ONC+ Developer's Guide

Chapter 9 NIS+ Programming Guide

This chapter presents the fundamental principles of the NIS+ (pronounced “niss-plus”) applications programming interface and a detailed sample program. The NIS+ API is for programmers who need to build applications for Solaris networks. It provides the essential features for supporting enterprise-wide applications.

This chapter covers the following topics:

The NIS+ network name service addresses the requirements of client/server networks ranging in size from 10 clients supported by a few servers on a local area network to 10,000 multi-vendor clients supported by 20 to 100 specialized servers located in sites throughout the world and connected by several public networks.

NIS+ Overview

This section describes various aspects of the NIS+ network name service.

NIS+ Domains

NIS+ supports hierarchical domains, as illustrated in the following figure.

Figure 9–1 NIS+ Domain

This graphic depicts a typical domain tree, with four subdomains branching off a main domain.

A NIS+ domain is a set of data describing the workstations, users, and network services in a portion of an organization. NIS+ domains can be administered independently of each other. This independence enables NIS+ to be used in a range of networks, from small to very large.

NIS+ and Servers

Each domain is supported by a set of servers. The principal server is called the master server, and the backup servers are called replicas. Both master and replica servers run NIS+ server software. The master server stores the original tables, and the backup servers store copies.

NIS+ accepts incremental updates to the replicas. Changes are first made on the master server. Then they are automatically propagated to the replica servers and are soon available to the entire namespace.

NIS+ Tables

NIS+ stores information in tables instead of maps or zone files. NIS+ provides 16 types of predefined, or system, tables, which are named in the following list:

Each table stores a different type of information. For instance, the Hosts table stores host name/Internet address pairs, and the Password table stores information about users of the network.

NIS+ tables have two major improvements over NIS maps. First, a NIS+ table can be accessed by any column, not just the first column, which is sometimes referred to as the “key.” This access eliminates the need for duplicate maps, such as the hosts.byname and hosts.byaddr maps of NIS. Second, access to the information in NIS+ tables can be controlled at three levels of granularity: the table level, the entry level, and the column level.

NIS+ Security

The NIS+ security model provides both authorization and authentication mechanisms. For authorization, every object in the namespace specifies the type of operation it accepts and from whom. NIS+ attempts to authenticate every requestor accessing the namespace. After it identifies the originator of the request, it determines whether the object has authorized that particular operation for that particular principal. Based on its authentication and the object's authorization, NIS+ carries out or denies the access request.

Name Service Switch

NIS+ works in conjunction with a separate facility called the Name Service Switch. The Name Service Switch, sometimes referred to as “the Switch,” enables Solaris-based workstations to obtain their information from more than one network information service. They can get the information from local, or /etc files, from NIS maps, from DNS zone files, or from NIS+ tables. The Switch not only offers a choice of sources, but allows a workstation to specify different sources for different types of information. The name service is configured through the file /etc/nsswitch.conf.

NIS+ Administration Commands

NIS+ provides a full set of commands for administering a namespace, as listed in the following table.

Table 9–1 NIS+ Namespace Administration Commands




Changes the group owner of a NIS+ object. 


Changes an object's access rights. 


Changes the owner of a NIS+ object. 


Creates or destroys a NIS+ group, or displays a list of its members. Also adds members to a group, removes them, or tests them for membership in the group. 


Displays the contents of NIS+ tables. 


Searches for entries in a NIS+ table. 


Lists the contents of a NIS+ directory. 


Searches for entries in a NIS+ table. 


Adds information from /etc files or NIS maps into NIS+ tables.


Creates or deletes NIS+ tables, and adds, modifies, or deletes entries in a NIS+ table.  


Creates credentials for NIS+ principals and stores them in the Cred table. 


Changes password information stored in the NIS+ Passwd table.  


Updates the public keys stored in a NIS+ object. 


Initializes a NIS+ client or server. 


Creates a NIS+ directory and specifies its master and replica servers. 


Removes NIS+ directories and replicas from the namespace. 


Creates org_dir and groups_dir directories and a complete set of (unpopulated) NIS+ tables for a NIS+ domain.


The NIS+ server process. 


Starts the NIS+ Cache Manager on a NIS+ client. 


Changes a NIS+ object's time to live value. 


