Page tree
Skip to end of metadata
Go to start of metadata

Namespace: urn:mace:shibboleth:2.0:metadata
Schema: http://shibboleth.net/schema/idp/shibboleth-metadata.xsd

Overview

The DynamicHTTPMetadataProvider fetches entity metadata just-in-time from a remote HTTP server. The metadata request URL is constructed by applying a transform to the entityID. The transform strategy is configurable, with a simple way to configure support for the Metadata Query Protocol.

Metadata is cached in memory subject to a complex set of interacting settings and the cache indicators within the metadata itself, and also can be saved to disk and reloaded back into memory at reload or startup time to restore the state of the cache. This isn't a fully redundant safety net but can be used as part of an overall strategy to reduce the risk of relying on remote sources in real-time. Ultimately, remote sources have to be bulletproof or there will be outages. This can be mitigated but not fully eliminated as a risk.

Use this provider with remote metadata

The DynamicHTTPMetadataProvider is used with remote metadata. See the MetadataManagementBestPractices topic for more information.

Reference

NameTypeDefaultDescription
maxConnectionsTotalInteger100The maximum total number of simultaneous connections allowed by the HTTP client's connection pool manager. This attribute is incompatible with httpClientRef.
maxConnectionsPerRouteInteger100The maximum number of simultaneous connections per route allowed by the HTTP client's connection pool manager. This attribute is incompatible with httpClientRef.

supportedContentTypes

Comma-delimited Strings"application/samlmetadata+xml, application/xml, text/xml"

The MIME types supported by this provider when requesting metadata from the HTTP server. The Content-Type response header is validated against this list. This value cannot be specified as a bean property.

NameCardinalityDescription
<TLSTrustEngine>0 or 1A custom TrustEngine used to evaluate TLS server certificates. This element conflicts with and is overridden by the httpClientSecurityParametersRef attribute. It contains a single <security:TrustEngine> element.
<MetadataQueryProtocol>


0 or 1

Constructs the metadata request URL based on the requirements of the Metadata Query Protocol
<Template>Constructs the metadata request URL by means of a simple transform based on substitution
<Regex>Constructs the metadata request URL by means of a complex transform based on a regular expression

At most one of the <MetadataQueryProtocol><Template>, or <Regex> child elements is allowed. If none are configured, the provider constructs the metadata request URL directly from the entityID. This corresponds to the "well-known location" mechanism defined in the SAML 2.0 Metadata specification, section 4.1.

Don't forget to configure a child element

If you forget to configure a child element, the provider will default to the well-known location strategy. This constrains the entityID to be an URL (not an URN) but the provider does not check the URL scheme. If the scheme on the entityID is "http:", the metadata exchange will be vulnerable to a man-in-the-middle attack. For this reason, the well-known location strategy should be avoided in most cases.

<MetadataQueryProtocol>  Child Element

If the <MetadataQueryProtocol> child element is used, the metadata request URL is constructed according to the SAML Profile for the Metadata Query Protocol, which itself is based on the Metadata Query Protocol specification. The content of the <MetadataQueryProtocol> child element will be used as the "Base URL" defined in that specification.

The <MetadataQueryProtocol> child element has the following optional attribute:

NameTypeDescription
transformRefBean IDA reference to a transform function for the entityID. If used, the child element must be empty.

<Template> Child Element

If the <Template> child element is used, the metadata request URL is constructed by means of a simple transform. Specifically, the value of the entityID is substituted into the template parameter "${entityID}".

The <Template> child element has the following attributes:

NameTypeDefaultDescription
encodingStyle

"none", "form", "path", or "fragment"

"form"

Determines whether and how the entityID value will be URL encoded prior to replacement.  Allowed values are:

  • "none" : No encoding is performed.
  • "form" : Encoded using URL form parameter encoding (for query parameters).
  • "path" : Encoded using URL path encoding.
  • "fragment" : Encoded using URL fragment encoding.

The precise definition of these terms is defined in the documentation for the methods of the Guava library's UrlEscapers class.

transformRefBean ID
A reference to a transform function for the entityID. If used, the child element must be empty.
velocityEngineBean IDshibboleth.VelocityEngineBean ID of a custom VelocityEngine to use, generally unneeded

<Regex> child element

If the <Regex> child element is used, the metadata request URL is constructed by means of a regular expression transform. The entityID value is first matched against the regular expression contained in the <Regex> element's match attribute. Then, the <Regex> element's content is treated as a replacement expression to run based on the results of the match.

