends. For more information, see Encryption from the load balancer to the backends.

HTTP/2 max concurrent streams

The HTTP/2 SETTINGS_MAX_CONCURRENT_STREAMS setting describes the maximum number of streams that an endpoint accepts, initiated by the peer. The value advertised by an HTTP/2 client to a Google Cloud load balancer is effectively meaningless because the load balancer doesn't initiate streams to the client.

In cases where the load balancer uses HTTP/2 to communicate with a server that is running on a VM, the load balancer respects the SETTINGS_MAX_CONCURRENT_STREAMS value advertised by the server. If a value of zero is advertised, the load balancer can't forward requests to the server, and this might result in errors.

HTTP/2 limitations

• HTTP/2 between the load balancer and the instance can require significantly more TCP connections to the instance than HTTP(S). Connection pooling, an optimization that reduces the number of these connections with HTTP(S), is not currently available with HTTP/2. As a result, you might see high backend latencies because backend connections are made more frequently.
• HTTP/2 between the load balancer and the backend does not support running the WebSocket Protocol over a single stream of an HTTP/2 connection (RFC 8441).
• HTTP/2 between the load balancer and the backend does not support server push.
• The gRPC error rate and request volume aren't visible in the Google Cloud API or the Google Cloud console. If the gRPC endpoint returns an error, the load balancer logs and the monitoring data report the OK 200 HTTP response code.

HTTP/3 support

HTTP/3 is a next-generation internet protocol. It is built on top of IETF QUIC, a protocol developed from the original Google QUIC protocol. HTTP/3 is supported between the external Application Load Balancer, Cloud CDN, and clients.

Specifically:

• IETF QUIC is a transport layer protocol that provides congestion control and reliability similar to TCP, uses TLS 1.3 for security, and improved performance.
• HTTP/3 is an application layer built on top of IETF QUIC, and it relies on QUIC to handle multiplexing, congestion control, loss detection, and retransmission.
• HTTP/3 allows faster client connection initiation, eliminates head-of-line blocking in multiplexed streams, and supports connection migration when a client's IP address changes.
• HTTP/3 is supported for connections between clients and the load balancer, not connections between the load balancer and its backends.
• HTTP/3 connections use the BBR congestion control algorithm.

HTTP/3 on your load balancer can improve web page load times, reduce video rebuffering, and improve throughput on higher latency connections.

The following table specifies the HTTP/3 support for external Application Load Balancers in each mode.

How HTTP/3 is negotiated

When HTTP/3 is enabled, the load balancer advertises this support to clients, allowing clients that support HTTP/3 to attempt to establish HTTP/3 connections with the HTTPS load balancer.

• Properly implemented clients always fall back to HTTPS or HTTP/2 when they cannot establish an HTTP/3 connection.
• Clients that support HTTP/3 use their cached prior knowledge of HTTP/3 support to save unnecessary round-trips in the future.
• Because of this fallback, enabling or disabling HTTP/3 in the load balancer does not disrupt the load balancer's ability to connect to clients.

Support is advertised in the Alt-Svc HTTP response header. When HTTP/3 is enabled, responses from the load balancer include the following alt-svc header value:

alt-svc: h3=":443"; ma=2592000,h3-29=":443"; ma=2592000"

If HTTP/3 has been explicitly set to DISABLE, responses do not include an alt-svc response header.

When you have HTTP/3 enabled on your HTTPS load balancer, some circumstances can cause your client to fall back to HTTPS or HTTP/2 instead of negotiating HTTP/3. These include the following:

• When a client supports versions of HTTP/3 that are not compatible with the HTTP/3 versions supported by the HTTPS load balancer.
• When the load balancer detects that UDP traffic is blocked or rate-limited in a way that would prevent HTTP/3 from working.
• The client does not support HTTP/3 at all, and thus does not attempt to negotiate an HTTP/3 connection.

