| HTTP Working Group | M. Nottingham |
| Internet-Draft | August 3, 2021 |
| Obsoletes: 3205 (if approved) | |
| Intended status: Best Current Practice | |
| Expires: February 4, 2022 |
Applications often use HTTP as a substrate to create HTTP-based APIs. This document specifies best practices for writing specifications that use HTTP to define new application protocols. It is written primarily to guide IETF efforts to define application protocols using HTTP for deployment on the Internet, but might be applicable in other situations.¶
RFC EDITOR: please remove this section before publication ¶
Discussion of this draft takes place on the HTTP working group mailing list (ietf-http-wg@w3.org), which is archived at https://lists.w3.org/Archives/Public/ietf-http-wg/.¶
Working Group information can be found at http://httpwg.github.io/; source code and issues list for this draft can be found at https://github.com/httpwg/http-extensions/labels/bcp56bis.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as “work in progress”.¶
This Internet-Draft will expire on February 4, 2022.¶
Copyright (c) 2021 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.¶
Applications other than Web browsing often use HTTP [I-D.ietf-httpbis-semantics] as a substrate, a practice sometimes referred to as creating "HTTP-based APIs", "REST APIs" or just "HTTP APIs". This is done for a variety of reasons, including:¶
These protocols are often ad hoc; they are intended for only deployment by one or a few servers, and consumption by a limited set of clients. As a result, a body of practices and tools has arisen around defining HTTP-based APIs that favours these conditions.¶
However, when such an application has multiple, separate implementations, is deployed on multiple uncoordinated servers, and is consumed by diverse clients -- as is often the case for HTTP APIs defined by standards efforts -- tools and practices intended for limited deployment can become unsuitable.¶
This mismatch is largely because the API's clients and servers will implement and evolve at different paces, leading to a need for deployments with different features and versions to co-exist. As a result, the designers of HTTP-based APIs intended for such deployments need to more carefully consider how extensibility of the service will be handled and how different deployment requirements will be accommodated.¶
More generally, an application protocol using HTTP faces a number of design decisions, including:¶
This document contains best current practices for the specification of such applications. Section 2 defines when it applies; Section 3 surveys the properties of HTTP that are important to preserve, and Section 4 conveys best practices for specifying them.¶
It is written primarily to guide IETF efforts to define application protocols using HTTP for deployment on the Internet, but might be applicable in other situations. Note that the requirements herein do not necessarily apply to the development of generic HTTP extensions.¶
This document obsoletes [RFC3205], to reflect experience and developments regarding HTTP in the intervening time.¶
Different applications have different goals when using HTTP. The requirements in this document apply when a specification defines an application that:¶
Additionally, when a specification is using HTTP, all of the requirements of the HTTP protocol suite are in force (in particular, [I-D.ietf-httpbis-semantics], but also other specifications such as the specific version of HTTP in use, and any extensions in use).¶
Note that this document is intended to apply to applications, not generic extensions to HTTP. Furthermore, while it is intended for IETF-specified applications, other standards organisations are encouraged to adhere to its requirements.¶
An application can rely upon HTTP without meeting the criteria for using it defined above. For example, an application might wish to avoid re-specifying parts of the message format, but change other aspects of the protocol's operation; or, it might want to use application-specific methods.¶
Doing so brings more freedom to modify protocol operations, but loses at least a portion of the benefits outlined in Section 3, as most HTTP implementations won't be easily adaptable to these changes, and the benefit of mindshare will be lost.¶
Such specifications MUST NOT use HTTP's URI schemes, transport ports, ALPN protocol IDs or IANA registries; rather, they are encouraged to establish their own.¶
This section examines the characteristics of HTTP that are important to consider when using HTTP to define an application protocol.¶
Much of the value of HTTP is in its generic semantics -- that is, the protocol elements defined by HTTP are potentially applicable to every resource, not specific to a particular context. Application-specific semantics are best expressed in message content and in header fields, not status codes or methods (although the latter do have generic semantics that relate to application state).¶
This generic/application-specific split allows a HTTP message to be handled by common software (e.g., HTTP servers, intermediaries, client implementations, and caches) without understanding the specific application. It also allows people to leverage their knowledge of HTTP semantics without special-casing them for a particular application.¶
Therefore, applications that use HTTP MUST NOT re-define, refine or overlay the semantics of generic protocol elements such as methods, status codes or existing header fields. Instead, they should focus their specifications on protocol elements that are specific to that application; namely their HTTP resources.¶
When writing a specification, it's often tempting to specify exactly how HTTP is to be implemented, supported and used. However, this can easily lead to an unintended profile of HTTP's behaviour. For example, it's common to see specifications with language like this:¶
A `POST` request MUST result in a `201 Created` response.