The <Regex> child element has the following required attribute:

NameTypeReq?Description
matchStringYA regular expression against which the entityID is evaluated

Note that only numeric/positional group references (e.g., $1) are supported.

NameTypeDefaultDescription
The following attributes are required on all metadata provider types:
idString
Identifier for logging, identification for command line reload, etc.
xsi:typeString
Specifies the exact type of provider to use (from those listed above, or a custom extension type)
The following attributes are common to all metadata provider types except the ChainingMetadataProvider type:

requireValidMetadata

Booleantrue

Whether candidate metadata found by the resolver must be valid in order to be returned (where validity is implementation specific, but in SAML cases generally depends on a validUntil attribute.) If this flag is true, then invalid candidate metadata will not be returned.

failFastInitialization

BooleantrueWhether to fail initialization of the underlying MetadataResolverService (and possibly the IdP as a whole) if the initialization of a metadata provider fails. When false, the IdP may start, and will continue to attempt to reload valid metadata if configured to do so, but operations that require valid metadata will fail until it does.
sortKeyInteger
Defines the order in which metadata providers are searched (see below), can only be specified on top level <MetadataProvider> elements.
The following are advanced settings supporting a new low-level feature allowing metadata lookup by keys other than the unique entityID and are rarely of use to a deployer.
criterionPredicateRegistryRefBean ID
Identifies the a custom CriterionPredicateRegistry bean used in resolving predicates from non-predicate input criteria
useDefaultPredicateRegistryBooleantrueFlag which determines whether the default CriterionPredicateRegistry will be used if a custom one is not supplied explicitly
satisfyAnyPredicatesBooleanfalseFlag which determines whether predicates used in filtering are connected by a logical 'OR' (true) or by logical 'AND' (false)

The following attributes are common to all dynamic metadata providers (i.e., DynamicHTTPMetadataProvider and LocalDynamicMetadataProvider):
NameTypeDefaultDescription
parserPoolRefBean IDshibboleth.ParserPoolIdentifies a Spring bean for the XML parser used to parse metadata. Generally should not be changed.
taskTimerRefBean ID
Identifies a Spring bean containing a Java TaskTimer used to schedule reloads. When not set, an internal timer is created. Generally should not be changed.
refreshDelayFactorReal Number (strictly between 0.0 and 1.0)0.75A factor applied to the initially determined refresh time in order to determine the next refresh time (typically to ensure refresh takes place prior to the metadata's expiration). Attempts to refresh metadata will generally begin around the product of this number and the maximum refresh delay.
minCacheDurationDurationPT10MThe minimum duration for which metadata will be cached before it is refreshed
maxCacheDurationDurationPT8HThe maximum duration for which metadata will be cached before it is refreshed
maxIdleEntityDataDurationPT8H

The maximum duration for which metadata will be allowed to be idle (no requests for it) before it is removed from the cache

removeIdleEntityDataBooleantrueFlag indicating whether idle metadata should be removed
cleanupTaskIntervalDurationPT30M

The interval at which the internal cleanup task should run. This task performs background maintenance tasks, such as the removal of expired and idle metadata.

persistentCacheManagerRefBean ID

The optional manager for the persistent cache store for resolved metadata. On metadata provider initialization, data present in the persistent cache will be loaded to memory, effectively restoring the state of the provider as closely as possible to that which existed before the previous shutdown. Each individual cache entry will only be loaded if 1) the entry is still valid as determined by the internal provider logic, and 2) the entry passes the (optional) predicate supplied via initializationFromCachePredicateRef.

persistentCacheManagerDirectoryDirectory path

The directory used for an internally-constructed filesystem-based persistent cache. This is a convenience parameter to avoid specifying a full bean via persistentCacheManagerRef. This option will be ignored if persistentCacheManagerRef is specified.