Lists a NIS+ object's default values: domain name, group name, workstation name, NIS+ principal name, access rights, directory search path, and time-to-live. 


Creates a symbolic link between two NIS+ objects. 


Removes NIS+ objects (except directories) from the namespace. 


Lists the contents of the NIS+ shared cache maintained by the NIS+ Cache Manager.


The NIS+ application programming interface (API) is a group of functions that can be called by an application to access and modify NIS+ objects. The NIS+ API has 54 functions that fall into nine categories:

The functions in each category are summarized in the following table. The category names match the names by which they are grouped in the NIS+ man pages.

Table 9–2 NIS+ API Functions




Locates and manipulates objects. 


Returns a copy of a NIS+ object. Can follow links. Though it cannot search for an entry object, if a link points to one, it can return an entry object.  


Adds a NIS+ object to the namespace. 


Removes a NIS+ object in the namespace. 


Modifies a NIS+ object in the namespace.  


Searches and updates tables.  


Searches a table in the NIS+ namespace and returns entry objects that match the search criteria. Can follow links and search paths from one table to another. 


Adds an entry object to a NIS+ table. Can be instructed to either fail or overwrite if the entry object already exists. Can return a copy of the resulting object if the operation was successful. 


Frees all memory associated with a nis_result structure.


Removes one or more entry objects from a NIS+ table. Can identify the object to be removed by using search criteria or by pointing to a cached copy of the object. If using search criteria, can remove all objects that match the search criteria; therefore, with the proper search criteria, can remove all entries in a table. Can return a copy of the resulting object if the operation was successful. 


Modifies one or more entry objects in a NIS+ table. Can identify the object to be modified by using search criteria or by pointing to a cached copy of the object.  


Returns a copy of the first entry object in a NIS+ table.  


Returns a copy of the “next” entry object in a NIS+ table. Because a table can be updated and entries removed or modified between calls to this function, the order of entries returned might not match the actual order of entries in the table.  


Gets default names for the current process. 


Returns the name of the workstation's NIS+ domain. 


Returns the fully qualified name of the workstation. A fully qualified name has the form host-name.domain-name.


Returns the name of the current NIS+ group, which is specified by the environment variable NIS_GROUP. 


Returns the name of the NIS+ principal that has a UID associated with the calling process. 


Returns a list of possible expansions to a particular name. 


Frees the memory containing the list generated by nis_getnames().


Group manipulation and authorization. 


Tests whether a principal is a member of a group. 


Adds a member to a group. The member can be a principal, a group, or a domain. 


Deletes a member from a group. 


Creates a group object. 


Deletes a group object. 


Tests whether a group object exists. 


Lists the principals that are members of a group object. 


Various services for NIS+ applications. 


Creates the databases to support service for a named directory on a specified host.  


Removes the directory from a host. 


Sets and reads state variables of NIS+ servers and flushes internal caches.  


Retrieves statistics about a server's performance.  


Returns a list of servers that support a particular domain.  


Frees the list of servers returned by nis_getservlist().


Frees the memory associated with the results of nis_servstate() and nis_stats().


Interface between the NIS+ server and the database. Not to be used by a NIS+ client. 


Returns a copy of the first entry of the specified table. 


Returns a copy of the entry succeeding the specified entry. 


Terminates a first/next entry sequence. 


Returns copies of entries that meet specified attributes. 


Removes all entries that meet specified attributes. 


Replaces an entry in a table identified by specified attributes with a copy of the specified object, or adds the object to the table. 


Reorganizes the contents of a table to make access to the table more efficient. 


Advises the database manager to release resources. 


Functions that supply descriptive strings equivalent to NIS+ status values. 


Returns a pointer to the appropriate string constant. 


Displays the appropriate string constant on standard output. 


Sends the appropriate string constant to syslog.


Returns a pointer to a statically allocated string to be used or to be copied with strdup().


Transaction logging functions used by servers. 


Used by the master server of a directory to timestamp it. This forces replicas of the directory to be updated. 


Forces logged data to be stored in the table on disk. 


Functions to help operate on NIS+ names and objects. 


Returns the first label in a NIS+ name. The returned name does not have a trailing dot.  


Removes all domain-related labels and returns only the unique object portion of the name. The name passed to the function must be either in the local domain or in one of its child domains, or the function returns NULL. 


Returns the name of the domain in which an object resides. The returned name ends in a dot.  


