Method and apparatus for providing a network configuration database

Abstract

A network database. The network database is arranged in a plurality of domains in a logical hierarchy. Each domain of the hierarchy represents a body of information associated with a logically related group of users or related group of computers. A relative naming scheme is implemented in which a domain stores the names of only its parent domain and child domains. This permits reconfiguration of the network to be accomplished without changing the database structure. Each domain stores information in a hierarchical structure known as a "directory." Each directory consists of a list of zero or more "properties," each having an associated name and ordered list of values.

Claims

I claim:

1. A method of storing configuration information for a network of computers, said method comprising the steps of:

on at least one computer of said network, executing operations for creating a domain hierarchy consisting of a plurality of domains, said domain hierarchy representing a logical organization of said network, each domain associated with said computers on said network, each domain including a name of each domain logically located above and logically located below said each domain in said domain hierarchy;

said operations creating a directory tree within each of said domains, each said directory tree including one or more directories for storing information associated with each domain and information associated with resources and services of said computers available to each domain; and,

said operations storing each directory tree on storage media of said computers within the domain associated with said directory tree.

2. The method of claim 1 wherein each directory stores information in an ordered list of properties, where each property consists of a property name and an ordered list of values associated with said property name.

3. The method of claim 1 wherein copies of each said directory tree are stored on said computers in said domain associated with said directory tree.

4. A method for storing logical configuration information in a network database, said method comprising the steps of:

executing on at least one computer of said network operations for creating an ordered list of properties of a directory of a directory tree of a domain associated with said computers of said network, each property of said list of properties containing a name and an ordered list of values, said each property defining a subset of said logical configuration information;

said operations defining a name for each property;

said operations defining an ordered list of values associated with each said name, said values representing said subset of said logical configuration information associated with said property; and,

said operations storing said property list, names and values on a storage media of said computers.

5. A method of storing configuration information for a network of computers, said method comprising the steps of:

executing on at least one computer of said network operations for creating a domain hierarchy of one or more domains, said logical domain hierarchy representing a logical organization of said network, each domain associated with one or more computers on said network, each domain including a name of each domain logically located above and logically located below said domain in said domain hierarchy;

said operations creating a directory tree within each of said domains, each said directory tree including one or more directories for storing information associated with each domain and resources and information associated with resources and services of said computers available to each domain, each directory storing information in an ordered list of properties, where each property consists of a property name and an ordered list of values associated with said property name;

restricting access by users to values associated with a first property by creating a second property associated with said first property, said second property containing an ordered list of values that identify permissible access to said first property's values;

said operations storing each directory tree on a storage media of said computers within the domain associated with said directory tree;

said operations comparing a user's identification with said second property's values when said user attempts to access said first property's values; and

said operations denying access by said user to said first property's values when said second property's values do not include said user's identification.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to the field of networks and in particular to maintaining configuration information in a collection of databases.

2. Background Art

In modern computing environments, it is commonplace to employ multiple computers or workstations linked together in a network to communicate between, and share data with, network users. A network also may include resources, such as printers, modems, file servers, etc., and may also include services, such as electronic mail. Information about the computers on the network, and the users, resources and services available to those computers, is referred to as "configuration information."

In some prior art systems, configuration information is typically stored in flat ASCII files. A disadvantage of storing configuration information in such files occurs when users seek access to configuration information to identify available resources and there is typically more information available in that file than the user requires. As a result, the user must search through unneeded information to find the information desired. It may be desired to define different levels and amounts of information associated with different levels of users.

One disadvantage of many prior art network systems is that only one level of user is defined. In other words, each user or computer is considered "equal" in the network hierarchy. This limits the ability to define different levels of information Another drawback of prior art network systems is that they lack flexible directory schemes. Only one level of directories typically can be defined, and directories are typically "read only." Writing to directories is possible only in a few special circumstances. Another disadvantage of some prior art systems is that if backup copies of the network database are stored on different computers, the only method of updating the database is to transfer complete copies of the master database to the backup systems.

Other prior art network databases also typically have highly restrictive directories. The ISO Directory Service (X.500) is one such prior art network system. The directories require a schema to describe the structure of the directory. There is typically no mechanism for creating new schemas, so creating new directories is not possible. In addition, there is no replication capability to accomplish network database backup.