NOTE: Do NOT specify a directory containing other metadata or any other content under local control, as the cache manager may alter those files, or fail to process them correctly, or both. Use a dedicated directory for each provider.

persistentCacheKeyGeneratorRefBean IDinternal default instanceIdentifies a Spring bean for a Function which generates the string key used with the cache manager. The default implementation produces the lower-case hex-encoded SHA-1 digest of the entityID of the EntityDescriptor.
initializeFromPersistentCacheInBackgroundBooleantrueFlag indicating whether should initialize from the persistent cache in the background. Initializing from the cache in the background will improve IdP startup times.
backgroundInitializationFromCacheDelayDurationPT2S

The delay after which to schedule the background initialization from the persistent cache when initializeFromPersistentCacheInBackground=true.

initializationFromCachePredicateRefBean ID"always true" predicate

Identifies a Spring bean for an optional Predicate which determines whether a given entity should be loaded from the persistent cache at resolver initialization time.

Configure the Dynamic Attributes for the desired cache behavior. In particular, the minCacheDuration attribute and/or the maxCacheDuration attribute should be adjusted based on the life cycle of the metadata. Note that the cacheDuration attribute in metadata (if any) contributes to the overall cache behavior.

The following attributes are common to all HTTP metadata providers (i.e., DynamicHTTPMetadataProvider and FileBackedHTTPMetadataProvider).

An HTTP metadata provider includes a default implementation of the org.apache.http.client.HttpClient interface. The attributes in the following subsections control the behavior of the default HTTP client. In most cases, the default behavior is sufficient.

To override the default client implementation, configure the following attribute:

NameTypeDescription
httpClientRefBean ID

A reference to an externally defined Spring bean that specifies an org.apache.http.client.HttpClient object. This attribute conflicts with and overrides all of the other HTTP attributes. See the HttpClientConfiguration topic for more information.

Use of the httpClientRef attribute precludes the use of any and all of the HTTP attributes in the following subsections.

The following attributes apply to the HTTP connections obtained and managed by an HTTP metadata provider:

NameTypeDefaultDescription
connectionRequestTimeoutDurationDepends on provider typeThe maximum amount of time to wait for a connection to be returned from the HTTP client's connection pool manager. Set to PT0S to disable.
connectionTimeoutDurationDepends on provider typeThe maximum amount of time to wait to establish a connection with the remote server. Set to PT0S to disable.
socketTimeoutDurationDepends on provider typeThe maximum amount of time to wait between two consecutive packets while reading from the socket connected to the remote server. Set to PT0S to disable.

The following security-related attributes apply to any HTTP metadata provider:

NameTypeDefaultDescription
disregardTLSCertificateBooleanfalse

If true, no TLS certificate checking will take place over an HTTPS connection.

Be careful with this setting, it is typically only used during testing. See the HttpClientConfiguration topic for more information.

httpClientSecurityParametersRefBean ID

A reference to an externally defined Spring bean that specifies an HttpClientSecurityParameters instance, which consolidates all HTTP security parameters including advanced TLS usage.

This attribute conflicts with and overrides any explicit TrustEngine implementation configured as an inline <TLSTrustEngine> element. See the HttpClientConfiguration topic for more information.

The following attributes configure an HTTP proxy for use with an HTTP metadata provider:

NameTypeDescription
proxyHostString

The hostname of the HTTP proxy through which connections will be made

proxyPortString

The port of the HTTP proxy through which connections will be made

proxyUserString

The username used with the HTTP proxy through which connections will be made

proxyPasswordString

The password used with the HTTP proxy through which connections will be made

For a DynamicHTTPMetadataProvider, the HTTP Connection Attributes each have an aggressive 5 second timeout default. These may be tightened further if desired.

Examples

A typical use case is to load entity metadata dynamically from a metadata query server (i.e., a server that supports the Metadata Query Protocol). Here is a complete example:

Load entity metadata from a remote MDQ server
<!--
    Load entity metadata from a remote HTTP server via the Metadata
    Query Protocol: https://github.com/iay/md-query

    The sample configuration below implicitly formulates a Metadata Query
    Protocol URL from the given base URL. For example, if the entityID is 
    https://sso.example.org/sp, the provider will request the following 
    resource: 

      https://mdq.example.org/global/entities/https%3A%2F%2Fsso.example.org%2Fsp

    The sample configuration below assumes: (1) the top-level element of 
    the XML document is signed; (2) the top-level element of the XML 
    document is decorated with a validUntil attribute; (3) the validity 
    interval is two weeks (P14D) in duration; and (4) the server conforms 
    to the Metadata Query Protocol specification.

