authenticatorfactory

Documentation

Index

• func NewFromTokens(tokens map[string]*user.DefaultInfo, audiences authenticator.Audiences) authenticator.Request
• func RecordRequestLatency(ctx context.Context, code string, latency float64)
• func RecordRequestTotal(ctx context.Context, code string)
• type DelegatingAuthenticatorConfig
  • func (c DelegatingAuthenticatorConfig) New() (authenticator.Request, *spec.SecurityDefinitions, error)
• type RequestHeaderConfig

Functions

func NewFromTokens

func NewFromTokens(tokens map[string]*user.DefaultInfo, audiences authenticator.Audiences) authenticator.Request

NewFromTokens returns an authenticator.Request or an error

func RecordRequestLatency

func RecordRequestLatency(ctx context.Context, code string, latency float64)

RecordRequestLatency measures request latency in seconds for the delegated authentication. Broken down by status code.

func RecordRequestTotal

func RecordRequestTotal(ctx context.Context, code string)

RecordRequestTotal increments the total number of requests for the delegated authentication.

Types

type DelegatingAuthenticatorConfig

type DelegatingAuthenticatorConfig struct {
	Anonymous bool

	// TokenAccessReviewClient is a client to do token review. It can be nil. Then every token is ignored.
	TokenAccessReviewClient authenticationclient.AuthenticationV1Interface

	// TokenAccessReviewTimeout specifies a time limit for requests made by the authorization webhook client.
	TokenAccessReviewTimeout time.Duration

	// WebhookRetryBackoff specifies the backoff parameters for the authentication webhook retry logic.
	// This allows us to configure the sleep time at each iteration and the maximum number of retries allowed
	// before we fail the webhook call in order to limit the fan out that ensues when the system is degraded.
	WebhookRetryBackoff *wait.Backoff

	// CacheTTL is the length of time that a token authentication answer will be cached.
	CacheTTL time.Duration

	// CAContentProvider are the options for verifying incoming connections using mTLS and directly assigning to users.
	// Generally this is the CA bundle file used to authenticate client certificates
	// If this is nil, then mTLS will not be used.
	ClientCertificateCAContentProvider dynamiccertificates.CAContentProvider

	APIAudiences authenticator.Audiences

	RequestHeaderConfig *RequestHeaderConfig
}

DelegatingAuthenticatorConfig is the minimal configuration needed to create an authenticator built to delegate authentication to a kube API server

func (DelegatingAuthenticatorConfig) New

func (c DelegatingAuthenticatorConfig) New() (authenticator.Request, *spec.SecurityDefinitions, error)

type RequestHeaderConfig

type RequestHeaderConfig struct {
	// UsernameHeaders are the headers to check (in order, case-insensitively) for an identity. The first header with a value wins.
	UsernameHeaders headerrequest.StringSliceProvider
	// GroupHeaders are the headers to check (case-insensitively) for a group names.  All values will be used.
	GroupHeaders headerrequest.StringSliceProvider
	// ExtraHeaderPrefixes are the head prefixes to check (case-insentively) for filling in
	// the user.Info.Extra.  All values of all matching headers will be added.
	ExtraHeaderPrefixes headerrequest.StringSliceProvider
	// CAContentProvider the options for verifying incoming connections using mTLS.  Generally this points to CA bundle file which is used verify the identity of the front proxy.
	//	It may produce different options at will.
	CAContentProvider dynamiccertificates.CAContentProvider
	// AllowedClientNames is a list of common names that may be presented by the authenticating front proxy.  Empty means: accept any.
	AllowedClientNames headerrequest.StringSliceProvider
}