Some prior art network systems support multiple levels of hierarchy such that different user levels may be defined. However, in such systems, the number of levels is limited, and all levels must be used. For many networks, hierarchical flexibility is highly desirable.

Another disadvantage of prior art network systems is their general lack of flexibility in the ability to make desired changes to the structure of the network. For example, when new computers or users are added, moved or deleted, a network administrator is often required to implement the changes.

Therefore, one object of this invention to provide a network system with a plurality of user hierarchies.

Another object of this invention is to provide a network system that has flexible directories which permit reading and writing of properties and permits the creation of new directories.

It is yet another object of this invention to provide a network system that allows network operations to be performed from any location on the network.

Another object of this invention is to provide a network system that allows for incremental replication of the network database.

Another further object of this invention is to provide a network system that permits flexible reorganization of the network structure.

It is another object of this invention to provide a network system that allows security levels to be easily defined.

Other objects and advantages of the present invention will become apparent upon reading the specification and drawings, in which like reference numerals refer to like parts throughout.

SUMMARY OF THE INVENTION

This invention is directed to a network database system for storing and sharing information on a network. The information may be network configuration data, such as user names, user ID's, computer addresses, passwords, personal data, user preference default settings, printers, services, etc. The network database information can also include any data that is to be accessible to a number of network users at the same time.

The network database system of this invention is arranged in a hierarchy of "domains," each of which includes certain information. While a domain can store virtually any information, it contains information about other domains, and typically contains information about the users associated with the computers and any groups the users may belong to, along with resources and services available to that group. A domain serves one or more computers of network.

At the top of the domain hierarchy is the root domain. There may be beneath the root domain subdomains that contain zero or more domains. The information available to each domain is organized into a hierarchy of directories. A directory might contain a list of users, for a list of groups of users. Each directory stores information as ordered lists of "properties". Each property consists of a "name" to identify the property and an ordered list of values associated with the name.

This invention also uses a method referred to as "relative naming." Relative naming refers to the fact that each domain in the domain hierarchy is named relative to its parent domain. A parent domain is the domain logically above a domain in the hierarchy. A domain need not store information about any other domains other than its parent domain and children domains. This allows restructuring of, additions to, or deletions from the domain hierarchy to be implemented by updating only those domains affected by the change.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A is a hierarchical view of an organization.

FIG. 1B is a view of a network configuration of the organization of FIG. 1A.

FIG. 2 is a view of the domain hierarchy of the network of FIG. 1B.

FIG. 3 is a view of the domain hierarchy of FIG. 2 and associated directories.

FIG. 4 is a view of the domain hierarchy of FIG. 3 illustrating one possible computer assignment.

FIG. 5A is a detailed view of a directory.

FIG. 5B is a view of directory ID's.

FIG. 6 is a flow diagram of a directory modification.

FIG. 7A is a flow diagram of a master/clone update cycle.

FIG. 7B is a flow diagram illustrating a clone update cycle.

FIG. 7C is a flow diagram of a master server check sum cycle.

FIG. 7D is a flow diagram illustrating a clone check sum cycle.

FIG. 8A is a block diagram of two computers coupled to a network link.

FIG. 8B is a block diagram of the domain assignment of the computers of FIG. 8A.

FIG. 9A is a diagram of the directory of domain L of FIG. 8B.

FIG. 9B is a diagram of the directory of domain J of FIG. 8B.

FIG. 9C is a diagram of the directory of domain K of FIG. 8B.

FIG. 10A is a block diagram of an example domain hierarchy.

FIG. 10B illustrates the domain of FIG. 10A after a splitting operation.

DETAILED DESCRIPTION OF THIS INVENTION

A method and apparatus for providing a network database is described. In the following description, numerous specific details, such as number of nodes, command formats, etc., are set forth in order to provide a more thorough description of the invention. It will be apparent, however, to one skilled in the art, that this invention may be practiced without these specific details. In other instances, well known features have not been described in detail so as not to obscure the invention.

Computers, workstations and other devices are often linked together in a network to enable multiple users to communicate and to share information. Configuration information for the network, that is, information relating to the computers, users, resources and services available to the network, is stored in a network database. However, not all of the information stored in the network database is of interest to all users. For example, a network may consists of computers in remote, physical locations. Suppose that printers are resources that are shared on this network and that a first printer is located in a first building and a second printer is located in a second building. The users in the first building typically have no interest in the status of the second printer because it is unlikely that those users will use that second printer. Rather, the users in the first building are typically interested only in status information concerning the first printer. Therefore, the network database of this invention is configured to create subsets of configuration information associated with logically related groups of computers and/or groups of users.

