F5BigCnePool

Overview

F5BigCnePool is a namespaced Custom Resource in the k8s.f5net.com API group. It represents a Network Function (NF) pool: a list of NF endpoints (members) plus optional health monitors.

CR Parameters

metadata

Parameter	Type	Description
name	string	Unique name of the custom resource.
namespace	string	Namespace where the resource is created (because the CRD is namespaced).

spec

spec defines pool behavior and membership.

Parameter	Type	Description	Default
minActiveMembers	integer	Minimum number of members that must be available for a priority group to activate. Minimum 0, maximum 65535.	0
vrf	string	The name of the VRF (tenant) all pool members are in. It is a soft-reference to a VRF CR. If not specified, the pool members will exist in the default route domain. Maximum length 255.	""
members	array(object)	List of server endpoints in this pool. Each item requires address and port.	[]
members.address	string	IPv4 or IPv6 address of this server endpoint. Max length 46, min length 7. Must match one of the IPv4/IPv6 patterns (no prefix).
members.port	integer	Port for the monitor to check. Minimum 0, maximum 65535.
members.priorityGroup	integer	Priority group assignment for this pool member. Higher priority groups are used first. Minimum 0, maximum 8.	0
dynamicMembers	array(object)	Specifies a list of kubernetes services containing endpoints. Any endpoints associated with the specified services will be dynamically added to this pool as members, once for each target port in the service.	[]
dynamicMembers.serviceNamespace	string	The namespace of the kubernetes service from which to discover endpoints dynamically. If this field is empty, it defaults to the same namespace as the pool. If the referenced service is in a different namespace than the pool, it MUST be given the annotation "k8s.f5.com/allow-cross-namespace-refs"="true".	""
dynamicMembers.serviceName	string	The name of the kubernetes service from which to discover endpoints dynamically.
monitors	object	Monitor configuration for checking pool members at a set interval.	{}
monitors.http2	array(object)	HTTP/2 monitor configurations.	[]
monitors.http2.interval	integer	Frequency (seconds) when resource is down/unknown. Format int32. Minimum 0, maximum 4294967295.	5
monitors.http2.timeout	integer	Seconds target has to respond. Must be ≤ interval (documented). Minimum 0, maximum 4294967295.	16
monitors.http2.upInterval	integer	Frequency (seconds) when resource is up. 0 means disabled (use interval). Minimum 0, maximum 4294967295.	0
monitors.http2.timeUntilUp	integer	Seconds after first successful response before node is marked up. Minimum 0, maximum 4294967295.	0
monitors.http2.sendString	string	String sent to the target. Maximum length 64000.	"GET / HTTP/1.1"
monitors.http2.receiveString	string	Regex indicating target is up. Maximum length 64000.	'200'
monitors.http2.receiveDisableString	string	Regex indicating target should be disabled. Maximum length 64000.	''
monitors.http2.serversslProfileName	string	Server-side SSL profile used by the HTTP/2 monitor. Maximum length 255.	''
monitors.http	array(object)	HTTP monitor configurations.	[]
monitors.http.interval	integer	Frequency (seconds) when resource is down/unknown. Format int64. Minimum 0, maximum 4294967295.	5
monitors.http.timeout	integer	Seconds target has to respond. Minimum 0, maximum 4294967295.	16
monitors.http.sendString	string	String sent to the target. Maximum length 64000.	"GET / HTTP/1.0\r\n\r\n"
monitors.http.receiveString	string	Regex indicating target is up. Maximum length 64000.	'200'
monitors.http.receiveDisableString	string	Regex indicating target should be disabled. Maximum length 64000.	''
monitors.http.username	string	Username for authenticating to the target. Maximum length 255.	''
monitors.http.password	string	Password for authenticating to the target. Maximum length 255.	''