This forms an expectation in the client that the response will always be 201 Created, when in fact there are a number of reasons why the status code might differ in a real deployment; for example, there might be a proxy that requires authentication, or a server-side error, or a redirection. If the client does not anticipate this, the application's deployment is brittle.¶
See Section 4.2 for more details.¶
Another common practice is assuming that the HTTP server's name space (or a portion thereof) is exclusively for the use of a single application. This effectively overlays special, application-specific semantics onto that space, precludes other applications from using it.¶
As explained in [RFC8820], such "squatting" on a part of the URL space by a standard usurps the server's authority over its own resources, can cause deployment issues, and is therefore bad practice in standards.¶
Instead of statically defining URI components like paths, it is RECOMMENDED that applications using HTTP define and use links, to allow flexibility in deployment.¶
Using runtime links in this fashion has a number of other benefits -- especially when an application is to have multiple implementations and/or deployments (as is often the case for those that are standardised).¶
For example, navigating with a link allows a request to be routed to a different server without the overhead of a redirection, thereby supporting deployment across machines well.¶
It also becomes possible to "mix and match" different applications on the same server, and offers a natural mechanism for extensibility, versioning and capability management, since the document containing the links can also contain information about their targets.¶
Using links also offers a form of cache invalidation that's seen on the Web; when a resource's state changes, the application can change its link to it so that a fresh copy is always fetched.¶
HTTP offers a number of features to applications, such as:¶
Applications that use HTTP are encouraged to utilise the various features that the protocol offers, so that their users receive the maximum benefit from it, and to allow it to be deployed in a variety of situations. This document does not require specific features to be used, since the appropriate design tradeoffs are highly specific to a given situation. However, following the practices in Section 4 is a good starting point.¶
This section contains best practices for specifying the use of HTTP by applications, including practices for specific HTTP protocol elements.¶
Specifications should use [I-D.ietf-httpbis-semantics] as the primary reference for HTTP; it is not necessary to reference all of the specifications in the HTTP suite unless there are specific reasons to do so (e.g., a particular feature is called out).¶
Because HTTP is a hop-by-hop protocol, a connection can be handled by implementations that are not controlled by the application; for example, proxies, CDNs, firewalls and so on. Requiring a particular version of HTTP makes it difficult to use in these situations, and harms interoperability. Therefore, it is NOT RECOMMENDED that applications using HTTP specify a minimum version of HTTP to be used.¶
However, if an application's deployment would benefit from the use of a particular version of HTTP (for example, HTTP/2's multiplexing), this ought be noted.¶
Applications using HTTP MUST NOT specify a maximum version, to preserve the protocol's ability to evolve.¶
When specifying examples of protocol interactions, applications should document both the request and response messages, with complete header sections, preferably in HTTP/1.1 format. For example:¶
GET /thing HTTP/1.1 Host: example.com Accept: application/things+json User-Agent: Foo/1.0
HTTP/1.1 200 OK Content-Type: application/things+json Content-Length: 500 Server: Bar/2.2 [content here]
The server-side behaviours of an application are most effectively specified by defining the following protocol elements:¶
An application can define its operation by composing these protocol elements to define a set of resources that are identified by link relations and that implement specified behaviours, including:¶
For example, an application might specify:¶
Resources linked to with the "example-widget" link relation type are Widgets. The state of a Widget can be fetched in the "application/example-widget+json" format, and can be updated by PUT to the same link. Widget resources can be deleted. The "Example-Count" response header field on Widget representations indicates how many Widgets are held by the sender. The "application/example-widget+json" format is a JSON [RFC8259] format representing the state of a Widget. It contains links to related information in the link indicated by the Link header field value with the "example-other-info" link relation type.
