Table Of Contents

1 About this document.

1.1 History of the document.

2 Aim of the project.

3 Decisions for the overall system.

4 Overall system Architecture

4.1 Central Application: signal handler.

4.2 Data processing.

4.3 Web interface

4.4 Priority mechanism.

4.5 The dataflow diagram.

4.6 Sending messages to the central gnucomo system.

4.7 Getting data into the database.

5 The database.

5.1 Database conventions.

5.2 Database design.

5.3 Actual design.

6 Warnings that can be detected.

7 User Interface.

8 The installation process.

8.1 Supported platforms.

8.2 Installation on the server.

8.3 Installation on a UNIX-client.

9 Dependency on other free software.

9.1 Related projects

10 Settings on the server machine.

10.1 Required.

10.2 Suggested.

11 Settings on the client machine.

11.1 Required.

11.2 Suggested.

12 Appendices.

12.1 Appendix A. GNU Public License Version 2, June 1991.

1 About this document.

This document describes the technical specifications for the Gnucomo project. It will describe the functionality achieved, design specifications and choices made. The document will be the manifest for the developers to work in the same direction and not run into unneeded disappointments.

1.1 History of the document.

2 Aim of the project.

The number of log-files in a system and the tools in general do not make monitoring a system simple. Quite often there is so much information that attention for logfiles seems to fade away. For that reason monitoring routers and clients is often not an option. Instead of timely detection of problems, logs more often are used to find out what went wrong. When a system is under attack early signals can often easily be detected¹ and preventive measures could have been taken if the time was only available.

The Gnucomo project is meant to make pro-active monitoring of computers and devices easier. It will contain a set of applications that will retrieve all types of monitoring information from devices and place it into a database. Devices can be server computers, desktop computers, PBX'es or other systems. By running checks the manual process of watching log files can be reduced. Also an intelligent script can see things that a human being will easy overlook. Gnucomo won't relieve an administrator of all manual work on log-files, but will increase the changes on actual monitoring.

With all this data being available at one central location the gnucomo-server acts as some kind of black box for computer systems. If something does happen the evidence will be available at a remote location available for exploration². Forensics will be made easier. Also the collected data is available from multiple locations make it easier to found more about an attack. The change that you overlook things will be reduced.

Based on the entered data all sorts of analysis will be performed to discover abnormalities a normal maintenance tool or IDS wouldn't be looking at. An example could be a website in Dutch that suddenly obtains a lot of attention from Greek visitors (based on the location of their IP³). These abnormalities will be presented as user-friendly as possible to increase awareness of the state of the system. By doing that the system can also be used for more complex analysis. One than can look at long-term trends like hacking attempts that take place, new exploits that are tried all of a sudden or signs of distributed attacks or a certain pattern of attempts⁴.

Also extra data will be gathered where ever possible to save time to the administrator. If IP address attract attention of gnucomo the next logical step would be to use tools like dig, whois and visiting arin-related websites. This data will automatically be collected in an early stage and stored in the database.

The results will be gathered in a warning system. Those warnings will be presented to the administrative person who is responsible for that particular machine or network. Having multiple systems in the system can add to the intelligence that can be gathered. The interface will be web-based and aimed at user-friendliness.

Since multiple systems can enter data into the Gnucomo database more intelligent hardware and security detection can be done. When things do go wrong Gnucomo contains as much information as possible to figure out what happened and assist in the forensics. Since research on the signals afterwards is broader more attention will be given as much data as possible and do less filtering.

With the data in the database also policies can be checked automatically retro-actively so that the security leaves some room to stretch some rules. This may sound not logic, because one of the main functions of security is to enforce rules. But some rules are made with the different meanings. For instance: a rule may be that private browsing isn't allowed. The background of such a rule might be to try and reduce the amount of data traffic (too many people downloading mp3's, or people not getting to work anymore). By monitoring that specific fact (bandwidth spent on non-business related Internet use) it's possible to be relaxed on that rule and enable many good-willing users reading his/hers daily newspaper.

The scope of the project is clearly limited to monitoring and not to offer automated maintenance or web-based maintenance. There are other projects currently being able to provide that functionality. We will focus on intelligence and user-friendliness of representation of facts presented. A warning should trigger an administrator to get up and do something using the tools he or she values the most.

Also we are not aiming to replace great tools like SNORT as a real-time IDS. These tools can do thing, that in the beginning won't be a part of gnucomo out of performance reasons. Also there is no need to duplicate what they already did. If that energy is placed in intelligent we're very complimentary. Gnucomo can however be a reality check on an existing NIDS (Network Intrusion Detection System). For instance if warnings keep coming it may be time to rethink the rules that have been set.

This evolves to the following list of functions gnucomo can provide:

• Intrusion detection
• Detecting hacker attempts
• Early detection of system failures
• Exhaustion of system resources
• Capacity planning for future expansion
• Spotting bottlenecks in a system.
• Verifying system integrity
• Assistance with troubleshooting
• Perform post-mortem forensics
• Incident response system.

3 Decisions for the overall system.

In order to get the project running we have to make some decisions before we can start. Of course are the decisions always open for review, but initially our main aim is to get a system running. This doesn't mean that we allow a lesser architecture, but more that we create an environment that will lead to results.

