github.com/hobbeswalsh/terraform@v0.3.7-0.20150619183303-ad17cf55a0fa/website/source/docs/providers/google/r/compute_target_pool.html.markdown (about)

     1  ---
     2  layout: "google"
     3  page_title: "Google: google_compute_target_pool"
     4  sidebar_current: "docs-google-resource-target_pool"
     5  description: |-
     6    Manages a Target Pool within GCE.
     7  ---
     8  
     9  # google\_compute\_target\_pool
    10  
    11  Manages a Target Pool within GCE.  This is a collection of instances used as
    12  target of a network load balancer (Forwarding Rule).  For more information see
    13  [the official
    14  documentation](https://cloud.google.com/compute/docs/load-balancing/network/target-pools)
    15  and [API](https://cloud.google.com/compute/docs/reference/latest/targetPools).
    16  
    17  
    18  ## Example Usage
    19  
    20  ```
    21  resource "google_compute_target_pool" "default" {
    22      name = "test"
    23      instances = [ "us-central1-a/myinstance1", "us-central1-b/myinstance2" ]
    24      health_checks = [ "${google_compute_http_health_check.default.name}" ]
    25  }
    26  ```
    27  
    28  ## Argument Reference
    29  
    30  The following arguments are supported:
    31  
    32  * `backup_pool` - (Optional) URL to the backup target pool.  Must also set
    33    failover\_ratio.
    34  
    35  * `description` - (Optional) Textual description field.
    36  
    37  * `failover_ratio` - (Optional) Ratio (0 to 1) of failed nodes before using the
    38    backup pool (which must also be set).
    39  
    40  * `health_checks` - (Optional) List of zero or one healthcheck names.
    41  
    42  * `instances` - (Optional) List of instances in the pool.  They can be given as
    43    URLs, or in the form of "zone/name".  Note that the instances need not exist
    44    at the time of target pool creation, so there is no need to use the Terraform
    45    interpolators to create a dependency on the instances from the target pool.
    46  
    47  * `name` - (Required) A unique name for the resource, required by GCE.  Changing
    48    this forces a new resource to be created.
    49  
    50  * `session_affinity` - (Optional) How to distribute load.  Options are "NONE" (no affinity).  "CLIENT\_IP" (hash of the source/dest addresses / ports), and "CLIENT\_IP\_PROTO" also includes the protocol (default "NONE").
    51  
    52  
    53  ## Attributes Reference
    54  
    55  The following attributes are exported:
    56  
    57  * `self_link` - The URL of the created resource.
    58