An application's expectations for client behaviour ought to be closely aligned with those of Web browsers, to avoid interoperability issues when they are used.¶
One way to do this is to define it in terms of [FETCH], since that is the abstraction that browsers use for HTTP.¶
Some client behaviours (e.g., automatic redirect handling) and extensions (e.g., Cookies) are not required by HTTP, but nevertheless have become very common. If their use is not explicitly specified by applications using HTTP, there may be confusion and interoperability problems. In particular:¶
Applications using HTTP should not statically require HTTP features that are usually negotiated to be supported by clients. For example, requiring that clients support responses with a certain content-coding ([I-D.ietf-httpbis-semantics], Section 8.4.1) instead of negotiating for it ([I-D.ietf-httpbis-semantics], Section 12.5.3) means that otherwise conformant clients cannot interoperate with the application. Applications can encourage the implementation of such features, though.¶
In HTTP, the resources that clients interact with are identified with URLs [RFC3986]. As [RFC8820] explains, parts of the URL are designed to be under the control of the owner (also known as the "authority") of that server, to give them the flexibility in deployment.¶
This means that in most cases, specifications for applications that use HTTP won't contain fixed application URLs or paths; while it is common practice for a specification of a single-deployment API to specify the path prefix "/app/v1" (for example), doing so in an IETF specification is inappropriate.¶
Therefore, the specification writer needs some mechanism to allow clients to discovery an application's URLs. Additionally, they need to specify what URL scheme(s) the application should be used with, and whether to use a dedicated port, or reuse HTTP's port(s).¶
Generally, a client will begin interacting with a given application server by requesting an initial document that contains information about that particular deployment, potentially including links to other relevant resources. Doing so assures that the deployment is as flexible as possible (potentially spanning multiple servers), allows evolution, and also gives the application the opportunity to tailor the 'discovery document' to the client.¶
There are a few common patterns for discovering that initial URL.¶
The most straightforward mechanism for URL discovery is to configure the client with (or otherwise convey to it) a full URL. This might be done in a configuration document, or through another discovery mechanism.¶
However, if the client only knows the server's hostname and the identity of the application, there needs to be some way to derive the initial URL from that information.¶
An application cannot define a fixed prefix for its URL paths; see [RFC8820]. Instead, a specification for such an application can use one of the following strategies:¶
Once the discovery document is located, it can be fetched, cached for later reuse (if allowed by its metadata), and used to locate other resources that are relevant to the application, using full URIs or URL Templates.¶
In some cases, an application may not wish to use such a discovery document; for example, when communication is very brief, or when the latency concerns of doing so precludes the use of a discovery document. These situations can be addressed by placing all of the application's resources under a well-known location.¶
Applications that use HTTP will typically employ the "http" and/or "https" URI schemes. "https" is RECOMMENDED to provide authentication, integrity and confidentiality, as well as mitigate pervasive monitoring attacks [RFC7258].¶
However, application-specific schemes can also be defined. When defining an URI scheme for an application using HTTP, there are a number of tradeoffs and caveats to keep in mind:¶
Applications can use the applicable default port (80 for HTTP, 443 for HTTPS), or they can be deployed upon other ports. This decision can be made at deployment time, or might be encouraged by the application's specification (e.g., by registering a port for that application).¶
If a non-default port is used, it needs to be reflected in the authority of all URLs for that resource; the only mechanism for changing a default port is changing the URI scheme (see Section 4.4.2).¶
Using a port other than the default has privacy implications (i.e., the protocol can now be distinguished from other traffic), as well as operability concerns (as some networks might block or otherwise interfere with it). Privacy implications should be documented in Security Considerations.¶
Applications that use HTTP MUST confine themselves to using registered HTTP methods such as GET, POST, PUT, DELETE, and PATCH.¶
New HTTP methods are rare; they are required to be registered in the HTTP Method Registry with IETF Review (see [I-D.ietf-httpbis-semantics]), and are also required to be generic. That means that they need to be potentially applicable to all resources, not just those of one application.¶