To organize network information and the network itself into a system that reflects the organization of the network users, the network database of this invention has a hierarchical structure. A typical hierarchical chart is used here to help illustrate the present invention. By way of example, an enterprise named "company" is organized into a number of divisions and departments as illustrated in the organization tree of FIG. 1A. At the top of the tree is the company itself. At the next successive level, the company is defined as having three divisions; "marketing," "engineering" and "corporate." The marketing division is defined as having two departments; "print" and "TV". The engineering division is also defined as having two departments; "research" and "development." The "corporate" division is defined as having no departments in this example.

A network database of the information to be shared among the divisions and departments of the company can be structured analogously to this tree structure of the company of FIG. 1A. The information to be stored in the network database includes administrative information, such as a user's name, network address, the name of the user's computer and the user's relationship, if any, to other users in the organization. Non-administrative information may also be stored in the network database of this information. For example, telephone numbers and addresses of users may be included, "log in" procedures, default user preference settings, etc., may also be included.

As more fully described below, this network database information is distributed among a plurality of domains. Each domain includes network database information that relates to, and is used by, a particular group of users on the network.

A hierarchical tree is defined as in FIG. 1B representing the collection of network information groups. For example, node A corresponds to the collection of information that relates to, and used by, the entire company. Node B represents the information relating to the marketing group, node C corresponds to engineering and node D to corporate. The marketing node B has two departments, print, (node E) and TV, (node F). The engineering division, node C, has two departments--research, (node G) and development, (node H). Each node at a logical level above another node is known as a "parent" node with respect to the lower node. Each lower node is considered to be a "child" node with respect to the higher node. Certain nodes may be both parent and child. For example, the marketing node B is a child of A but a parent of E and F.

Domain Hierarchy

Each collection of information represented by the structure of FIG. 1A is referred to as a "domain" in this invention. Here, a domain consists of one or more network computers and the information associated with the computer's configuration. As shown in FIG. 2, eight domains are defined in the structure of the example company. For purposes of example, each domain is identified by its associated letter A through H. In the present invention, each domain contains the identification of any child domains associated with that particular domain. For example, domain A, the root domain, contains identification information associated with marketing, engineering and corporate domains. Similarly, domain B (the marketing domain) contains the names of domains E and F ("print" and "TV"), which are the children domains of B. Domain C, the engineering domain, contains the identification information associated with domains G and H ("research" and "development"), which are the children domains of C. Domains E, F, G and H do not contain any identification information because these domains do not have any associated children domains. Similarly, domain D ("corporate") does not contain any identification information. The domain hierarchy of this invention may consist of only one domain, or it may consist of a plurality of domains having any number of levels of associated branches and children domains.

Relative Naming

This invention uses a technique referred to herein as "relative naming" of domain hierarchies. Each domain only contains information about the domains immediately above it in the hierarchy or immediately below it, (if any exist). If this convention is maintained consistently throughout the network domain hierarchy, a communication path can be created between any two domains in the hierarchy.

For example, a communication path can be created between domain C and domain E as shown in FIG. 2. Domain C only contains the names of its two children domains ("research" and development") G and H, and information indicating that domain A is its parent domain. Domain C contains no information about domain E or about any other domain on the network. Domain E only contains information indicating that domain B is its parent. Similarly, domain E contains no information about domain C or about any other domain. To establish a communication path between domain C and domain E, domain C can access domain A to achieve such communication. Domain A, storing information about its children domains, can then access domain B, (marketing). Domain B, storing information about its parent (domain A) and its children, (domains E and F) can access domain E and thus a link between domains C and E is created.

By using the relative naming scheme of the present invention, information about the absolute position of a domain in the domain hierarchy is unnecessary.

Directories

Within each domain of the domain hierarchy, information is contained in a hierarchical structure known as a "directory tree." A directory tree is represented here by an inverted tree having its root at the top and branches extending downward. The directory at the top of the tree is called the root directory. The root directory branches into several other directories, called its "subdirectories" (because of their relation to the root directory). Each subdirectory may also have its own associated subdirectories. A path from the root directory to a subdirectory is referred to as a "branch" of the tree. For convention a directory is "above" its associated subdirectories. A directory is a parent of its associated subdirectories. Conversely, a directory is a subdirectory, or child, of its associated parent directory.

