nsupdate
Zones that are under dynamic control via nsupdate or a DHCP server should not be edited by hand. Manual edits could conflict with dynamic updates and cause data to be lost.
The resource records that are dynamically added or removed with nsupdate have to be in the same zone. Requests are sent to the zone's master server. This is identified by the MNAME field of the zone's SOA record.
The -d option makes nsupdate operate in debug mode. This provides tracing information about the update requests that are made and the replies received from the name server.
The -D option makes nsupdate report additional debugging information to -d.
Transaction signatures can be used to authenticate the Dynamic DNS updates. These use the TSIG resource record type described in RFC2845 or the SIG(0) record described in RFC3535 and RFC2931 or GSS-TSIG as described in RFC3645. TSIG relies on a shared secret that should only be known to nsupdate and the name server. Currently, the only supported encryption algorithm for TSIG is HMAC-MD5, which is defined in RFC 2104. Once other algorithms are defined for TSIG, applications will need to ensure they select the appropriate algorithm as well as the key when authenticating each other. For instance, suitable key and server statements would be added to /etc/named.conf so that the name server can associate the appropriate secret key and algorithm with the IP address of the client application that will be using TSIG authentication. SIG(0) uses public key cryptography. To use a SIG(0) key, the public key must be stored in a KEY record in a zone served by the name server. nsupdate does not read /etc/named.conf. GSS-TSIG uses Kerberos credentials.
nsupdate uses the -y or -k option to provide the shared secret needed to generate a TSIG record for authenticating Dynamic DNS update requests, default type HMAC-MD5. These options are mutually exclusive. With the -k option, nsupdate reads the shared secret from the file keyfile, whose name is of the form K{name}.+157.+{random}.private. For historical reasons, the file K{name}.+157.+{random}.key must also be present. When the -y option is used, a signature is generated from [hmac:]keyname:secret. keyname is the name of the key, and secret is the base64 encoded shared secret. Use of the -y option is discouraged because the shared secret is supplied as a command line argument in clear text. This may be visible in the output from ps(1) or in a history file maintained by the user's shell.
The -k may also be used to specify a SIG(0) key used to authenticate Dynamic DNS update requests. In this case, the key specified is not an HMAC-MD5 key.
The -g and -o specify that GSS-TSIG is to be used. The -o should only be used with old Microsoft Windows 2000 servers.
By default, nsupdate uses UDP to send update requests to the name server unless they are too large to fit in a UDP request in which case TCP will be used. The -v option makes nsupdate use a TCP connection. This may be preferable when a batch of update requests is made.
The -t option sets the maximum time an update request can take before it is aborted. The default is 300 seconds. Zero can be used to disable the timeout.
The -u option sets the UDP retry interval. The default is 3 seconds. If zero, the interval will be computed from the timeout interval and number of UDP retries.
The -r option sets the number of UDP retries. The default is 3. If zero, only one update request will be made.
The -R randomdev option specifies a source of randomness. If the operating system does not provide a /dev/random or equivalent device, the default source of randomness is keyboard input. randomdev specifies the name of a character device or file containing random data to be used instead of the default. The special value keyboard indicates that keyboard input should be used. This option may be specified multiple times.
nsupdate reads input from filename or standard input. Each command is supplied on exactly one line of input. Some commands are for administrative purposes. The others are either update instructions or prerequisite checks on the contents of the zone. These checks set conditions that some name or set of resource records (RRset) either exists or is absent from the zone. These conditions must be met if the entire update request is to succeed. Updates will be rejected if the tests for the prerequisite conditions fail.
Every update request consists of zero or more prerequisites and zero or more updates. This allows a suitably authenticated update request to proceed if some specified resource records are present or missing from the zone. A blank input line (or the send command) causes the accumulated commands to be sent as one Dynamic DNS update request to the name server.
The command formats and their meaning are as follows:
server {servername} [port]
local {address} [port]
zone {zonename}
class {classname}
ttl {seconds}
key {name} {secret}
prereq nxdomain {domain-name}
prereq yxdomain {domain-name}
prereq nxrrset {domain-name} [class] {type}
prereq yxrrset {domain-name} [class] {type}
prereq yxrrset {domain-name} [class] {type} {data...}
update delete {domain-name} [ttl] [class] [type [data...]]
update add {domain-name} {ttl} [class] {type} {data...}
show
send
answer
debug
Lines beginning with a semicolon are comments and are ignored.
The examples below show how nsupdate could be used to insert and delete resource records from the example.com zone. Notice that the input in each example contains a trailing blank line so that a group of commands are sent as one dynamic update request to the master name server for example.com.
# nsupdate > update delete oldhost.example.com A > update add newhost.example.com 86400 A 172.16.1.1 > send
Any A records for oldhost.example.com are deleted. And an A record for newhost.example.com with IP address 172.16.1.1 is added. The newly-added record has a 1 day TTL (86400 seconds).
# nsupdate > prereq nxdomain nickname.example.com > update add nickname.example.com 86400 CNAME somehost.example.com > send
The prerequisite condition gets the name server to check that there are no resource records of any type for nickname.example.com. If there are, the update request fails. If this name does not exist, a CNAME for it is added. This ensures that when the CNAME is added, it cannot conflict with the long-standing rule in RFC1034 that a name must not exist as any other record type if it exists as a CNAME. (The rule has been updated for DNSSEC in RFC2535 to allow CNAMEs to have RRSIG, DNSKEY and NSEC records.)
/etc/resolv.conf
K{name}.+157.+{random}.key
K{name}.+157.+{random}.private
RFC2136(), RFC3007(), RFC2104(), RFC2845(), RFC1034(), RFC2535(), RFC2931(), named(8), dnssec-keygen(8).
The TSIG key is redundantly stored in two separate files. This is a consequence of nsupdate using the DST library for its cryptographic operations, and may change in future releases.