While historically some applications (e.g., [RFC4791]) have defined non-generic methods, [I-D.ietf-httpbis-semantics] now forbids this.¶
When authors believe that a new method is required, they are encouraged to engage with the HTTP community early, and document their proposal as a separate HTTP extension, rather than as part of an application's specification.¶
GET is the most common and useful HTTP method; its retrieval semantics allow caching, side-effect free linking and underlies many of the benefits of using HTTP.¶
Queries can be performed with GET, often using the query component of the URL; this is a familiar pattern from Web browsing, and the results can be cached, improving efficiency of an often expensive process. In some cases, however, GET might be unwieldy for expressing queries, because of the limited syntax of the URI; in particular, if binary data forms part of the query terms, it needs to be encoded to conform to URI syntax.¶
While this is not an issue for short queries, it can become one for larger query terms, or ones which need to sustain a high rate of requests. Additionally, some HTTP implementations limit the size of URLs they support -- although modern HTTP software has much more generous limits than previously (typically, considerably more than 8000 octets, as required by [I-D.ietf-httpbis-semantics]).¶
In these cases, an application using HTTP might consider using POST to express queries in the request's content; doing so avoids encoding overhead and URL length limits in implementations. However, in doing so it should be noted that the benefits of GET such as caching and linking to query results are lost. Therefore, applications using HTTP that feel a need to allow POST queries ought consider allowing both methods.¶
Applications should not change their state or have other side effects that might be significant to the client, since implementations can and do retry HTTP GET requests that fail, and some GET requests protected by TLS Early Data might be vulnerable to replay attacks (see [RFC8470]). Note that this does not include logging and similar functions; see [I-D.ietf-httpbis-semantics], Section 9.2.1.¶
Finally, note that while HTTP allows GET requests to have content syntactically, this is done only to allow parsers to be generic; as per [I-D.ietf-httpbis-semantics], Section 9.3.1, content on a GET has no meaning, and will be either ignored or rejected by generic HTTP software.¶
The OPTIONS method was defined for metadata retrieval, and is used both by WebDAV [RFC4918] and CORS [FETCH]. Because HTTP-based APIs often need to retrieve metadata about resources, it is often considered for their use.¶
However, OPTIONS does have significant limitations:¶
Instead of OPTIONS, one of these alternative approaches might be more appropriate:¶
HTTP status codes convey semantics both for the benefit of generic HTTP components -- such as caches, intermediaries, and clients -- and applications themselves. However, applications can encounter a number of pitfalls in their use.¶
First, status codes are often generated by components other than the application itself. This can happen, for example, when network errors are encountered, a captive portal, proxy or Content Delivery Network is present, when a server is overloaded, or it thinks it is under attack. They can even be generated by generic client software when certain error conditions are encountered. As a result, if an application assigns specific semantics to one of these status codes, a client can be misled about its state, because the status code was generated by a generic component, not the application itself.¶
Furthermore, mapping application errors to individual HTTP status codes one-to-one often leads to a situation where the finite space of applicable HTTP status codes is exhausted. This, in turn, leads to a number of bad practices -- including minting new, application-specific status codes, or using existing status codes even though the link between their semantics and the application's is tenuous at best.¶
Instead, applications using HTTP should define their errors to use the most applicable status code, making generous use of the general status codes (200, 400 and 500) when in doubt. Importantly, they should not specify a one-to-one relationship between status codes and application errors, thereby avoiding the exhaustion issue outlined above.¶
To distinguish between multiple error conditions that are mapped to the same status code, and to avoid the misattribution issue outlined above, applications using HTTP should convey finer-grained error information in the response's message content and/or header fields. [RFC7807] provides one way to do so.¶
Because the set of registered HTTP status codes can expand, applications using HTTP should explicitly point out that clients ought to be able to handle all applicable status codes gracefully (i.e., falling back to the generic n00 semantics of a given status code; e.g., 499 can be safely handled as 400 by clients that don't recognise it). This is preferable to creating a "laundry list" of potential status codes, since such a list won't be complete in the foreseeable future.¶