Compares any two NIS+ names. The comparison ignores case and states whether the names are the same, descendants of each other, or not related. 


Creates an exact duplicate of a NIS+ object.


Destroys an object created by nis_clone_object().


Prints the contents of a NIS+ object structure to stdout.

NIS+ Sample Program

This sample program performs the following tasks:

The example program is not a typical application. In a normal situation the directories and tables would be created or removed through the command line interface, and applications would manipulate NIS+ entry objects.

Unsupported Macros

The sample program uses unsupported macros that are defined in the file <rpcsvc/nis.h>. These macros are not public APIs and can change or disappear in the future. They are used for illustration purposes only and if you choose to use them, you do so at your own risk. The macros used are:

Functions Used in the Example

The use of the following NIS+ C API functions is illustrated through this example.

Program Compilation

The NIS+ principal running this application has permission to create directory objects in the local domain. The program is compiled by typing:

% cc -o example.c example -lnsl

It is invoked by typing:

% example [dir]

where dir is the NIS+ directory in which the program creates all the NIS+ objects. Specifying no directory argument causes the objects to be created in the parent directory of the local domain. Note that for the call to nis_lookup(), a space and the name of the local domain are appended to the string that names the directory. The argument is the name of the NIS+ directory in which to create the NIS+ objects. The principal running this program should have create permission in the directory.

The following code example shows the routine is called by main() to create directory objects.

Example 9–1 NIS+ Routine to Create Directory Objects

dir_create (dir_name, dirobj)
	nis_name			dir_name;
	nis_object			*dirobj;
	nis_result 	*cres;
	nis_error			err;

	printf ("\n Adding Directory %s to namespace ... \n",
	cres = nis_add (dir_name, dirobj);

	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status, "unable to add directory foo.");
		exit (1);

	(void) nis_freeresult (cres);

	 * NOTE: you need to do a nis_mkdir to create the table to
store the
	 * contents of the directory you are creating.
	err = nis_mkdir (dir_name,

	if (err != NIS_SUCCESS) {
		(void) nis_remove (dir_name, 0);

		nis_perror (err,
				"unable to create table for  directory object foo.");
		exit (1);

This routine is called by main() to create the group object. Because nis_creategroup() works only on group objects, the “groups_dir” literal is not needed in the group name.

Example 9–2 NIS+ Routine to Create Group Objects

grp_create (grp_name)
	nis_name			grp_name;
	nis_error			err;

	printf ("\n Adding %s group to namespace ... \n", grp_name);
	err = nis_creategroup (grp_name, 0);
	if (err != NIS_SUCCESS) {
		nis_perror (err, "unable to create group.");
		exit (1);

The routine shown in the following example is called by main() to create a table object laid out as shown in the following table.

Table 9–3 NIS+ Table Objects








Searchable, Text 

Searchable, Text 

Access Rights 



The TA_SEARCHABLE constant indicates to the service that the column is searchable. Only TEXT (the default) columns are searchable. TA_CASE indicates to the service that the column value is to be treated in a case-insensitive manner during searches.

Example 9–3 NIS+ Routine to Create Table Objects

#define			TABLE_MAXCOLS 2
#define			TABLE_COLSEP ':'
#define			TABLE_PATH 0

tbl_create (dirobj, table_name)
	nis_object 	*dirobj;		/* need to use some of the fields */
	nis_name			table_name;
	nis_result				*cres;
	static nis_object					tblobj;
	static table_col					tbl_cols[TABLE_MAXCOLS] = {

	tblobj.zo_owner = dirobj->zo_owner;
	tblobj.zo_group = dirobj->zo_group;
	tblobj.zo_access = DEFAULT_RIGHTS;    /* macro defined in nis.h  */
	tblobj.zo_data.zo_type = TABLE_OBJ;   /* enumerated type in nis.h */
	tblobj.TA_data.ta_type = TABLE_TYPE;
	tblobj.TA_data.ta_maxcol = TABLE_MAXCOLS;
	tblobj.TA_data.ta_sep = TABLE_COLSEP;
	tblobj.TA_data.ta_path = TABLE_PATH;
	tblobj.TA_data.ta_cols.ta_cols_len =
		tblobj.TA_data.ta_maxcol;				 /* ALWAYS ! */
	tblobj.TA_data.ta_cols.ta_cols_val = tbl_cols;

 * Use a fully qualified table name i.e. the "org_dir" literal

 * be embedded in the table name. This is necessary because
 * operates on all types of NIS+ objects and needs the full path
 * if a table is created.
	printf ("\n Creating table %s ... \n", table_name);
	cres = nis_add (table_name, &tblobj);
	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status, "unable to add table.");
		exit (1);
	(void) nis_freeresult (cres);

The routine shown in the following example is called by main() to add entry objects to the table object. Two entries are added to the table object. Note that the column width in both entries is set to include the NULL character for a string terminator.