    The metadata is cached for efficiency. The minCacheDuration attribute 
    (default: PT10M) and the maxCacheDuration attribute (default: PT8H) 
    strongly influence the life cycle of metadata in the local cache. (Any 
    cacheDuration and validUntil attributes in the metadata itself also 
    influence the behavior of the local cache.) The goal is to avoid needless 
    interaction with the HTTP server. To achieve this goal, you need to 
    understand the life cycle of the metadata on the server. For this reason, 
    it is best to ask your federation operator for specific recommendations.

    The HTTP Connection Attributes include the connectionRequestTimeout 
    attribute (default: PT5S), the connectionTimeout attribute (default: PT5S), 
    and the socketTimeout attribute (default: PT5S). The default values of these 
    attributes are overridden in the example below.
-->
<MetadataProvider id="DynamicEntityMetadata" xsi:type="DynamicHTTPMetadataProvider"
                  connectionRequestTimeout="PT2S"
                  connectionTimeout="PT2S"
                  socketTimeout="PT4S">
     <!--
        Verify the signature on the root element of the metadata 
        using a trusted metadata signing certificate.
    -->
    <MetadataFilter xsi:type="SignatureValidation" requireSignedRoot="true"
        certificateFile="%{idp.home}/credentials/mdq-cert.pem"/>
 
    <!--
        Require a validUntil XML attribute on the root element and
        make sure its value is no more than 14 days into the future.
    -->
    <MetadataFilter xsi:type="RequiredValidUntil" maxValidityInterval="P14D"/>
 
    <!-- Specify the base URL for the Metadata Query Protocol -->
    <MetadataQueryProtocol>https://mdq.example.org/global/</MetadataQueryProtocol>
 
</MetadataProvider>

Note that the <MetadataQueryProtocol> child element encodes the base URL of the Metadata Query Protocol. For example, consider the following child element:

Using a <MetadataQueryProtocol> child element
<!-- Specify the base URL for the Metadata Query Protocol -->
<MetadataQueryProtocol>https://mdq.example.org/global/</MetadataQueryProtocol>

The previous <MetadataQueryProtocol> child element is equivalent to the following <Template> child element:

Using a <Template> child element
<!-- 
    The Template element specifies a simple template with a single parameter. 
    By default, the entityID is percent-encoded before substitution. 
--> 
<Template>https://mdq.example.org/global/entities/${entityID}</Template>

The above configuration explicitly formulates an MDQ protocol URL. This example is for illustration purposes only. If the server supports the Metadata Query Protocol, a <MetadataQueryProtocol> child element should be used instead. This intentionally hides the details of the Metadata Query Protocol.

Finally, here is an example of the well-known location strategy:

Example of Well-Known Location
<MetadataProvider id="WellKnownEntityMetadata" xsi:type="DynamicHTTPMetadataProvider">
 
    <!--
      Use the well-known location strategy to get SP metadata. The
      entityID is not configured here; it is determined from the
      AuthnRequest's Issuer element, as sent by the requester.
      
      In this case, the entityID MUST be in the form of a URL (rather 
      than a URN). It is STRONGLY RECOMMENDED that https URLs be used 
      to protect against man-in-the-middle attacks.
    -->
</MetadataProvider>

Frequently Asked Questions

What does “dynamic” mean?

A DynamicHTTPMetadataProvider fetches entity metadata as needed. We say that the IdP queries for SP metadata just-in-time.

Compare this to a FileBackedHTTPMetadataProvider that batch loads all of the entity descriptors in a metadata file whether or not the individual entity descriptors are actually needed. In contrast, a DynamicHTTPMetadataProvider loads exactly those entities that are needed—no more, no less. In this sense, a DynamicHTTPMetadataProvider is much more efficient.

OTOH, all metadata query protocols are synchronous protocols by definition. Basically the IdP is blocked until it obtains the metadata it needs.

How does metadata query work?

