Status of This Memo

This document is not an Internet Standards Track specification; it is published for the historical record.¶

This document defines a Historic Document for the Internet community. 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). Not all documents approved by the IESG are candidates for any level of Internet Standard; see 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 at https://www.rfc-editor.org/info/rfc9327.¶

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.¶

This document may contain material from IETF Documents or IETF Contributions published or made publicly available before November 10, 2008. The person(s) controlling the copyright in some of this material may not have granted the IETF Trust the right to allow modifications of such material outside the IETF Standards Process. Without obtaining an adequate license from the person(s) controlling the copyright in such materials, this document may not be modified outside the IETF Standards Process, and derivative works of it may not be created outside the IETF Standards Process, except to format it for publication as an RFC or to translate it into languages other than English.¶

Table of Contents

4. Commands

Commands consist of the header and optional data field shown in Figure 1. When present, the data field contains a list of identifiers or assignments in the form <<identifier>>[=<<value>>],<<identifier>>[=<<value>>],... where <<identifier>> is the ASCII name of a system or peer variable such as the ones specified in RFC 5905 and <<value>> is expressed as a decimal, hexadecimal, or string constant in the syntax of the C programming language. Where no ambiguity exists, the "sys." or "peer." prefixes can be suppressed. Space characters (ASCII nonprinting format effectors) can be added to improve readability for simple monitoring programs that do not reformat the data field. Representations of note are as follows:¶

• IPv4 internet addresses are written in the form [n.n.n.n], where n is in decimal notation and the brackets are optional¶
• IPv6 internet addresses are formulated based on the guidelines defined in [RFC5952].¶
• Timestamps (including reference, originate, receive, and transmit values) and the logical clock are represented in units of seconds and fractions, preferably in hexadecimal notation.¶
• Delay, offset, dispersion, and distance values are represented in units of milliseconds and fractions, preferably in decimal notation.¶
• All other values are represented as is, preferably in decimal notation.¶

Implementations may define variables other than those described in RFC 5905; called "extramural variables", these are distinguished by the inclusion of some character type other than alphanumeric or "." in the name. For those commands that return a list of assignments in the response data field, if the command data field is empty, it is expected that all available variables defined in RFC 5905 will be included in the response. For the read commands, if the command data field is nonempty, an implementation may choose to process this field to individually select which variables are to be returned.¶

Commands are interpreted as follows:¶

Read Status (1):

The command data field is empty or contains a list of identifiers separated by commas. The command operates in two ways depending on the value of the Association ID. If this identifier is nonzero, the response includes the peer identifier and status word. Optionally, the response data field may contain other information, such as described in the Read Variables command. If the association identifier is zero, the response includes the system identifier (0) and status word; the data field contains a list of binary-coded pairs <<Association ID>> <<status word>>, one for each currently defined association.¶

Read Variables (2):

The command data field is empty or contains a list of identifiers separated by commas. If the Association ID is nonzero, the response includes the requested peer identifier and status word; the data field contains a list of peer variables and values as described above. If the Association ID is zero, the data field contains a list of system variables. If a peer has been selected as the synchronization source, the response includes the peer identifier and status word; otherwise, the response includes the system identifier (0) and status word.¶

Write Variables (3):

The command data field contains a list of assignments as described above. The variables are updated as indicated. The response is as described for the Read Variables command.¶

Read Clock Variables (4):

The command data field is empty or contains a list of identifiers separated by commas. The Association ID selects the system clock variables or peer clock variables in the same way as in the Read Variables command. The response includes the requested clock identifier and status word; the data field contains a list of clock variables and values, including the last timecode message received from the clock.¶

Write Clock Variables (5):

The command data field contains a list of assignments as described above. The clock variables are updated as indicated. The response is as described for the read clock variables command.¶

Set Trap Address/Port (6):

The command Association ID, status, and data fields are ignored. The address and port number for subsequent trap messages are taken from the source address and port of the control message itself. The initial trap counter for trap response messages is taken from the sequence field of the command. The response association identifier, status, and data fields are not significant. Implementations should include logical timeouts that prevent trap transmissions if the monitoring program does not renew this information after a lengthy interval.¶

Trap Response (7):

This message is sent when a system, peer, or clock exception event occurs. The opcode field is 7 and the R bit is set. The trap counter is incremented by one for each trap sent and the sequence field set to that value. The trap message is sent using the IP address and port fields established by the set trap address/port command. If a system trap, the Association ID field is set to zero and the status field contains the system status word. If a peer trap, the Association ID field is set to that peer and the status field contains the peer status word. Optional ASCII-coded information can be included in the data field.¶

Configure (8):

