» google_compute_backend_service

A Backend Service defines a group of virtual machines that will serve traffic for load balancing. This resource is a global backend service, appropriate for external load balancing or self-managed internal load balancing. For managed internal load balancing, use a regional backend service instead.

Currently self-managed internal load balancing is only available in beta.

To get more information about BackendService, see:

» Example Usage - Backend Service Basic

resource "google_compute_backend_service" "default" {
  name          = "backend-service"
  health_checks = [google_compute_http_health_check.default.self_link]
}

resource "google_compute_http_health_check" "default" {
  name               = "health-check"
  request_path       = "/"
  check_interval_sec = 1
  timeout_sec        = 1
}

» Example Usage - Backend Service Traffic Director Round Robin

resource "google_compute_backend_service" "default" {
  provider = google-beta

  name                  = "backend-service"
  health_checks         = [google_compute_health_check.health_check.self_link]
  load_balancing_scheme = "INTERNAL_SELF_MANAGED"
  locality_lb_policy    = "ROUND_ROBIN"
}

resource "google_compute_health_check" "health_check" {
  provider = google-beta

  name = "health-check"
  http_health_check {
    port = 80
  }
}

» Example Usage - Backend Service Traffic Director Ring Hash

resource "google_compute_backend_service" "default" {
  provider = google-beta

  name                  = "backend-service"
  health_checks         = [google_compute_health_check.health_check.self_link]
  load_balancing_scheme = "INTERNAL_SELF_MANAGED"
  locality_lb_policy    = "RING_HASH"
  session_affinity      = "HTTP_COOKIE"
  circuit_breakers {
    max_connections = 10
  }
  consistent_hash {
    http_cookie {
      ttl {
        seconds = 11
        nanos   = 1111
      }
      name = "mycookie"
    }
  }
  outlier_detection {
    consecutive_errors = 2
  }
}

resource "google_compute_health_check" "health_check" {
  provider = google-beta

  name = "health-check"
  http_health_check {
    port = 80
  }
}

» Argument Reference

The following arguments are supported:

  • health_checks - (Required) The set of URLs to the HttpHealthCheck or HttpsHealthCheck resource for health checking this BackendService. Currently at most one health check can be specified, and a health check is required. For internal load balancing, a URL to a HealthCheck resource must be specified instead.

  • name - (Required) Name of the resource. Provided by the client when the resource is created. The name must be 1-63 characters long, and comply with RFC1035. Specifically, the name must be 1-63 characters long and match the regular expression [a-z]([-a-z0-9]*[a-z0-9])? which means the first character must be a lowercase letter, and all following characters must be a dash, lowercase letter, or digit, except the last character, which cannot be a dash.


  • affinity_cookie_ttl_sec - (Optional) Lifetime of cookies in seconds if session_affinity is GENERATED_COOKIE. If set to 0, the cookie is non-persistent and lasts only until the end of the browser session (or equivalent). The maximum allowed value for TTL is one day. When the load balancing scheme is INTERNAL, this field is not used.

  • backend - (Optional) The set of backends that serve this BackendService. Structure is documented below.

  • circuit_breakers - (Optional, Beta) Settings controlling the volume of connections to a backend service. This field is applicable only when the load_balancing_scheme is set to INTERNAL_SELF_MANAGED. Structure is documented below.

  • consistent_hash - (Optional, Beta) Consistent Hash-based load balancing can be used to provide soft session affinity based on HTTP headers, cookies or other properties. This load balancing policy is applicable only for HTTP connections. The affinity to a particular destination host will be lost when one or more hosts are added/removed from the destination service. This field specifies parameters that control consistent hashing. This field only applies if the load_balancing_scheme is set to INTERNAL_SELF_MANAGED. This field is only applicable when locality_lb_policy is set to MAGLEV or RING_HASH. Structure is documented below.

  • cdn_policy - (Optional) Cloud CDN configuration for this BackendService. Structure is documented below.

  • connection_draining_timeout_sec - (Optional) Time for which instance will be drained (not accept new connections, but still work to finish started).

  • custom_request_headers - (Optional, Beta) Headers that the HTTP/S load balancer should add to proxied requests.

  • description - (Optional) An optional description of this resource.

  • enable_cdn - (Optional) If true, enable Cloud CDN for this BackendService.

  • iap - (Optional) Settings for enabling Cloud Identity Aware Proxy Structure is documented below.

  • load_balancing_scheme - (Optional) Indicates whether the backend service will be used with internal or external load balancing. A backend service created for one type of load balancing cannot be used with the other. Must be EXTERNAL or INTERNAL_SELF_MANAGED for a global backend service. Defaults to EXTERNAL.

