Status of This Memo

This is an Internet Standards Track document.¶

This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Further information on Internet Standards is available in Section 2 of RFC 7841.¶

Information about the current status of this document, any errata, and how to provide feedback on it may be obtained athttps://www.rfc-editor.org/info/rfc9289.¶

Copyright Notice

Copyright (c) 2022 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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶

Table of Contents

1.Introduction

In 2014 the IETF published a document entitled "Pervasive Monitoring Is anAttack"[RFC7258], whichrecognized that unauthorized observation of network traffic had becomewidespread and was a subversive threat to all who make use of the Internet atlarge. It strongly recommended that newly defined Internet protocols shouldmake a genuine effort to mitigate monitoring attacks. Typically, thismitigation includes encrypting data in transit.¶

The Remote Procedure Call version 2 protocol has been a Proposed Standard forthree decades (see[RFC5531]and its antecedents). Over twenty years ago, Eisler et al. first introducedRPCSEC_GSS as an in-transit encryption mechanism for RPC[RFC2203]. However, experiencehas shown that RPCSEC_GSS with in-transit encryption can be challenging to usein practice due to the following:¶

• Parts of each RPC header remain in cleartext,constituting a loss of metadata confidentiality.¶
• Offloading the Generic Security Service (GSS) privacy service is not practicalin large multi-user deploymentssince each message is encrypted using a key basedon the issuing RPC user.¶

However strong GSS-provided confidentiality is,it cannot provide any security if the challengesof using it result in choosing not to deploy it at all.¶

Moreover, the use of AUTH_SYSremains common despite the adverse effectsthat acceptance of User Identifiers (UIDs) and Group Identifiers (GIDs)from unauthenticated clients brings with it.Continued use is in part because:¶

• Per-client deployment and administrative costsfor the only well-defined alternative to AUTH_SYSare expensive at scale.For instance, administrators must provide keying materialfor each RPC client, including transient clients.¶
• GSS host identity management and user identity management typically must beenforced in the same security realm. However, cloud providers, for instance,might prefer to remain authoritative for host identity but allow tenants tomanage user identities within their private networks.¶

In view of the challenges with the currently available mechanisms forauthenticating and protecting the confidentiality of RPC transactions, thisdocument specifies a transport-layer security mechanism that complements theexisting ones. The TLS[RFC8446] and Datagram Transport Layer Security (DTLS)[RFC9147]protocols are well-established Internet building blocks that protect manystandard Internet protocols such as the Hypertext Transfer Protocol (HTTP)[RFC9110].¶

Encrypting at the RPC transport layer accords several significant benefits:¶

Encryption by Default:

Transport encryption can be enabledwithout additional administrative tasks such asidentifying client systems to a trust authorityandproviding each with keying material.¶

Encryption Offload:

Hardware support for the GSS privacy service has not appeared in the marketplace.However, the use of a well-established transport encryption mechanismthat is employed by other ubiquitous network protocolsmakes it more likely that encryption offload for RPCis practicable.¶

Securing AUTH_SYS:

Most critically, transport encryption cansignificantly reduce several security issuesinherent in the current widespread use of AUTH_SYS(i.e., acceptance of UIDs and GIDsgenerated by an unauthenticated client).¶

Decoupled User and Host Identities:

TLS can be used to authenticate peer hostswhile other security mechanisms can handle user authentication.¶

Compatibility:

The imposition of encryption at the transport layerprotects any upper-layer protocol that employs RPC,without alteration of the upper-layer protocol.¶

Further,Section 6of the current document defines policies in line with[RFC7435]that enable RPC-with-TLS to be deployed opportunistically in environments thatcontain RPC implementations that do not support TLS. However, specificationsfor RPC-based upper-layer protocols should choose to require even stricterpolicies that guarantee encryption and host authentication are used for all RPCtransactions to mitigate against pervasive monitoring attacks[RFC7258]. Enforcing the use ofRPC-with-TLS is of particular importance for existing upper-layer protocolswhose security infrastructure is weak.¶

The protocol specification in the current document assumes that support forONC RPC[RFC5531], TLS[RFC8446], PKIX[RFC5280], DNSSEC/DNS-BasedAuthentication of Named Entities (DANE)[RFC6698], and optionally RPCSEC_GSS[RFC2203]is available within the platformwhere RPC-with-TLS support is to be added.¶

