When a client application connects to the database server, it specifies which PostgreSQL user name it wants to connect as, much the same way one logs into a Unix computer as a particular user. Within the SQL environment the active database user name determines access privileges to database objects -- see Chapter 7 for more information about that. It is therefore obviously essential to restrict which database user name(s) a given client can connect as.
Authentication is the process by which the database server establishes the identity of the client, and by extension determines whether the client application (or the user who runs the client application) is permitted to connect with the user name that was requested.
PostgreSQL offers a number of different client authentication methods. The method to be used can be selected on the basis of (client) host and database; some authentication methods allow you to restrict by user name as well.
PostgreSQL database user names are logically separate from user names of the operating system in which the server runs. If all the users of a particular server also have accounts on the server's machine, it makes sense to assign database user names that match their operating system user names. However, a server that accepts remote connections may have many users who have no local account, and in such cases there need be no connection between database user names and OS user names.
Client authentication is controlled by the file pg_hba.conf in the data directory, e.g., /usr/local/pgsql/data/pg_hba.conf. (HBA stands for host-based authentication.) A default pg_hba.conf file is installed when the data area is initialized by initdb.
The general format of the pg_hba.conf file is of a set of records, one per line. Blank lines and lines beginning with a hash character ("#") are ignored. A record is made up of a number of fields which are separated by spaces and/or tabs. Records cannot be continued across lines.
Each record specifies a connection type, a client IP address range (if relevant for the connection type), a database name or names, and the authentication method to be used for connections matching these parameters. The first record that matches the type, client address, and requested database name of a connection attempt is used to do the authentication step. There is no "fall-through" or "backup": if one record is chosen and the authentication fails, the following records are not considered. If no record matches, the access will be denied.
A record may have one of the three formats
local database authentication-method [ authentication-option ]
host database IP-address IP-mask authentication-method [ authentication-option ]
hostssl database IP-address IP-mask authentication-method [ authentication-option ]
The meaning of the fields is as follows:
This record pertains to connection attempts over Unix domain sockets.
This record pertains to connection attempts over TCP/IP networks. Note that TCP/IP connections are completely disabled unless the server is started with the -i switch or the equivalent configuration parameter is set.
This record pertains to connection attempts with SSL over TCP/IP. To make use of this option the server must be built with SSL support enabled. Furthermore, SSL must be enabled with the -l option or equivalent configuration setting when the server is started. (Note: host records will match either SSL or non-SSL connection attempts, but hostssl records match only SSL connections.)
Specifies the database that this record applies to. The value all specifies that it applies to all databases, while the value sameuser identifies the database with the same name as the connecting user. Otherwise, this is the name of a specific PostgreSQL database.
These two fields specify to which client machines a
host or hostssl
record applies, based on their IP
address. (Of course IP addresses can be spoofed but this
consideration is beyond the scope of
PostgreSQL.) The precise logic is that
must be zero for the record to match.
Specifies the method that users must use to authenticate themselves when connecting under the control of this authentication record. The possible choices are summarized here, details are in Section 4.2.
The connection is allowed unconditionally. This method allows any user that has login access to the client host to connect as any PostgreSQL user whatsoever.
The connection is rejected unconditionally. This is mostly useful to "filter out" certain hosts from a group.
The client is required to supply a password which is required to match the database password that was set up for the user.
An optional file name may be specified after the password keyword. This file is expected to contain a list of users who may connect using this record, and optionally alternative passwords for them.
The password is sent over the wire in clear text. For better protection, use the md5 or crypt methods.
Like the password method, but the password is sent over the wire encrypted using a simple challenge-response protocol. This protects against incidental wire-sniffing. This is now the recommended choice for password-based authentication.
The name of a file may follow the md5 keyword. It contains a list of users who may connect using this record.
Like the md5 method but uses older crypt encryption, which is needed for pre-7.2 clients. md5 is preferred for 7.2 and later clients. The crypt method is not compatible with encrypting passwords in pg_shadow, and may fail if client and server machines have different implementations of the crypt() library routine.
Kerberos V4 is used to authenticate the user. This is only available for TCP/IP connections.
Kerberos V5 is used to authenticate the user. This is only available for TCP/IP connections.
The identity of the user as determined on login to the operating system is used by PostgreSQL to determine whether the user is allowed to connect as the requested database user. For TCP/IP connections the user's identity is determined by contacting the ident server on the client host. (Note that this is only as reliable as the remote ident server; ident authentication should never be used for remote hosts whose administrators are not trustworthy.) On operating systems supporting SO_PEERCRED requests for Unix domain sockets, ident authentication is possible for local connections; the system is then asked for the connecting user's identity.
On systems without SO_PEERCRED requests, ident authentication is only available for TCP/IP connections. As a workaround, it is possible to specify the localhost address 127.0.0.1 and make connections to this address.
The authentication option following the ident keyword specifies the name of an ident map that specifies which operating system users equate with which database users. See below for details.
This authentication type operates similarly to password, with the main difference that it will use PAM (Pluggable Authentication Modules) as the authentication mechanism. The authentication option following the pam keyword specifies the service name that will be passed to PAM. The default service name is postgresql. For more information about PAM, please read the Linux-PAM Page and/or the Solaris PAM Page.
This field is interpreted differently depending on the authentication method, as described above.
Since the pg_hba.conf records are examined sequentially for each connection attempt, the order of the records is very significant. Typically, earlier records will have tight connection match parameters and weaker authentication methods, while later records will have looser match parameters and stronger authentication methods. For example, one might wish to use trust authentication for local TCP connections but require a password for remote TCP connections. In this case a record specifying trust authentication for connections from 127.0.0.1 would appear before a record specifying password authentication for a wider range of allowed client IP addresses.
An example of a pg_hba.conf file is shown in Example 4-1. See below for details on the different authentication methods.