  • locality_lb_policy - (Optional, Beta) The load balancing algorithm used within the scope of the locality. The possible values are - ROUND_ROBIN - This is a simple policy in which each healthy backend is selected in round robin order. LEAST_REQUEST - An O(1) algorithm which selects two random healthy hosts and picks the host which has fewer active requests. RING_HASH - The ring/modulo hash load balancer implements consistent hashing to backends. The algorithm has the property that the addition/removal of a host from a set of N hosts only affects 1/N of the requests. RANDOM - The load balancer selects a random healthy host. ORIGINAL_DESTINATION - Backend host is selected based on the client connection metadata, i.e., connections are opened to the same address as the destination address of the incoming connection before the connection was redirected to the load balancer. MAGLEV - used as a drop in replacement for the ring hash load balancer. Maglev is not as stable as ring hash but has faster table lookup build times and host selection times. For more information about Maglev, refer to https://ai.google/research/pubs/pub44824 This field is applicable only when the load_balancing_scheme is set to INTERNAL_SELF_MANAGED.

  • outlier_detection - (Optional, Beta) Settings controlling eviction of unhealthy hosts from the load balancing pool. This field is applicable only when the load_balancing_scheme is set to INTERNAL_SELF_MANAGED. Structure is documented below.

  • port_name - (Optional) Name of backend port. The same name should appear in the instance groups referenced by this service. Required when the load balancing scheme is EXTERNAL.

  • protocol - (Optional) The protocol this BackendService uses to communicate with backends. Possible values are HTTP, HTTPS, HTTP2, TCP, and SSL. The default is HTTP. NOTE: HTTP2 is only valid for beta HTTP/2 load balancer types and may result in errors if used with the GA API.

  • security_policy - (Optional) The security policy associated with this backend service.

  • session_affinity - (Optional) Type of session affinity to use. The default is NONE. Session affinity is not applicable if the protocol is UDP.

  • timeout_sec - (Optional) How many seconds to wait for the backend before considering it a failed request. Default is 30 seconds. Valid range is [1, 86400].

  • log_config - (Optional, Beta) This field denotes the logging options for the load balancer traffic served by this backend service. If logging is enabled, logs will be exported to Stackdriver. Structure is documented below.

  • project - (Optional) The ID of the project in which the resource belongs. If it is not provided, the provider project is used.

The backend block supports:

  • balancing_mode - (Optional) Specifies the balancing mode for this backend. For global HTTP(S) or TCP/SSL load balancing, the default is UTILIZATION. Valid values are UTILIZATION, RATE (for HTTP(S)) and CONNECTION (for TCP/SSL).

  • capacity_scaler - (Optional) A multiplier applied to the group's maximum servicing capacity (based on UTILIZATION, RATE or CONNECTION). Default value is 1, which means the group will serve up to 100% of its configured capacity (depending on balancingMode). A setting of 0 means the group is completely drained, offering 0% of its available Capacity. Valid range is [0.0,1.0].

  • description - (Optional) An optional description of this resource. Provide this property when you create the resource.

  • group - (Required) The fully-qualified URL of an Instance Group or Network Endpoint Group resource. In case of instance group this defines the list of instances that serve traffic. Member virtual machine instances from each instance group must live in the same zone as the instance group itself. No two backends in a backend service are allowed to use same Instance Group resource. For Network Endpoint Groups this defines list of endpoints. All endpoints of Network Endpoint Group must be hosted on instances located in the same zone as the Network Endpoint Group. Backend services cannot mix Instance Group and Network Endpoint Group backends. Note that you must specify an Instance Group or Network Endpoint Group resource using the fully-qualified URL, rather than a partial URL.

  • max_connections - (Optional) The max number of simultaneous connections for the group. Can be used with either CONNECTION or UTILIZATION balancing modes. For CONNECTION mode, either maxConnections or one of maxConnectionsPerInstance or maxConnectionsPerEndpoint, as appropriate for group type, must be set.

  • max_connections_per_instance - (Optional) The max number of simultaneous connections that a single backend instance can handle. This is used to calculate the capacity of the group. Can be used in either CONNECTION or UTILIZATION balancing modes. For CONNECTION mode, either maxConnections or maxConnectionsPerInstance must be set.

  • max_connections_per_endpoint - (Optional) The max number of simultaneous connections that a single backend network endpoint can handle. This is used to calculate the capacity of the group. Can be used in either CONNECTION or UTILIZATION balancing modes. For CONNECTION mode, either maxConnections or maxConnectionsPerEndpoint must be set.