2.Requirements Language

The key words "MUST", "MUST NOT","REQUIRED", "SHALL", "SHALL NOT","SHOULD", "SHOULD NOT","RECOMMENDED", "NOT RECOMMENDED","MAY", and "OPTIONAL" in this document are to beinterpreted as described in BCP 14[RFC2119][RFC8174]when, and only when, they appear in all capitals, as shown here.¶

3.Terminology

This document adopts the terminology introduced inSection 3 of [RFC6973]and assumes a working knowledge ofthe RPC version 2 protocol[RFC5531]andthe TLS version 1.3 protocol[RFC8446].¶

Note also that the NFS community long ago adopted theuse of the term "privacy" from documents such as[RFC2203].In the current document, the authors use the term"privacy" only when referring specificallyto the historic GSS privacy service defined in[RFC2203].Otherwise, the authors use the term "confidentiality",following the practices of contemporary security communities.¶

We adhere to the convention that a "client"is a network host that actively initiates an association,anda "server" is a network host that passively accepts an association request.¶

RPC documentation historically refers tothe authentication of a connecting host as "machine authentication"or "host authentication".TLS documentation refers to the same as "peer authentication".In the current document, there is little distinction between these terms.¶

The term "user authentication" in the current document refers specifically tothe RPC caller's credential, provided in the "cred" and "verf" fields in eachRPC Call.¶

4.RPC-with-TLS in Operation

5.TLS Requirements

When peers negotiate a TLS session that is to transport RPC,the following restrictions apply:¶

• ImplementationsMUST NOT negotiate TLS versions prior to 1.3(for TLS[RFC8446]or DTLS[RFC9147],respectively). Support for mandatory-to-implement cipher suites for thenegotiated TLS version isREQUIRED.¶
• ImplementationsMUST conform to the recommendations for TLSusage specified in BCP 195[RFC7525]. Although RFC 7525 permits the use of TLS 1.2, therequirement to use TLS 1.3 or later for RPC-with-TLS takes precedence.Further, because TLS 1.3 ciphers are qualitatively different than ciphersuites in previous versions of TLS, and RFC 7525 predates TLS 1.3, the ciphersuite recommendations in RFC 7525 do not apply to RPC-with-(D)TLS. A strictTLS mode for RPC-with-TLS that protects against STRIPTLS attacks is discussedin detail inSection 6.1.1.¶
• ImplementationsMUST support certificate-based mutualauthentication. Support for Pre-Shared Key (PSK) mutual authentication isOPTIONAL; seeSection 5.2.2for further details.¶
• Negotiation of a cipher suite providing confidentiality aswell as integrity protection isREQUIRED.¶

Client implementationsMUST include the"application_layer_protocol_negotiation(16)" extension[RFC7301]in their "ClientHello" messageandMUST include the protocol identifierdefined inSection 7.2in that message's ProtocolNameList value.¶

Similarly, in response to the "ClientHello" message,server implementationsMUST include the"application_layer_protocol_negotiation(16)" extension[RFC7301]in their "ServerHello" messageandMUST include only the protocol identifierdefined inSection 7.2in that message's ProtocolNameList value.¶

If the server responds incorrectly(for instance, if the "ServerHello" message does not conform to the above requirements),the clientMUST NOT establish a TLS session for use with RPCon this connection.See[RFC7301]for further details about how to form these messages properly.¶

6.Security Considerations

One purpose of the mechanism described in the current documentis to protect RPC-based applications against threatsto the confidentiality of RPC transactionsandRPC user identities.A taxonomy of these threats appears inSection 5 of [RFC6973].Also,Section 6 of [RFC7525]contains a detailed discussionof technologies used in conjunction with TLS.Section 8 of [RFC5280]covers important considerations about handling certificate material securely.Implementers should familiarize themselves with these materials.¶

Once a TLS session is established,the RPC payload carried on TLS version 1.3 is forward secure.However, implementers need to be aware that replay attackscan occur during session establishment.Remedies for such attacks are discussed in detail inSection 8 of [RFC8446].Further, the current document does notprovide a profile that defines the use of 0-RTT data(seeAppendix E.5 of [RFC8446]).Therefore, RPC-with-TLS implementationsMUST NOTuse 0-RTT data.¶

7.IANA Considerations

Appendix A.Known Weaknesses of the AUTH_SYS Authentication Flavor

The ONC RPC protocol, as specified in[RFC5531], provides several modes of security, commonlyreferred to as "authentication flavors". Some of these flavors provide muchmore than an authentication service. We refer to these as authenticationflavors, security flavors, or simply, flavors. One of the earliest and mostbasic flavors is AUTH_SYS, also known as AUTH_UNIX.Appendix A of [RFC5531]specifies AUTH_SYS.¶