Applications using HTTP MUST NOT re-specify the semantics of HTTP status codes, even if it is only by copying their definition. It is NOT RECOMMENDED they require specific reason phrases to be used; the reason phrase has no function in HTTP, is not guaranteed to be preserved by implementations, and is not carried at all in the HTTP/2 [RFC7540] message format.¶
Applications MUST only use registered HTTP status codes. As with methods, new HTTP status codes are rare, and required (by [I-D.ietf-httpbis-semantics]) to be registered with IETF Review. Similarly, HTTP status codes are generic; they are required (by [I-D.ietf-httpbis-semantics]) to be potentially applicable to all resources, not just to those of one application.¶
When authors believe that a new status code is required, they are encouraged to engage with the HTTP community early, and document their proposal as a separate HTTP extension, rather than as part of an application's specification.¶
The 3xx series of status codes specified in Section 15.4 of [I-D.ietf-httpbis-semantics] direct the user agent to another resource to satisfy the request. The most common of these are 301, 302, 307 and 308, all of which use the Location response header field to indicate where the client should resend the request.¶
There are two ways that the members of this group of status codes differ:¶
This table summarises their relationships:¶
| Permanent | Temporary | |
|---|---|---|
| Allows changing the request method from POST to GET | 301 | 302 |
| Does not allow changing the request method | 308 | 307 |
The 303 See Other status code can be used to inform the client that the result of an operation is available at a different location using GET.¶
As noted in [I-D.ietf-httpbis-semantics], a user agent is allowed to automatically follow a 3xx redirect that has a Location response header field, even if they don't understand the semantics of the specific status code. However, they aren't required to do so; therefore, if an application using HTTP desires redirects to be automatically followed, it needs to explicitly specify the circumstances when this is required.¶
Redirects can be cached (when appropriate cache directives are present), but beyond that they are not 'sticky' -- i.e., redirection of a URI will not result in the client assuming that similar URIs (e.g., with different query parameters) will also be redirected.¶
Applications using HTTP are encouraged to specify that 301 and 302 responses change the subsequent request method from POST (but no other method) to GET, to be compatible with browsers. Generally, when a redirected request is made, its header fields are copied from the original request's. However, they can be modified by various mechanisms; e.g., sent Authorization ([I-D.ietf-httpbis-semantics], Section 11) and Cookie ([RFC6265]) header fields will change if the origin (and sometimes path) of the request changes. An application using HTTP should specify if any request header fields that it defines need to be modified or removed upon a redirect; however, this behaviour cannot be relied upon, since a generic client (like a browser) will be unaware of such requirements.¶
Applications often define new HTTP header fields. Typically, using HTTP header fields is appropriate in a few different situations:¶
When the conditions above are not met, it is usually better to convey application-specific information in other places; e.g., the message content or the URL query string.¶
New header fields MUST be registered, as per Section 16.3 of [I-D.ietf-httpbis-semantics].¶
See Section 16.3.2 of [I-D.ietf-httpbis-semantics] for guidelines to consider when minting new header fields. [RFC8941] provides a common structure for new header fields, and avoids many issues in their parsing and handling; it is RECOMMENDED that new header fields use it.¶
It is RECOMMENDED that header field names be short (even when field compression is used, there is an overhead) but appropriately specific. In particular, if a header field is specific to an application, an identifier for that application can form a prefix to the header field name, separated by a "-".¶
For example, if the "example" application needs to create three header fields, they might be called "example-foo", "example-bar" and "example-baz". Note that the primary motivation here is to avoid consuming more generic field names, not to reserve a portion of the namespace for the application; see [RFC6648] for related considerations.¶
The semantics of existing HTTP header fields MUST NOT be re-defined without updating their registration or defining an extension to them (if allowed). For example, an application using HTTP cannot specify that the Location header field has a special meaning in a certain context.¶