  • max_rate - (Optional) The max requests per second (RPS) of the group. Can be used with either RATE or UTILIZATION balancing modes, but required if RATE mode. For RATE mode, either maxRate or one of maxRatePerInstance or maxRatePerEndpoint, as appropriate for group type, must be set.

  • max_rate_per_instance - (Optional) The max requests per second (RPS) that a single backend instance can handle. This is used to calculate the capacity of the group. Can be used in either balancing mode. For RATE mode, either maxRate or maxRatePerInstance must be set.

  • max_rate_per_endpoint - (Optional) The max requests per second (RPS) that a single backend network endpoint can handle. This is used to calculate the capacity of the group. Can be used in either balancing mode. For RATE mode, either maxRate or maxRatePerEndpoint must be set.

  • max_utilization - (Optional) Used when balancingMode is UTILIZATION. This ratio defines the CPU utilization target for the group. The default is 0.8. Valid range is [0.0, 1.0].

The circuit_breakers block supports:

  • connect_timeout - (Optional) The timeout for new network connections to hosts. Structure is documented below.

  • max_requests_per_connection - (Optional) Maximum requests for a single backend connection. This parameter is respected by both the HTTP/1.1 and HTTP/2 implementations. If not specified, there is no limit. Setting this parameter to 1 will effectively disable keep alive.

  • max_connections - (Optional) The maximum number of connections to the backend cluster. Defaults to 1024.

  • max_pending_requests - (Optional) The maximum number of pending requests to the backend cluster. Defaults to 1024.

  • max_requests - (Optional) The maximum number of parallel requests to the backend cluster. Defaults to 1024.

  • max_retries - (Optional) The maximum number of parallel retries to the backend cluster. Defaults to 3.

The connect_timeout block supports:

  • seconds - (Required) Span of time at a resolution of a second. Must be from 0 to 315,576,000,000 inclusive.

  • nanos - (Optional) Span of time that's a fraction of a second at nanosecond resolution. Durations less than one second are represented with a 0 seconds field and a positive nanos field. Must be from 0 to 999,999,999 inclusive.

The consistent_hash block supports:

  • http_cookie - (Optional) Hash is based on HTTP Cookie. This field describes a HTTP cookie that will be used as the hash key for the consistent hash load balancer. If the cookie is not present, it will be generated. This field is applicable if the sessionAffinity is set to HTTP_COOKIE. Structure is documented below.

  • http_header_name - (Optional) The hash based on the value of the specified header field. This field is applicable if the sessionAffinity is set to HEADER_FIELD.

  • minimum_ring_size - (Optional) The minimum number of virtual nodes to use for the hash ring. Larger ring sizes result in more granular load distributions. If the number of hosts in the load balancing pool is larger than the ring size, each host will be assigned a single virtual node. Defaults to 1024.

The http_cookie block supports:

  • ttl - (Optional) Lifetime of the cookie. Structure is documented below.

  • name - (Optional) Name of the cookie.

  • path - (Optional) Path to set for the cookie.

The ttl block supports:

  • seconds - (Required) Span of time at a resolution of a second. Must be from 0 to 315,576,000,000 inclusive.

  • nanos - (Optional) Span of time that's a fraction of a second at nanosecond resolution. Durations less than one second are represented with a 0 seconds field and a positive nanos field. Must be from 0 to 999,999,999 inclusive.

The cdn_policy block supports:

  • cache_key_policy - (Optional) The CacheKeyPolicy for this CdnPolicy. Structure is documented below.

  • signed_url_cache_max_age_sec - (Optional) Maximum number of seconds the response to a signed URL request will be considered fresh, defaults to 1hr (3600s). After this time period, the response will be revalidated before being served. When serving responses to signed URL requests, Cloud CDN will internally behave as though all responses from this backend had a "Cache-Control: public, max-age=[TTL]" header, regardless of any existing Cache-Control header. The actual headers served in responses will not be altered.

The cache_key_policy block supports:

  • include_host - (Optional) If true requests to different hosts will be cached separately.

  • include_protocol - (Optional) If true, http and https requests will be cached separately.

  • include_query_string - (Optional) If true, include query string parameters in the cache key according to query_string_whitelist and query_string_blacklist. If neither is set, the entire query string will be included. If false, the query string will be excluded from the cache key entirely.

  • query_string_blacklist - (Optional) Names of query string parameters to exclude in cache keys. All other parameters will be included. Either specify query_string_whitelist or query_string_blacklist, not both. '&' and '=' will be percent encoded and not treated as delimiters.