AUTH_SYS assumes that the RPC client and serverboth use POSIX-style user and group identifiers(each user and group can be distinctly representedas a 32-bit unsigned integer).It also assumes that the client and serverboth use the same mapping of user and group to an integer.One user ID, one primary group ID, and up to 16 supplemental group IDsare associated with each RPC request.The combination of these identifies the entity on the clientthat is making the request.¶

A string identifies peers (hosts) in each RPC request.[RFC5531]does not specify any requirements for this stringother than that it is no longer than 255 octets.It does not have to be the same from request to request.Also, it does not have to match the DNS hostname of the sending host.For these reasons,even though most implementations fill in their hostname in this field,receivers typically ignore its content.¶

Appendix A of [RFC5531]contains a brief explanation of security considerations:¶

It should be noted that use of this flavor of authentication does notguarantee any security for the users or providers of a service, initself. The authentication provided by this scheme can be consideredlegitimate only when applications using this scheme and the networkcan be secured externally, and privileged transport addresses areused for the communicating end-points (an example of this is the useof privileged TCP/UDP ports in UNIX systems -- note that not allsystems enforce privileged transport address mechanisms).¶

It should be clear, therefore, that AUTH_SYS by itself(i.e., without strong client authentication)offers little to no communication security:¶

1. It does not protect the confidentiality or integrity ofRPC requests,users,orpayloads,relying instead on "external" security.¶
2. It does not provide authentication of RPC peer machines,other than inclusion of an unprotected domain name.¶
3. The use of 32-bit unsigned integers as user and group identifiersis problematic because these data types arenot cryptographically signed or otherwise verified by any authority.In addition, the mapping of these integers to users and groupshas to be consistent amongst a server and its cohort of clients.¶
4. Because the user and group ID fields are not integrity protected,AUTH_SYS does not provide non-repudiation.¶

Appendix B.ASN.1 Module

The following module adheres to ASN.1 specifications[X.680] and[X.690].¶

<CODE BEGINS>RPCwithTLS-2021  { iso(1) identified-organization(3) dod(6) internet(1)  security(5) mechanisms(5) pkix(7) id-mod(0)  id-mod-rpcWithTLS-2021(105) }DEFINITIONS IMPLICIT TAGS ::=BEGIN-- OID Arcid-kp OBJECT IDENTIFIER ::=  { iso(1) identified-organization(3) dod(6) internet(1)    security(5) mechanisms(5) pkix(7) kp(3) }-- Extended Key Usage Valuesid-kp-rpcTLSClient OBJECT IDENTIFIER ::= { id-kp 33 }id-kp-rpcTLSServer OBJECT IDENTIFIER ::= { id-kp 34 }END<CODE ENDS>

Acknowledgments

Special mention goes toCharles Fisher,author of"Encrypting NFSv4 with Stunnel TLS". His article inspired themechanism described in the current document.¶

Many thanks toBenjamin Coddington,Tigran Mkrtchyan,andRick Macklemfor their work on prototype implementations and feedback on the current document.Also, thanks toBenjamin Kadukfor his expert guidance on the use of PKIX and TLSand toRuss Housley for his ASN.1 expertise and for providing other proper finishing touches.In addition, the authors thank the other members of the IESG fortheir astute review comments. These contributors made this a significantly better document.¶

Thanks toDerrell Piperfor numerous suggestions that improved boththis simple mechanismandthe current document's security-related discussion.¶

Many thanks toTransport Area DirectorMagnus Westerlundfor his sharp questions and careful readingof the final revisions of the current document.The text ofSection 5.1.2is mostly his contribution.¶

The authors are additionally grateful toBill Baker,David Black,Alan DeKok,Lars Eggert,Olga Kornievskaia,Greg Marsden,Alex McDonald,Justin Mazzola Paluska,Tom Talpey,Martin Thomson,andNico Williamsfor their input and support of this work.¶

Finally, special thanks toNFSV4 Working Group Chair and document shepherdDavid Noveck,NFSV4 Working Group ChairsSpencer SheplerandBrian Pawlowski, and NFSV4 Working Group SecretaryThomas Haynesfor their guidance and oversight.¶

Authors' Addresses