When a connection falls back to HTTPS or HTTP/2, we do not count this as a failure of the load balancer.

Before you enable HTTP/3, ensure that the previously described behaviors are acceptable for your workloads.

Configure HTTP/3

Both NONE (the default) and ENABLE enable HTTP/3 support for your load balancer.

When HTTP/3 is enabled, the load balancer advertises it to clients, which allows clients that support it to negotiate an HTTP/3 version with the load balancer. Clients that do not support HTTP/3 do not negotiate an HTTP/3 connection. You do not need to explicitly disable HTTP/3 unless you have identified broken client implementations.

External Application Load Balancers provide three ways to configure HTTP/3 as shown in the following table.

To explicitly enable (or disable) HTTP/3, follow these steps.

gcloud compute target-https-proxies create HTTPS_PROXY_NAME \
    --global \
    --quic-override=QUIC_SETTING

POST https://www.googleapis.com/v1/compute/projects/PROJECT_ID/global/targetHttpsProxies/TARGET_PROXY_NAME/setQuicOverride

{
  "quicOverride": QUIC_SETTING
}

WebSocket support

Google Cloud HTTP(S)-based load balancers support the websocket protocol when you use HTTP or HTTPS as the protocol to the backend. The load balancer does not require any configuration to proxy websocket connections.

The websocket protocol provides a full-duplex communication channel between clients and the load balancer. For more information, see RFC 6455.

The websocket protocol works as follows:

1. The load balancer recognizes a websocket Upgrade request from an HTTP(S) client. The request contains the Connection: Upgrade, Upgrade: websocket headers, followed by other relevant websocket related request headers.
2. Backend sends a websocket Upgrade response. The backend instance sends a 101 switching protocol response with Connection: Upgrade, Upgrade: websocket headers, and other other websocket related response headers.
3. The load balancer proxies bidirectional traffic for the duration of the current connection.

If the backend instance returns an error response with response code 426 or 502, the load balancer closes the connection.

Session affinity for websockets works the same as for any other request. For more information, see Session affinity.

gRPC support

gRPC is an open-source framework for remote procedure calls. It is based on the HTTP/2 standard. Use cases for gRPC include the following:

• Low-latency, highly scalable, distributed systems
• Developing mobile clients that communicate with a cloud server
• Designing new protocols that must be accurate, efficient, and language-independent
• Layered design to enable extension, authentication, and logging

To use gRPC with your Google Cloud applications, you must proxy requests end-to-end over HTTP/2. To do this:

1. Configure an HTTPS load balancer.
2. Enable HTTP/2 as the protocol from the load balancer to the backends.

The load balancer negotiates HTTP/2 with clients as part of the SSL handshake by using the ALPN TLS extension.

The load balancer may still negotiate HTTPS with some clients or accept insecure HTTP requests on a load balancer that is configured to use HTTP/2 between the load balancer and the backend instances. Those HTTP or HTTPS requests are transformed by the load balancer to proxy the requests over HTTP/2 to the backend instances.

You must enable TLS on your backends. For more information, see Encryption from the load balancer to the backends.

If you want to configure an external Application Load Balancer by using HTTP/2 with Google Kubernetes Engine Ingress or by using gRPC and HTTP/2 with Ingress, see HTTP/2 for load balancing with Ingress.

For information about troubleshooting problems with HTTP/2, see Troubleshooting issues with HTTP/2 to the backends.

For information about HTTP/2 limitations, see HTTP/2 limitations.

TLS support

By default, an HTTPS target proxy accepts only TLS 1.0, 1.1, 1.2, and 1.3 when terminating client SSL requests.

When the load balancer uses HTTPS as the backend service protocol, it can negotiate TLS 1.0, 1.1, 1.2, or 1.3 to the backend.