The command data is parsed and applied as if supplied in the daemon configuration file.¶

Save Configuration (9):

Writes a snapshot of the current configuration to the file name supplied as the command data. Further, the command is refused unless a directory in which to store the resulting files has been explicitly configured by the operator.¶

Read Most Recently Used (MRU) list (10):

Retrieves records of recently seen remote addresses and associated statistics. This command supports all of the state variables defined in Section 9 of [RFC5905]. Command data consists of name=value pairs controlling the selection of records, as well as a requestor-specific nonce previously retrieved using this command or opcode 12 (Request Nonce). The response consists of name=value pairs where some names can appear multiple times using a dot followed by a zero-based index to distinguish them and to associate elements of the same record with the same index. A new nonce is provided with each successful response.¶

Read ordered list (11):

Retrieves a list ordered by IP address (IPv4 information precedes IPv6 information). If the command data is empty or is the seven characters "ifstats", the associated statistics, status, and counters for each local address are returned. If the command data is the characters "addr_restrictions", then the set of IPv4 remote address restrictions followed by the set of IPv6 remote address restrictions (access control lists) are returned. Other command data returns error code 5 (unknown variable name). Similar to Read MRU, response information uses zero-based indexes as part of the variable name preceding the equals sign and value, where each index relates information for a single address or network. This opcode requires authentication.¶

Request Nonce (12):

Retrieves a 96-bit nonce specific to the requesting remote address, which is valid for a limited period. Command data is not used in the request. The nonce consists of a 64-bit NTP timestamp and 32 bits of hash derived from that timestamp, the remote address, and salt known only to the server, which varies between daemon runs. Inclusion of the nonce by a management agent demonstrates to the server that the agent can receive datagrams sent to the source address of the request, making source address "spoofing" more difficult in a similar way as TCP's three-way handshake.¶

Unset Trap (31):

Removes the requesting remote address and port from the list of trap receivers. Command data is not used in the request. If the address and port are not in the list of trap receivers, the error code is 4 (bad association).¶

5. IANA Considerations

This document has no IANA actions.¶

6. Security Considerations

A number of security vulnerabilities have been identified with these control messages.¶

NTP's control query interface allows reading and writing of system, peer, and clock variables remotely from arbitrary IP addresses using commands mentioned in Section 4. Overwriting these variables, but not reading them, requires authentication by default. However, this document argues that an NTP host must authenticate all control queries and not just ones that overwrite these variables. Alternatively, the host can use an access control list to explicitly list IP addresses that are allowed to control query the clients. These access controls are required for the following reasons:¶

NTP as a Distributed Denial-of-Service (DDoS) vector:

NTP timing query and response packets (modes 1-2, 3-4, and 5) are usually short in size. However, some NTP control queries generate a very long packet in response to a short query. As such, there is a history of use of NTP's control queries, which exhibit such behavior, to perform DoS attacks. These off-path attacks exploit the large size of NTP control queries to cause UDP-based amplification attacks (e.g., mode 7 monlist command generates a very long packet in response to a small query [CVE-DOS]). These attacks only use NTP as a vector for DoS attacks on other protocols, but do not affect the time service on the NTP host itself. To limit the sources of these malicious commands, NTP server operators are recommended to deploy ingress filtering [RFC3704].¶

Time-shifting attacks through information leakage/overwriting:

NTP hosts save important system and peer state variables. An off-path attacker who can read these variables remotely can leverage the information leaked by these control queries to perform time-shifting and DDoS attacks on NTP clients. These attacks do affect time synchronization on the NTP hosts. For instance:¶

• In the client/server mode, the client stores its local time when it sends the query to the server in its xmt peer variable. This variable is used to perform TEST2 to non-cryptographically authenticate the server (i.e., if the origin timestamp field in the corresponding server response packet matches the xmt peer variable, then the client accepts the packet). An off-path attacker with the ability to read this variable can easily spoof server response packets for the client, which will pass TEST2 and can deny service or shift time on the NTP client. The specific attack is described in [CVE-SPOOF].¶
• The client also stores its local time when the server response is received in its rec peer variable. This variable is used for authentication in interleaved-pivot mode. An off-path attacker with the ability to read this state variable can easily shift time on the client by passing this test. This attack is described in [CVE-SHIFT].¶

Fast-Scanning:

NTP mode 6 control messages are usually small UDP packets. Fast-scanning tools like ZMap can be used to spray the entire (potentially reachable) Internet with these messages within hours to identify vulnerable hosts. To make things worse, these attacks can be extremely low-rate, only requiring a control query for reconnaissance and a spoofed response to shift time on vulnerable clients.¶

The mode 6 and 7 messages are vulnerable to replay attacks [CVE-Replay]:

If an attacker observes mode 6/7 packets that modify the configuration of the server in any way, the attacker can apply the same change at any time later by simply sending the packets to the server again. The use of the nonce (Request Nonce command) provides limited protection against replay attacks.¶

NTP best practices recommend configuring NTP with the no-query parameter. The no-query parameter blocks access to all remote control queries. However, sometimes the hosts do not want to block all queries and want to give access for certain control queries remotely. This could be for the purpose of remote management and configuration of the hosts in certain scenarios. Such hosts tend to use firewalls or other middleboxes to blacklist certain queries within the network.¶

Significantly fewer hosts respond to mode 7 monlist queries as compared to other control queries because it is a well-known and exploited control query. These queries are likely blocked using blacklists on firewalls and middleboxes rather than the no-query option on NTP hosts. The remaining control queries that can be exploited likely remain out of the blacklist because they are undocumented in the current NTP specification [RFC5905].¶

This document describes all of the mode 6 control queries allowed by NTP and can help administrators make informed decisions on security measures to protect NTP devices from harmful queries and likely make those systems less vulnerable. The use of the legacy mode 6 interface is NOT RECOMMENDED. Regardless of which mode 6 commands an administrator may elect to allow, remote access to this facility needs to be protected from unauthorized access (e.g., strict Access Control Lists (ACLs)). Additionally, the legacy interface for mode 6 commands SHOULD NOT be utilized in new deployments or implementation of NTP.¶

Appendix A. NTP Remote Facility Message Format

The format of the NTP Remote Facility Message header, which immediately follows the UDP header, is shown in Figure 3. A description of its fields follows Figure 3. Bit positions marked as zero are reserved and should always be transmitted as zero.¶

 0                   1                   2                   3
 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|R|M| VN  |Mode |A|  Sequence   | Implementation|   Req Code    |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|  Err  |        Count          |  MBZ  |       Size            |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|                                                               |
/                    Data (up to 500 bytes)                     /
|                                                               |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|                Encryption KeyID (when A bit set)              |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|                                                               |
/          Message Authentication Code (when A bit set)         /
|                                                               |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Response Bit (R):

Set to 0 if the packet is a request. Set to 1 if the packet is a response.¶

More Bit (M):

Set to 0 if this is the last packet in a response; otherwise, set to 1 in responses requiring more than one packet.¶

Version Number (VN):

Set to the version number of the NTP daemon.¶

Mode:

Set to 7 for Remote Facility messages.¶

Authenticated Bit (A):

If set to 1, this packet contains authentication information.¶

Sequence:

For a multi-packet response, this field contains the sequence number of this packet. Packets in a multi-packet response are numbered starting with 0. The More Bit is set to 1 for all packets but the last.¶

Implementation:

The version number of the implementation that defined the request code used in this message. An implementation number of 0 is used for a request code supported by all versions of the NTP daemon. The value 255 is reserved for future extensions.¶

Request Code (Req Code):

An implementation-specific code that specifies the operation being requested. A request code definition includes the format and semantics of the data included in the packet.¶

Error (Err):

Set to 0 for a request. For a response, this field contains an error code relating to the request. If the Error is nonzero, the operation requested wasn't performed.¶

0:

no error¶

1:

incompatible implementation number¶

2:

unimplemented request code¶

3:

format error¶

4:

no data available¶

7:

authentication failure¶

Count:

The number of data items in the packet. Range is 0 to 500.¶

Must Be Zero (MBZ):

A reserved field set to 0 in requests and responses.¶

Size:

The size of each data item in the packet. Range is 0 to 500.¶

Data:

A variable-sized field containing request/response data. For requests and responses, the size in octets must be greater than or equal to the product of the number of data items (Count) and the size of a data item (Size). For requests, the data area is exactly 40 octets in length. For responses, the data area will range from 0 to 500 octets, inclusive.¶

Encryption KeyID:

A 32-bit unsigned integer used to designate the key used for the Message Authentication Code. This field is included only when the A bit is set to 1.¶

Message Authentication Code:

An optional Message Authentication Code defined by the version of the NTP daemon indicated in the Implementation field. This field is included only when the A bit is set to 1.¶

Acknowledgements

Tim Plunkett created the original version of this document. Aanchal Malhotra provided the initial version of the Security Considerations section.¶

Karen O'Donoghue, David Hart, Harlan Stenn, and Philip Chimento deserve credit for portions of this document due to their earlier efforts to document these commands.¶

Miroshav Lichvar, Ulrich Windl, Dieter Sibold, J Ignacio Alvarez-Hamelin, and Alex Campbell provided valuable comments on various draft versions of this document.¶

Author's Address