monitors.icmp	array(object)	ICMP monitor configurations.	[]
monitors.icmp.interval	integer	Frequency (seconds) when resource is down/unknown. Format int32. Minimum 0, maximum 4294967295.	5
monitors.icmp.timeout	integer	Seconds target has to respond. Must be ≤ interval (documented). Minimum 0, maximum 4294967295.	16
monitors.tcp	array(object)	TCP monitor configurations.	[]
monitors.tcp.interval	integer	Frequency (seconds) when resource is down/unknown. Format int32. Minimum 0, maximum 4294967295.	5
monitors.tcp.timeout	integer	Seconds target has to respond. Must be ≤ interval (documented). Minimum 0, maximum 4294967295.	16
monitors.tcp.sendString	string	String sent to the target. Maximum length 64000.	''
monitors.tcp.receiveString	string	Regex indicating target is up. Maximum length 64000.	''
monitors.tcp.receiveDisableString	string	Regex indicating target should be disabled. Maximum length 64000.	''
monitors.dns	array(object)	DNS monitor configurations.	[]
monitors.dns.interval	integer	Frequency (seconds) when resource is down/unknown. Minimum 0, maximum 4294967295.	5
monitors.dns.timeout	integer	Seconds target has to respond. Must be ≤ interval (documented). Minimum 0, maximum 4294967295.	16
monitors.dns.upInterval	integer	Frequency (seconds) when resource is up. 0 means disabled (use interval). Minimum 0, maximum 4294967295.	0
monitors.dns.timeUntilUp	integer	Seconds after first successful response before node is marked up. Minimum 0, maximum 4294967295.	0
monitors.dns.aliasAddress	string	Destination IP of the monitor. IPv4/IPv6 (no prefix). Minimum length 2, maximum length 39.	"::"
monitors.dns.aliasPort	integer	Destination port of the monitor. Minimum 0, maximum 65535.	0
monitors.dns.reverse	boolean	Reverse mode: successful receive match marks object down instead of up.	false
monitors.dns.recursionDesired	boolean	Sets the RD flag in query packets (required for monitoring recursive DNS servers).	false
monitors.dns.queryName	string	Domain name to query. Maximum length 255.
monitors.dns.queryType	string	DNS query type. Enum: a, aaaa.	a
monitors.dns.answerContains	string	Record types required in the answer section to mark up. Enum: query-type, any-type, anything.	query-type
monitors.dns.acceptRcode	string	RCODE required in the response for an up status. Enum: no-error, anything.	no-error
monitors.dns.recvAddress	string	Address the monitor looks for in DNS response RR sections. OneOf: empty/IPv4/IPv6. Minimum length 2, maximum length 39.	"::"
monitors.inband	array(object)	Passive monitoring based on actual traffic.	[]
monitors.inband.failures	integer	Number of failed attempts before marking a pool member down. Minimum 0, maximum 4294967295.	5
monitors.inband.failureInterval	integer	Interval (seconds) in which failures are counted. Minimum 0, maximum 4294967295.	30
monitors.inband.responseTime	integer	Interval (seconds) in which a pool member must respond with data. Minimum 0, maximum 4294967295.	10
monitors.inband.retryTime	integer	Wait time (seconds) before retrying whether a pool member is up. Minimum 0, maximum 4294967295.	300
monitors.tcpHalfOpen	array(object)	TCP half-open monitor configurations.	[]
monitors.tcpHalfOpen.interval	integer	Frequency (seconds) when resource is down/unknown. Format int64. Minimum 0, maximum 4294967295.	5
monitors.tcpHalfOpen.timeout	integer	Seconds target has to respond. Minimum 0, maximum 4294967295.	16

CR Example

apiVersion: k8s.f5net.com/v1
kind: F5BigCnePool
metadata:
  name: example-nf-pool
  namespace: "cnf-gateway"
spec:
  minActiveMembers: 1
  members:
  - address: "192.0.2.10"
    port: 8080
    priorityGroup: 1
  - address: "2001:db8::10"
    port: 8080
    priorityGroup: 0
  monitors:
    http:
    - interval: 5
      timeout: 16
      sendString: "GET /health HTTP/1.0\r\n\r\n"
      receiveString: "200"