Referring to FIG. 3, each domain of the domain hierarchy has an associated directory tree (e.g., A, B and C directory trees of Domains A, B and C, respectively). A directory tree is shown in greater detail in FIG. 5A. It has two "subdirectories," Y and X. Y in turn, has two other subdirectories, W and V. Directory X has no subdirectories.

Each directory has a numeric object identification ("ID") that is unique for the domain containing such directory. For example, the directories shown in FIG. 5A may have ID's as listed in FIG. 5B. Directory Z is identified by the ID 0; directory Y by the ID 1; directory X by the ID 7; and directories W and V by the ID's 10 and 44, respectively. Each numeric ID within a domain is unique. However, the same numeric ID may be used in different domains. If a directory is moved to another domain containing a directory with the same ID, the ID of the directory being moved must be modified to maintain uniqueness.

Directories may be arbitrarily created and assigned a unique ID. Similarly, directories may be arbitrarily destroyed if they do not have subdirectories. That is, the lowest directory of a branch may be arbitrarily removed. But directories having associated children cannot be removed unless the children directories are removed first.

Each directory also includes an ordered list of zero or more "properties." Each property has a "name" to identify the property and an ordered list of 0 or more "values." A property defines a relationship between a name and its associated values. The name of a property is comprised of alphanumeric characters, as is each value. An example of a property list is as follows:

    ______________________________________
    Property 1:
           name:      "User Name"
           value 1:   "Penny Smith"
    Property 2:
           name:      "Phone Numbers"
           value 1:   "555-1775"
           value 2:   "555-1675"
    Property 3:
           value 1:   "Full Time Employee"
           (no values)
    ______________________________________

    ______________________________________
    Property 1:
           name:      "User Name"
           value 1:   "Penny Smith"
    Property 2:
           value 1:   "Full Time Employee"
           (no values)
    ______________________________________


______________________________________
    Name ".sub.-- writers"
           Value 1    "Ellen"
           Value 1    "Frank"
    ______________________________________

    ______________________________________
    Name ".sub.-- writers"
           Value 1    "Ellen"
           Value 2    "Frank"
    ______________________________________

    ______________________________________
    Property 1:
           name:      "name"
           value 1:   "Ellen"
    Property 2:
           name:      "password"
           value 1:   "topsecret"
    Property 3:
           name:      ".sub.-- writers.sub.-- password"
           Value 1:   "Ellen"
    ______________________________________

    ______________________________________
    Property 1:
           name:      "name"
           value 1:   "Ellen"
    Property 2:
           name:      "password"
           value 1:   "topsecret"
    Property 3:
           name:      ".sub.-- writers.sub.-- password"
           Value 1:   "*"
    ______________________________________

    ______________________________________
           "Name"
           Value 1    "machines"
    ______________________________________

    ______________________________________
    Property 1:
           name:      "name"
           value 1:   "machine-2"
    Property 2:
           name:      "ip.sub.-- address"
           value 1:   "129.18.2.60"
    Property 3:
           name:      "serves"
           Value 1:   "Marketing/B"
           Value 2:   "./A"
    ______________________________________

    ______________________________________
    Property 1:
           name:      "name"
           value 1:   "machine-2"
    Property 2:
           name:      "ip.sub.-- address"
           value 1:   "129.18.2.60"
    Property 3:
           name:      "serves"
           Value 1:   "Frank/F"
           Value 2:   "./B"
           Value 3:   "../A"
    ______________________________________

    ______________________________________
           Name       "master"
           Value 1:   "machine-2/A"
    ______________________________________

• Inventors
  Taylor, Bradley;
• Assignee
  Next Computer, Inc. (Redwood City, CA)
• Published
  Apr-25-1995
• Current US Classes:
  340/825.02
  707/100
• Application #
  174213
• International Classes
  G06F 015/16; G06F 012/00
• Field of Search
  395/600 395/425 395/650
• Examiner
  Black; Thomas G.
• Agent
  Hecker & Harriman
• US Patent References:
  4731734
  4827411
  5055999