See Section 4.9 for the interaction between header fields and HTTP caching; in particular, request header fields that are used to "select" a response have impact there, and need to be carefully considered.¶
See Section 4.10 for considerations regarding header fields that carry application state (e.g., Cookie).¶
HTTP caching [I-D.ietf-httpbis-cache] is one of the primary benefits of using HTTP for applications; it provides scalability, reduces latency and improves reliability. Furthermore, HTTP caches are readily available in browsers and other clients, networks as forward and reverse proxies, Content Delivery Networks and as part of server software.¶
Even when an application using HTTP isn't designed to take advantage of caching, it needs to consider how caches will handle its responses, to preserve correct behaviour when one is interposed (whether in the network, server, client, or intervening infrastructure).¶
Assigning even a short freshness lifetime ([I-D.ietf-httpbis-cache], Section 4.2) -- e.g., 5 seconds -- allows a response to be reused to satisfy multiple clients, and/or a single client making the same request repeatedly. In general, if it is safe to reuse something, consider assigning a freshness lifetime.¶
The most common method for specifying freshness is the max-age response directive ([I-D.ietf-httpbis-cache], Section 5.2.2.1). The Expires header field ([I-D.ietf-httpbis-cache], Section 5.3) can also be used, but it is not necessary; all modern cache implementations support Cache-Control, and specifying freshness as a delta is usually more convenient and less error-prone.¶
It is not necessary to add the public response directive ([I-D.ietf-httpbis-cache], Section 5.2.2.9) to cache most responses; it is only necessary when it's desirable to store an authenticated response.¶
In some situations, responses without explicit cache freshness directives will be stored and served using a heuristic freshness lifetime; see [I-D.ietf-httpbis-cache], Section 4.2.2. As the heuristic is not under control of the application, it is generally preferable to set an explicit freshness lifetime, or make the response explicitly uncacheable.¶
If caching of a response is not desired, the appropriate response directive is "Cache-Control: no-store". Other directives are not necessary, and no-store only need be sent in situations where the response might be cached; see [I-D.ietf-httpbis-cache], Section 3. Note that "Cache-Control: no-cache" allows a response to be stored, just not reused by a cache without validation; it does not prevent caching (despite its name).¶
For example, this response cannot be stored or reused by a cache:¶
HTTP/1.1 200 OK Content-Type: application/example+xml Cache-Control: no-store [content]
Authors should understand that stale responses (e.g., with "Cache-Control: max-age=0") can be reused by caches when disconnected from the origin server; this can be useful for handling network issues.¶
If doing so is not suitable for a given response, the origin should use "Cache-Control: must-revalidate". See Section 4.2.4 of [I-D.ietf-httpbis-cache], and also [RFC5861] for additional controls over stale content.¶
Stale responses can be refreshed by assigning a validator, saving both transfer bandwidth and latency for large responses; see Section 13 of [I-D.ietf-httpbis-semantics].¶
When an application has a need to express a lifetime that's separate from the freshness lifetime, this should be conveyed separately, either in the response's content or in a separate header field. When this happens, the relationship between HTTP caching and that lifetime needs to be carefully considered, since the response will be used as long as it is considered fresh.¶
In particular, application authors need to consider how responses that are not freshly obtained from the origin server should be handled; if they have a concept like a validity period, this will need to be calculated considering the age of the response (see [I-D.ietf-httpbis-cache], Section 4.2.3).¶
One way to address this is to explicitly specify that all responses be fresh upon use.¶
If an application uses a request header field to change the response's header fields or content, authors should point out that this has implications for caching; in general, such resources need to either make their responses uncacheable (e.g., with the "no-store" cache-control directive defined in [I-D.ietf-httpbis-cache], Section 5.2.2.5) or send the Vary response header field ([I-D.ietf-httpbis-semantics], Section 12.5.5) on all responses from that resource (including the "default" response).¶
For example, this response:¶
HTTP/1.1 200 OK Content-Type: application/example+xml Cache-Control: max-age=60 ETag: "sa0f8wf20fs0f" Vary: Accept-Encoding [content]
can be stored for 60 seconds by both private and shared caches, can be revalidated with If-None-Match, and varies on the Accept-Encoding request header field.¶
Applications can use stateful cookies [RFC6265] to identify a client and/or store client-specific data to contextualise requests.¶