Example 9–4 NIS+ Routine to Add Objects to Table

#define			MAXENTRIES 2
	nis_name 	table_name;
	int			i;
	nis_object			entdata;
	nis_result			*cres;
	static entry_col ent_col_data[MAXENTRIES][TABLE_MAXCOLS] = {
			{0, 2, "1", 0, 5, "John"},
			{0, 2, "2", 0, 5, "Mary"}

	printf ("\n Adding entries to table ... \n");

	 * Look up the table object first since the entries being added
	 * should have the same owner, group owner and access rights as
	 * the table they go in.
	cres = nis_lookup (table_name, 0);

	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status, "Unable to lookup table");
	entdata.zo_owner = NIS_RES_OBJECT (cres)->zo_owner;
	entdata.zo_group = NIS_RES_OBJECT (cres)->zo_group;
	entdata.zo_access = NIS_RES_OBJECT (cres)->zo_access;
	/* Free cres, so that it can be reused. */
	(void) nis_freeresult (cres);

	entdata.zo_data.zo_type = ENTRY_OBJ; /* enumerated type in
nis.h */
	entdata.EN_data.en_type = TABLE_TYPE;
	entdata.EN_data.en_cols.en_cols_len = TABLE_MAXCOLS;
	for (i = 0; i < MAXENTRIES; ++i) {
		entdata.EN_data.en_cols.en_cols_val = &ent_col_data[i][0];
		cres = nis_add_entry (table_name, &entdata, 0);

		if (cres->status != NIS_SUCCESS) {
			nis_perror (cres->status, "unable to add entry.");
			exit (1);
		(void) nis_freeresult (cres);

The routine shown in the following example is the print function for the nis_list() call. When list_objs() calls nis_list(), a pointer to print_info() is one of the calling arguments. Each time the service calls this function, it prints the contents of the entry object. The return value indicates to the library to call with the next entry from the table.

Example 9–5 NIS+ Routine for nis_list Call

print_info (name, entry, cbdata)
	nis_name			name;			/* Unused */
	nis_object			*entry;			/* The NIS+ entry object */
	void			*cbdata;			/* Unused */
	static u_int				firsttime = 1;
	entry_col	*tmp;		/* only to make source more readable */
	u_int				i, terminal;

	if (firsttime) {
		printf ("\tId.\t\t\tName\n");
		printf ("\t---\t\t\t----\n");
		firsttime = 0;
	for (i = 0; i < entry->EN_data.en_cols.en_cols_len; ++i) {
		tmp = &entry->EN_data.en_cols.en_cols_val[i];
		terminal = tmp->ec_value.ec_value_len;
		tmp->ec_value.ec_value_val[terminal] = '\0';

	 * ENTRY_VAL is a macro that returns the value of a specific
	 * column value of a specified entry.
	printf("\t%s\t\t\t%s\n", ENTRY_VAL (entry, 0),
								ENTRY_VAL (entry, 1));
	return (0); /* always ask for more */

The routine shown in the following example is called by main() to list the contents of the group, table, and directory objects. The routine demonstrates the use of callbacks also. It retrieves and displays the membership of the group. The group membership list is not stored as the contents of the object. So, the list is queried through the nis_lookup() instead of the nis_list() call. You must use the “groups_dir” form of the group because nis_lookup() works on all types of NIS+ objects.

