Skip to main content

dtcloud_lb_balancing_pool (Resource)

Builds a listener, its pool, a health monitor and the members in a single call. A create-only shortcut; prefer the granular resources.

Example Usage​

# The whole chain in one call: listener, pool, health monitor and members.
# Every argument is ForceNew -- changing one rebuilds all four objects.
resource "dtcloud_lb_balancing_pool" "web" {
  lb_id = dtcloud_lb.public.id

  # Front end
  protocol      = "HTTP"
  protocol_port = 80

  # Back end
  backend_protocol      = "HTTP"
  backend_protocol_port = 8080
  lb_algorithm          = "ROUND_ROBIN"

  # Health monitor. `protocol` here is the probe type, not a network protocol.
  health_monitor {
    protocol            = "HTTP"
    url_path            = "/healthz"
    interval            = 10
    timeout             = 5
    healthy_threshold   = 2
    unhealthy_threshold = 3
  }

  member {
    compute_server_id = dtcloud_vm.web.id
    address           = dtcloud_vm.web.primary_ip
    protocol_port     = 8080
  }
}
danger

Every argument here forces a new resource. There is no update endpoint for the chain as a unit, so changing the probe interval recreates the listener and drops the members with it. Prefer dtcloud_lb_listener, dtcloud_lb_pool, dtcloud_lb_member and dtcloud_lb_health_monitor unless you specifically want the atomic shape, and never mix the two styles on one load balancer.

Schema​

Required​

  • backend_protocol (String) Protocol used to reach the members. Changing this forces a new resource to be created.
  • backend_protocol_port (Number) Port used to reach the members. Changing this forces a new resource to be created.
  • healthy_threshold (Number) Consecutive successes before a member counts as healthy. Changing this forces a new resource to be created.
  • interval (Number) Seconds between probes (5-300). Changing this forces a new resource to be created.
  • lb_algorithm (String) Balancing algorithm. Changing this forces a new resource to be created.
  • lb_id (String) ID of the load balancer. Changing this forces a new resource to be created.
  • lb_port (Number) Front-end port. Changing this forces a new resource to be created.
  • lb_protocol (String) Front-end protocol. One of: HTTP, HTTPS, TCP, UDP, TERMINATED_HTTPS. Changing this forces a new resource to be created.
  • sticky_session (Boolean) Whether to pin a client to the member it first reached. Changing this forces a new resource to be created.
  • timeout (Number) Seconds to wait for a probe response (5-60). Changing this forces a new resource to be created.
  • type (String) Health probe type. Changing this forces a new resource to be created.
  • unhealthy_threshold (Number) Consecutive failures before a member counts as unhealthy. Changing this forces a new resource to be created.

Optional​

  • allowed_cidrs (List of String) CIDRs permitted to reach the listener. Changing this forces a new resource to be created.
  • certificate (String, Sensitive) PEM certificate for TERMINATED_HTTPS. Changing this forces a new resource to be created.
  • connection_limit (Number) Maximum number of concurrent connections. -1 means no limit. Changing this forces a new resource to be created.
  • insert_headers (Block List, Max: 1) Headers the listener injects into requests forwarded to members. Changing this forces a new resource to be created. (see below for nested schema)
  • member (Block List) Back-end members created together with the pool. Changing this forces a new resource to be created. (see below for nested schema)
  • private_key (String, Sensitive) PEM private key for TERMINATED_HTTPS. Changing this forces a new resource to be created.
  • timeout_client_data (Number) Client inactivity timeout in SECONDS (note: dtcloud_lb_listener takes milliseconds). Changing this forces a new resource to be created.
  • timeout_member_connect (Number) Member connect timeout in SECONDS (note: dtcloud_lb_listener takes milliseconds). Changing this forces a new resource to be created.
  • timeout_member_data (Number) Member inactivity timeout in SECONDS (note: dtcloud_lb_listener takes milliseconds). Changing this forces a new resource to be created.
  • timeouts (Block, Optional) (see below for nested schema)
  • url_path (String) Path requested by HTTP and HTTPS probes. Changing this forces a new resource to be created.

Read-Only​

  • balancing_pool_id (String) ID of the balancing pool on its own, without the load balancer prefix.
  • health_monitor_id (String) ID of the health monitor the platform created.
  • id (String) The ID of this resource.
  • listener_id (String) ID of the listener the platform created.
  • members_state (List of String) Operating state of each member.
  • members_total (Number) Number of members in the pool.
  • status (String) Status reported by the platform.

Nested Schema for insert_headers​

Optional:

  • x_forwarded_for (String) Changing this forces a new resource to be created.
  • x_forwarded_port (String) Changing this forces a new resource to be created.
  • x_forwarded_proto (String) Changing this forces a new resource to be created.
  • x_ssl_client_cn (String) Changing this forces a new resource to be created.
  • x_ssl_client_dn (String) Changing this forces a new resource to be created.
  • x_ssl_client_has_cert (String) Changing this forces a new resource to be created.
  • x_ssl_client_not_after (String) Changing this forces a new resource to be created.
  • x_ssl_client_not_before (String) Changing this forces a new resource to be created.
  • x_ssl_client_sha1 (String) Changing this forces a new resource to be created.
  • x_ssl_client_verify (String) Changing this forces a new resource to be created.
  • x_ssl_issuer (String) Changing this forces a new resource to be created.

Nested Schema for member​

Required:

  • address (String) IP address of the member. Changing this forces a new resource to be created.
  • compute_server_id (String) ID of the VM behind this address. Changing this forces a new resource to be created.
  • protocol_port (Number) Port the member listens on. Changing this forces a new resource to be created.

Nested Schema for timeouts​

Optional:

  • create (String)
  • delete (String)

Behaviour worth knowing​

  • The three listener timings are seconds here, where dtcloud_lb_listener takes milliseconds. That asymmetry belongs to the API: this endpoint converts, the listener endpoint does not.
  • Left unset, the platform fills several values in -- the timings, url_path and connection_limit -- and reports what it chose. They are Computed for that reason; without it an omitted value would disagree with the report and, being ForceNew, propose a rebuild on every plan.
  • health_monitor.protocol is the probe type, not a network protocol.

Timeouts​

create and delete each default to 30 minutes. There is no update.

Import​

# <load-balancer-id>:<balancing-pool-id>. Import one only when it declares no
# `member` blocks: a member's port is reported by nothing, so declared members
# come back empty and the first plan proposes a rebuild.
terraform import dtcloud_lb_balancing_pool.web 3d372446-4c75-4327-9511-98442456162f:5b2a9c31-7e04-4d88-9c6e-44a2f1d0e7b3

member blocks do not survive an import: nothing reports a member's port. Import a balancing pool only when it declares no members, or accept that the first apply rebuilds it.