When used, it is important to carefully specify the scoping and use of cookies; if the application exposes sensitive data or capabilities (e.g., by acting as an ambient authority), exploits are possible. Mitigations include using a request-specific token to assure the intent of the client.¶
Clients often need to send multiple requests to perform a task.¶
In HTTP/1, parallel requests are most often supported by opening multiple connections. Application performance can be impacted when too many simultaneous connections are used, because connections' congestion control will not be coordinated. Furthermore, it can be difficult for applications to decide when and to select which connection to use to issue a given request, further impacting performance.¶
HTTP/2 and HTTP/3 offer multiplexing to applications, removing the need to use multiple connections. However, application performance can still be significantly affected by how the server chooses to prioritize responses. Depending on the application, it might be best for the server to determine the priority of responses, or for the client to hint its priorities to the server (see, e.g., [I-D.ietf-httpbis-priority]).¶
In all versions of HTTP, requests are made independently -- you can't rely on the relative order of two requests to guarantee processing order. This is because they might be sent over a multiplexed protocol by an intermediary, sent to different origin servers, or the server might even perform processing in a different order. If two requests need strict ordering, the only reliable way to assure the outcome is to issue the second request when the final response to the first has begun.¶
Applications MUST NOT make assumptions about the relationship between separate requests on a single transport connection; doing so breaks many of the assumptions of HTTP as a stateless protocol, and will cause problems in interoperability, security, operability and evolution.¶
Applications can use HTTP authentication Section 11 of [I-D.ietf-httpbis-semantics] to identify clients. As per [RFC7617], the Basic authentication scheme is not suitable for protecting sensitive or valuable information unless the channel is secure (e.g., using the "HTTPS" URI scheme). Likewise, [RFC7617] requires the Digest authentication scheme to be used over a secure channel.¶
When used, it is important to carefully specify the scoping and use of authentication; if the application exposes sensitive data or capabilities (e.g., by acting as an ambient authority), exploits are possible. Mitigations include using a request-specific token to assure the intent of the client.¶
Even if there is not an intent for an application to be used with a Web browser, its resources will remain available to browsers and other HTTP clients. This means that all such applications that use HTTP need to consider how browsers will interact with them, particularly regarding security.¶
For example, if an application's state can be changed using a POST request, a Web browser can easily be coaxed into cross-site request forgery (CSRF) from arbitrary Web sites.¶
Or, if an attacker gains control of content returned from the application's resources (for example, part of the request is reflected in the response, or the response contains external information that the attacker can change), they can inject code into the browser and access data and capabilities as if they were the origin -- a technique known as a cross-site scripting (XSS) attack.¶
This is only a small sample of the kinds of issues that applications using HTTP must consider. Generally, the best approach is to consider the application actually as a Web application, and to follow best practices for their secure development.¶
A complete enumeration of such practices is out of scope for this document, but some considerations include:¶
Depending on how they are intended to be deployed, specifications for applications using HTTP might require the use of these mechanisms in specific ways, or might merely point them out in Security Considerations.¶
An example of a HTTP response from an application that does not intend for its content to be treated as active by browsers might look like this:¶
HTTP/1.1 200 OK Content-Type: application/example+json X-Content-Type-Options: nosniff Content-Security-Policy: default-src 'none' Cache-Control: max-age=3600 Referrer-Policy: no-referrer [content]
Because many HTTP capabilities are scoped to the origin [RFC6454], applications also need to consider how deployments might interact with other applications (including Web browsing) on the same origin.¶
For example, if Cookies [RFC6265] are used to carry application state, they will be sent with all requests to the origin by default (unless scoped by path), and the application might receive cookies from other applications on the origin. This can lead to security issues, as well as collision in cookie names.¶