  • query_string_whitelist - (Optional) Names of query string parameters to include in cache keys. All other parameters will be excluded. Either specify query_string_whitelist or query_string_blacklist, not both. '&' and '=' will be percent encoded and not treated as delimiters.

The iap block supports:

The outlier_detection block supports:

  • base_ejection_time - (Optional) The base time that a host is ejected for. The real time is equal to the base time multiplied by the number of times the host has been ejected. Defaults to 30000ms or 30s. Structure is documented below.

  • consecutive_errors - (Optional) Number of errors before a host is ejected from the connection pool. When the backend host is accessed over HTTP, a 5xx return code qualifies as an error. Defaults to 5.

  • consecutive_gateway_failure - (Optional) The number of consecutive gateway failures (502, 503, 504 status or connection errors that are mapped to one of those status codes) before a consecutive gateway failure ejection occurs. Defaults to 5.

  • enforcing_consecutive_errors - (Optional) The percentage chance that a host will be actually ejected when an outlier status is detected through consecutive 5xx. This setting can be used to disable ejection or to ramp it up slowly. Defaults to 100.

  • enforcing_consecutive_gateway_failure - (Optional) The percentage chance that a host will be actually ejected when an outlier status is detected through consecutive gateway failures. This setting can be used to disable ejection or to ramp it up slowly. Defaults to 0.

  • enforcing_success_rate - (Optional) The percentage chance that a host will be actually ejected when an outlier status is detected through success rate statistics. This setting can be used to disable ejection or to ramp it up slowly. Defaults to 100.

  • interval - (Optional) Time interval between ejection sweep analysis. This can result in both new ejections as well as hosts being returned to service. Defaults to 10 seconds. Structure is documented below.

  • max_ejection_percent - (Optional) Maximum percentage of hosts in the load balancing pool for the backend service that can be ejected. Defaults to 10%.

  • success_rate_minimum_hosts - (Optional) The number of hosts in a cluster that must have enough request volume to detect success rate outliers. If the number of hosts is less than this setting, outlier detection via success rate statistics is not performed for any host in the cluster. Defaults to 5.

  • success_rate_request_volume - (Optional) The minimum number of total requests that must be collected in one interval (as defined by the interval duration above) to include this host in success rate based outlier detection. If the volume is lower than this setting, outlier detection via success rate statistics is not performed for that host. Defaults to 100.

  • success_rate_stdev_factor - (Optional) This factor is used to determine the ejection threshold for success rate outlier ejection. The ejection threshold is the difference between the mean success rate, and the product of this factor and the standard deviation of the mean success rate: mean - (stdev * success_rate_stdev_factor). This factor is divided by a thousand to get a double. That is, if the desired factor is 1.9, the runtime value should be 1900. Defaults to 1900.

The base_ejection_time block supports:

  • seconds - (Required) Span of time at a resolution of a second. Must be from 0 to 315,576,000,000 inclusive.

  • nanos - (Optional) Span of time that's a fraction of a second at nanosecond resolution. Durations less than one second are represented with a 0 seconds field and a positive nanos field. Must be from 0 to 999,999,999 inclusive.

The interval block supports:

  • seconds - (Required) Span of time at a resolution of a second. Must be from 0 to 315,576,000,000 inclusive.

  • nanos - (Optional) Span of time that's a fraction of a second at nanosecond resolution. Durations less than one second are represented with a 0 seconds field and a positive nanos field. Must be from 0 to 999,999,999 inclusive.

The log_config block supports:

  • enable - (Optional) Whether to enable logging for the load balancer traffic served by this backend service.

  • sample_rate - (Optional) This field can only be specified if logging is enabled for this backend service. The value of the field must be in [0, 1]. This configures the sampling rate of requests to the load balancer where 1.0 means all logged requests are reported and 0.0 means no logged requests are reported. The default value is 1.0.

» Attributes Reference

In addition to the arguments listed above, the following computed attributes are exported:

  • creation_timestamp - Creation timestamp in RFC3339 text format.

  • fingerprint - Fingerprint of this resource. A hash of the contents stored in this object. This field is used in optimistic locking.

  • self_link - The URI of the created resource.

» Timeouts

This resource provides the following Timeouts configuration options:

  • create - Default is 4 minutes.
  • update - Default is 4 minutes.
  • delete - Default is 4 minutes.

» Import

BackendService can be imported using any of these accepted formats:

$ terraform import google_compute_backend_service.default projects/{{project}}/global/backendServices/{{name}}
$ terraform import google_compute_backend_service.default {{project}}/{{name}}
$ terraform import google_compute_backend_service.default {{name}}

» User Project Overrides

This resource supports User Project Overrides.