When an IdP receives a SAML protocol request from a particular SP, the IdP must first obtain entity metadata for that SP. If the IdP has no such metadata in its possession, metadata resolution proceeds sequentially according to a configured chain of metadata providers. Upon encountering a DynamicHTTPMetadataProvider in the chain, the IdP consults an HTTP client that acts as an intermediary between the IdP and the query server.

The HTTP client implements a shared HTTP cache. (RFC 7234) If the desired metadata is already cached, and the stored response is fresh, the client immediately returns the cached metadata to the IdP. Otherwise the client issues an HTTP request to the query server. Upon receiving a response from the server, the client caches the response and finally returns the metadata to the IdP.

In either case, the IdP parses the metadata and applies any metadata filters configured on the DynamicHTTPMetadataProvider. The metadata that ultimately emerges from the configured metadata pipeline is cached locally (in memory) for future use.

The next time the IdP receives a SAML protocol request from this SP, it again traverses the chain of providers until it encounters the DynamicHTTPMetadataProvider. This time, however, the IdP does not bother to consult the HTTP client since the needed metadata is in the IdP’s local cache.

How long does the metadata remain in the IdP’s local cache?

The IdP’s local cache is governed by the Dynamic Attributes. In particular, the minCacheDuration and maxCacheDuration attributes strongly influence the life cycle of metadata in the local cache. Any cacheDuration and validUntil attributes in the metadata itself also influence the behavior of the local cache.

Does the HTTP client cache the response in memory?

Yes, by default the HTTP client caches responses in memory. Consequently, two copies of each entity descriptor reside in memory, one managed by the HTTP client as an HTTP response, and another "first-order" metadata object managed directly by the IdP.

The HTTP client may be overridden to perform file-based caching but that cache will not survive a restart so the overall benefit of file caching is low. In most cases, a memory cache is preferred, and the metadata plugin can perform its own persistent caching to disk, which does survive a restart.

Does the HTTP client support HTTP conditional GET?

Yes, the HTTP client supports HTTP conditional GET (RFC 7232) for optimal performance but the inner workings of the HTTP client are opaque to the IdP. If the IdP does in fact consult the HTTP client, and the client returns metadata to the IdP, the IdP blindly parses the metadata and applies the metadata filters. There are no optimizations implemented on the IdP side to prevent re-parsing the XML because the fragments are small enough to limit the benefit.

What if the metadata query server is down or unavailable?

When the HTTP client sends an HTTP request to a metadata query server, the SAML protocol exchange is blocked until a response is received from the server and returned to the IdP. If the client reports a failed request, the IdP continues with the next provider in the configured chain of providers. If the offending DynamicHTTPMetadataProvider is the last provider in the chain, metadata resolution fails.

What can I do to minimize the impact of metadata query?

There are at least three things you can do to help minimize the impact of metadata query:

  1. Configure minCacheDuration and/or maxCacheDuration

  2. Configure the HTTP Connection Attributes

  3. Configure a robust chain of metadata providers

As noted above, the minCacheDuration and maxCacheDuration attributes strongly influence the life cycle of metadata in the local cache. The goal is to avoid needless interaction with the HTTP server. To achieve this goal, you need to understand the life cycle of the metadata on the server. For this reason, it is best to ask your federation operator for specific recommendations.

OTOH, the federation operator may influence the life cycle of metadata in the IdP’s local cache by including a cacheDuration attribute in the metadata itself. In that case, the deployer has fewer configuration options to consider, by design.

The HTTP Connection Attributes include the following attributes:

  1. connectionRequestTimeout (default: PT5S): The maximum amount of time to wait for a connection to be returned from the HTTP client's connection pool manager.

  2. connectionTimeout (default: PT5S): The maximum amount of time to wait to establish a connection with the remote server.

  3. socketTimeout (default: PT5S): The maximum amount of time to wait between two consecutive packets while reading from the socket connected to the remote server.

As noted above, each of these attributes defaults to 5 seconds. You may want to tighten these timeout values further, depending on what you know about the route to the server or the server itself.

Regardless of the IdP configuration or the service-level agreement you have with the server operator, things will go wrong. One thing you can do to hedge your bets is to deploy a local query server as backup. Alternatively, one or more high-value SPs can be pre-loaded into memory.




  • No labels