Example 9–6 NIS+ Routine to List Objects

list_objs(dir_name, table_name, grp_name)
         nis_name   dir_name, table_name, grp_name;
	group_obj			*tmp; 	/* only to make source more readable */
	u_int			i;
	char			grp_obj_name [NIS_MAXNAMELEN];
	nis_result			*cres;
	char			index_name [BUFFER_SIZE];

	sprintf (grp_obj_name, "%s.groups_dir.%s",
				nis_leaf_of (grp_name), nis_domain_of (grp_name));
	printf ("\nGroup %s membership is: \n", grp_name);

	cres = nis_lookup(grp_obj_name, 0);

	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status, "Unable to lookup group object.");

	tmp = &(NIS_RES_OBJECT(cres)->GR_data);
	for (i = 0; i < tmp->gr_members.gr_members_len; ++i)
		printf ("\t %s\n", tmp->gr_members.gr_members_val[i]);
	(void) nis_freeresult (cres);

	 * Display the contents of the foo domain without using
	printf ("\nContents of Directory %s are: \n", dir_name);
	cres = nis_list (dir_name, 0, 0, 0);
	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status,
                                "Unable to list Contents of
Directory foo.");
	for (i = 0; i < NIS_RES_NUMOBJ(cres); ++i)
		printf("\t%s\n", NIS_RES_OBJECT(cres)[i].zo_name);
	(void) nis_freeresult (cres);

	 * List the contents of the table we created using the callback
	 * of nis_list().
	printf ("\n Contents of Table %s are: \n", table_name);
	cres = nis_list (table_name, 0, print_info, 0);
	if(cres->status != NIS_CBRESULTS && cres->status !=
		nis_perror (cres->status,
			"Listing entries using callback failed");
	(void) nis_freeresult (cres);


	 * List only one entry from the table we created. Use 
	 * indexed names to do this retrieval.

	printf("\n Entry corresponding to id 1 is:\n");
	 * The name of the column is usually extracted from the table
	 * object, which would have to be retrieved first.
	sprintf(index_name, "[Id=1],%s", table_name);
	cres = nis_list (index_name, 0, print_info, 0);
	if(cres->status != NIS_CBRESULTS && cres->status !=
		nis_perror (cres->status,
		"Listing entry using indexed names and callback failed");
	(void) nis_freeresult (cres);

The routine in the following table is called by cleanup() to remove a directory object from the namespace. The routine also informs the servers serving the directory about this deletion. Note that the memory containing result structure, pointed to by cres, must be freed after the result has been tested.

Example 9–7 NIS+ Routine to Remove Directory Objects

dir_remove(dir_name, srv_list, numservers)
	nis_name			dir_name;
	nis_server			*srv_list;
	u_int			numservers;
	nis_result			*cres;
	nis_error			err;
	u_int			i;

	printf ("\nRemoving %s directory object from namespace ...
	cres = nis_remove (dir_name, 0);
	if (cres->status != NIS_SUCCESS) {
		nis_perror (cres->status, "unable to remove directory");
		exit (1);
	(void) nis_freeresult (cres);

	for (i = 0; i < numservers; ++i) {
		err = nis_rmdir (dir_name, &srv_list[i]);
		if (err != NIS_SUCCESS) {
			nis_perror (err,
			"unable to remove server from directory");
			exit (1);

The following routine is called by main() to delete all the objects that were created in this example. Note the use of the REM_MULTIPLE flag in the call to nis_remove_entry(). You must delete all entries from a table before the table itself can be deleted.

Example 9–8 NIS+ Routine to Remove All Objects

cleanup(local_princip, grp_name, table_name, dir_name, dirobj)
	nis_name			local_princip, grp_name, table_name, dir_name;
	nis_object			*dirobj;
	char		grp_dir_name [NIS_MAXNAMELEN];
	char		org_dir_name [NIS_MAXNAMELEN];
	nis_error			err;
	nis_result			*cres;

	sprintf(grp_dir_name, "%s.%s", "groups_dir", dir_name);
	sprintf(org_dir_name, "%s.%s", "org_dir", dir_name);

	printf("\n\n\nStarting to Clean up ... \n");
	printf("\n\nRemoving principal %s from group %s \n",
						local_princip, grp_name);
	err = nis_removemember (local_princip, grp_name);

	if (err != NIS_SUCCESS) {
		nis_perror (err,
		"unable to delete local principal from group.");
		exit (1);