Mutual TLS support

Mutual TLS, or mTLS, is an industry standard protocol for mutual authentication between a client and a server. It ensures that both the client and server authenticate each other by verifying that each holds a valid certificate issued by a trusted certificate authority (CA). Unlike standard TLS, where only the server is authenticated, mTLS requires both the client and server to present certificates, confirming the identities of both parties before communication is established.

All of the Application Load Balancers support mTLS. With mTLS, the load balancer requests that the client send a certificate to authenticate itself during the TLS handshake with the load balancer. You can configure a Certificate Manager trust store that the load balancer then uses to validate the client certificate's chain of trust.

For more information about mTLS, see Mutual TLS authentication.

Limitations

• HTTPS load balancers do not send a close_notify closure alert when terminating SSL connections. That is, the load balancer closes the TCP connection instead of performing an SSL shutdown.
• HTTPS load balancers support only lowercase characters in domains in a common name (CN) attribute or a subject alternative name (SAN) attribute of the certificate. Certificates with uppercase characters in domains are returned only when set as the primary certificate in the target proxy.
• HTTPS load balancers do not use the Server Name Indication (SNI) extension when connecting to the backend, except for load balancers with Internet NEG backends. For more details, see Encryption from the load balancer to the backends.

• When using regional external Application Load Balancers with Cloud Run in a Shared VPC environment, standalone VPC networks in service projects can send traffic to any other Cloud Run services deployed in any other service projects within the same Shared VPC environment. This is a known issue and this form of access will be blocked in the future.

• Google Cloud doesn't guarantee that an underlying TCP connection can remain open for the entirety of the value of the backend service timeout. Client systems must implement retry logic instead of relying on a TCP connection to be open for long periods of time.

• You can't create a regional external Application Load Balancer in Premium Tier using the Google Cloud console. Additionally, only regions supporting Standard Tier are available for these load balancers in the Google Cloud console. Use either the gcloud CLI or the API instead. Use either the gcloud CLI or the API instead.

• Cloud CDN allows you to force an object or set of objects to be ignored by the cache by requesting a cache invalidation. When you're using a global external Application Load Balancer with Shared VPC cross-project service referencing, by default, service project admins won't have the required permissions to request cache invalidations. This is because cache invalidation is configured in the frontend project (that is, the project that has the forwarding rule, target proxy, and URL map of the load balancer). Thus, cache invalidations can only be issued by principals who have the IAM roles for configuring load balancer related resources in the frontend projects (for example, the Compute Network Admin role). Service admins, who control provisioning of the backend services in a separate project, will have to work with the load balancer admin of the frontend project to issue cache invalidation for their cross-project services.

What's next

• To learn how to deploy a global external Application Load Balancer, see Setting up an external Application Load Balancer with a Compute Engine backend.

• To learn how to deploy a regional external Application Load Balancer, see Setting up a regional external Application Load Balancer with a Compute Engine backend.

• If you are an existing user of the classic Application Load Balancer, make sure that you review Migration overview when you plan a new deployment with the global external Application Load Balancer.

• To learn how to automate your external Application Load Balancer setup with Terraform, see Terraform module examples for external Application Load Balancers.

• To learn how to configure advanced traffic management capabilities available with the global external Application Load Balancer, see Traffic management overview for global external Application Load Balancers.

• To learn how to configure advanced traffic management capabilities available with the regional external Application Load Balancer, see Traffic management overview for regional external Application Load Balancers.

• To learn about serving websites, see Serving websites.

• To find the locations for Google PoPs, see GFE locations.

• To learn about capacity management, see Capacity management with load balancing tutorial and Application capacity optimizations with global load balancing.

• To learn how to use Certificate Manager to provision and manage SSL certificates, see the Certificate Manager overview.

• To insert custom logic into the load balancing data path, configure Service Extensions plugins or callouts.

• For regional external Application Load Balancers only, you can try using Apigee Shadow API Discovery to find shadow APIs (also known as undocumented APIs) in your existing Google Cloud infrastructure. Make sure that you read the associated limitations before you enable Shadow API Discovery.