One solution to these issues is to require a dedicated hostname for the application, so that it has a unique origin. However, it is often desirable to allow multiple applications to be deployed on a single hostname; doing so provides the most deployment flexibility and enables them to be "mixed" together (See [RFC8820] for details). Therefore, applications using HTTP should strive to allow multiple applications on an origin.¶
To enable this, when specifying the use of Cookies, HTTP authentication realms [I-D.ietf-httpbis-semantics], or other origin-wide HTTP mechanisms, applications using HTTP should not mandate the use of a particular name, but instead let deployments configure them. Consideration should be given to scoping them to part of the origin, using their specified mechanisms for doing so.¶
HTTP/2 adds the ability for servers to "push" request/response pairs to clients in [RFC7540], Section 8.2. While server push seems like a natural fit for many common application semantics (e.g., "fanout" and publish/subscribe), a few caveats should be noted:¶
Applications wishing to optimise cases where the client can perform work related to requests before the full response is available (e.g., fetching links for things likely to be contained within) might benefit from using the 103 (Early Hints) status code; see [RFC8297].¶
Applications using server push directly need to enforce the requirements regarding authority in [RFC7540], Section 8.2, to avoid cross-origin push attacks.¶
It's often necessary to introduce new features into application protocols, and change existing ones.¶
In HTTP, backwards-incompatible changes are possible using a number of mechanisms:¶
This document has no requirements for IANA.¶
Section 4.10 discusses the impact of using stateful mechanisms in the protocol as ambient authority, and suggests a mitigation.¶
Section 4.4.2 recommends support for 'https' URLs, and discourages the use of 'http' URLs, to provide authentication, integrity and confidentiality, as well as mitigate pervasive monitoring attacks. Many applications using HTTP perform authentication and authorization with bearer tokens (e.g., in session cookies). If the transport is unencrypted, an attacker that can eavesdrop on HTTP communications can often escalate their privilege to perform operations on resources.¶
Section 4.13 highlights the implications of Web browsers' capabilities on applications that use HTTP.¶
Section 4.14 discusses the issues that arise when applications are deployed on the same origin as Web sites (and other applications).¶
Section 4.15 highlights risks of using HTTP/2 server push in a manner other than specified.¶
Applications that use HTTP in a manner that involves modification of implementations -- for example, requiring support for a new URI scheme, or a non-standard method -- risk having those implementations "fork" from their parent HTTP implementations, with the possible result that they do not benefit from patches and other security improvements incorporated upstream.¶
HTTP clients can expose a variety of information to servers. Besides information that's explicitly sent as part of an application's operation (for example, names and other user-entered data), and "on the wire" (which is one of the reasons https is recommended in Section 4.4.2), other information can be gathered through less obvious means -- often by connecting activities of a user over time.¶
This includes session information, tracking the client through fingerprinting, and mobile code.¶
Session information includes things like the IP address of the client, TLS session tickets, Cookies, ETags stored in the client's cache, and other stateful mechanisms. Applications are advised to avoid using session mechanisms unless they are unavoidable or necessary for operation, in which case these risks needs to be documented. When they are used, implementations should be encouraged to allow clearing such state.¶
Fingerprinting uses unique aspects of a client's messages and behaviours to connect disparate requests and connections. For example, the User-Agent request header field conveys specific information about the implementation; the Accept-Language request header field conveys the users' preferred language. In combination, a number of these markers can be used to uniquely identify a client, impacting its control over its data. As a result, applications are advised to specify that clients should only emit the information they need to function in requests.¶
Finally, if an application exposes the ability to run mobile code, great care needs to be taken, since any ability to observe its environment can be used as an opportunity to both fingerprint the client and to obtain and manipulate private data (including session information). For example, access to high-resolution timers (even indirectly) can be used to profile the underlying hardware, creating a unique identifier for the system. Applications are advised to avoid allowing the use of mobile code where possible; when it cannot be avoided, the resulting system's security properties need be carefully scrutinised.¶