	 * Delete the admins group. We do not use the "groups_dir" form
	 * of the group name since this API is applicable to groups
	 * It automatically embeds the groups_dir literal in the name
	 * the group.
	printf("\nRemoving %s group from namespace ... \n",
	err = nis_destroygroup (grp_name);
	if (err != NIS_SUCCESS) {
		nis_perror (err, "unable to delete group.");
		exit (1);

	printf("\n Deleting all entries from table %s ... \n",

	cres = nis_remove_entry(table_name, 0, REM_MULTIPLE);
	switch (cres->status) {
			nis_perror(cres->status, "Could not delete entries from
	(void) nis_freeresult (cres);

	printf("\n Deleting table %s itself ... \n", table_name);
	cres = nis_remove(table_name, 0);

	if (cres->status != NIS_SUCCESS) {
		nis_perror(cres->status, "Could not delete table.");
	(void) nis_freeresult (cres);

	/* delete the groups_dir, org_dir and foo  directory objects.

	dir_remove (grp_dir_name,
	dir_remove (org_dir_name,
	dir_remove (dir_name, dirobj-

Running the program displays on the screen, as shown in the following code example.

Example 9–9 NIS+ Program Execution

myhost% domainname
 myhost% ./sample
 Adding Directory foo.sun.com. to namespace ...
 Adding Directory groups_dir.foo.sun.com. to namespace ...
 Adding Directory org_dir.foo.sun.com. to namespace ...
 Adding admins.foo.sun.com. group to namespace ...
 Adding principal myhost.sun.com. to group admins.foo.sun.com. ...
 Creating table test_table.org_dir.foo.sun.com. ...
 Adding entries to table ...
 Group admins.foo.sun.com. membership is:
 Contents of Directory foo.sun.com. are:

 Contents of Table test_table.org_dir.foo.sun.com. are:
         Id.                     Name
         ---                     ----
         1                       John
         2                       Mary

 Entry corresponding to id 1 is:
         1                       John

 Starting to Clean up ...

 Removing principal myhost.sun.com. from group admins.foo.sun.com.
 Removing admins.foo.sun.com. group from namespace ...
 Deleting all entries from table test_table.org_dir.foo.sun.com. ...
 Deleting table test_table.org_dir.foo.sun.com. itself ...
 Removing groups_dir.foo.sun.com. directory object from namespace ...
 Removing org_dir.foo.sun.com. directory object from namespace ...
 Removing foo.sun.com. directory object from namespace ...

As a debugging aid, the same operations are performed by the following command sequence. The first command displays the name of the master server.

Substitute the master server name where the variable master appears in the following.

% niscat -o `domainname`
% nismkdir -m master foo.`domainname`.
 # Create the org_dir.foo subdirectory with the specified master
 % nismkdir -m master org_dir.foo.`domainname`.
 # Create the groups_dir.foo subdirectory with the specified master
 % nismkdir -m master groups_dir.foo.`domainname`.
 # Create the “admins” group
 % nisgrpadm -c admins.foo.`domainname`.
 # Add yourself as a member of this group
 % nisgrpadm -a admins.foo.`domainname`. `nisdefaults -p`
 # Create a test_table with two columns : Id and Name
 % nistbladm -c test_data id=SI Name=SI \
 # Add one entry to that table.
 % nistbladm -a id=1 Name=John test_table.org_dir.foo.`domainname`.
 # Add another entry to that table.
 % nistbladm -a id=2 Name=Mary test_table.org_dir.foo.`domainname`.
 # List the members of the group admins
 % nisgrpadm -l admins.foo.`domainname`.
 # List the contents of the foo directory
 % nisls foo.`domainname`.
 # List the contents of the test_table along with its header
 % niscat -h test_table.org_dir.foo.`domainname`.
 # Get the entry from the test_table where id = 1
 % nismatch id=1 test_table.org_dir.foo.`domainname`.
 # Delete all we created.
 # First, delete yourself from the admins group
 % nisgrpadm -r admins.foo.`domainname`. `nisdefaults -p`
 # Delete the admins group
 % nisgrpadm -d admins.foo.`domainname`.
 # Delete all the entries from the test_table
 % nistbladm -r “[],test_table.org_dir.foo.`domainname`.”
 # Delete the test_table itself.
 % nistbladm -d test_table.org_dir.foo.`domainname`.
 # Delete all three directories that we created
 % nisrmdir groups_dir.foo.`domainname`.
 % nisrmdir org_dir.foo.`domainname`.
 % nisrmdir foo.`domainname`.