The following decisions apply to the system in general:

• The major part of the system will be used for security features. The solution itself has to be secure.
• In general we strive to be as platform and application independent as possible. However to achieve software to get ready and to assure progress some selections will have to be made excluding options for applications⁵. Where possible we will enable ports to other applications and platforms.
• The computers collecting the data will only be Linux machines, with future support for other Unices. However it is not aim our to port this part of the project to other platforms just now.
• The database system will make heavy use of stored procedures and triggers and thus lock out some less feature-full database.
• Although we're not married to a database system initially this will be PostgreSQL only so that we can build for results. The code however shouldn't be build to deliberately lock outer systems. Other database that might be interesting to add: DB2, Oracle (if the rainfall of security advisories seems to be over) and Informix.
• The interface for the user will be a web-interface written in PHP with PostgreSQL-database access. Despite the fact that initially this will be the preferred way to communicate with our system, other interfaces are welcome and should be supported.
• The technologies used for the daemons on the central site and remote site are open for discussion. A decision is made when it is clear what exactly will be needed to code.
• The original logfiles on systems will not be harmed in any way, but will be saved the way they presented to gnucomo.
• In the database processed data as well as raw data will be stored.
• Our main aim is to ease the life of the administrator when dealing with the symptoms of a machine so that he doesn't miss important notifications. The project will be made around detection. We won't focus on making a maintenance applications but solely on monitoring, because other applications like webmin or linuxconf do already deliver such functionality. We are complementary to those applications. This doesn't mean that the system cannot send SMS'es, e-mails or make alarms sound.
• Any output of the logbook will be stored and sent within a certain interval which can be set. By doing that the mechanism will not flood the total mechanism and overload machines.
• When applications for gnucomo we will try and do the non-server part as platform independent as possible. This in order to prevent writing a same application multiple times. For instance one for Unices, one for Macintosh and one for Microsoft.
• To ensure the quality of software and to prevent any unwanted functionality rolling into the project all code will be reviewed by one of the lead members before it is accepted. Since this project is aimed at throughput speed no formal procedures will be enforced, but we promise not to let software in that hasn't been checked.

4 Overall system Architecture

The overall systems aims to make maintenance data better accessible and by doing that lowering the barrier to be on alert for intrusions, system failure and other misery that can happen to computer systems. Also more systems can be monitored and even focus can be placed on desktop computers, something that nowadays rarely happens. Since the solution is only aimed at monitoring (with some responses possible) other people can watch their system(s) without being Administrator⁶.

The main system will know to sides:

• Central Application - server
• Monitored System - client

The project has been setup as a two sided system in order to be able to guard many computers at the same time. However it may be obvious that both sides of the application can very well be installed on a single system.

To monitor a system, Gnucomo uses two kinds of input: event and parameters. Events occur on a system while it is running and reflect the transient behaviour of the system. Parameters reflect the current state of the system. The most obvious way to gather events from a monitored system is to read the system log files. Examples of events are IP packets that are rejected by the firewall or clients that access the http daemon. Parameters are obtained for example by reading configuration files or kernel data structures. Examples of parameters are the size and free space of a filesystem or the users that are listed in the password file. Both kinds of input are obtained actively or passively, i.e. by installing probe agents in the system which regularly aquire the system's parameters or passively by sending the output of programs to the Gnucomo server.

When signals⁷ arrive they will be stored in a file. When this file is delivered to a certain directory a daemon will detect this and start the transfer of the file. The file will be transferred to the central application or the client. This transfer will be triggered immediately after a process has finished or with a certain time-interval when it concerns a logbook. All output will be placed in a directory where the daemon detects it and ensures the transport. For transport currently only two mechanisms will be supported:

• Encrypted file copy relying on the SSH protocol. On Unix-based systems this will scp (secure copy).
• E-Mail. The e-mail is encrypted and signed using gpg and then sent to the server. Since the file format will identify the type-of-output the Subject-field of the e-mail is not really needed.

4.1 Central Application: signal handler.

Illustration 1 Basic overview of the processes on the server.

4.2 Data processing.

The data processing has four tasks:

• Extracting data from e-mail and store it in the input-buffer. This can be done by a daemon that checks the e-mail⁹ with a certain interval extracts the e-mails and leave the content in a file in the input buffer. This daemon has only rights to write to the directory it has to write to. We may as well have the email captured directly by a program with a "gnucomo: |/usr/local/bin/gnucomo-input" - like alias.
• Detect files in the input buffer decrypt the content and verify the signature. Another daemon will see the log-files and starts checking if the origin is correct (by verifying the signature) and decrypting the content. The legible files will be processed by entering the data into the database.
• The database will accept the data and perform a certain number of checks as the data comes in. During the processing of the data abnormalities will be detected and entered into a notification table. The database system will also carry out more complex tasks on given moments in time and enter them as well in the notification table.
• The database system must be able to undertake action when high alerts are being entered into the database. This can be a couple of things to begin with: send an e-mail or SMS. In a later stage other technologies may be added as well. In this scheme we also make a possibility to escalate problems when no action is taken in a certain amount of time.

4.2.1 System Parameters

Gnucomo maintains the operational parameters of a monitored system for a number of reasons. The most important reason is to create notifications when somthing about a parameter changes while the parameter is not supposed to change. Such a change may be intended by the system administrator, e.g. when a package is upgraded, or there may be something wrong. In any case, you will want to know about a change in your system when it happens. Furthermore, a change history of a parameter's values will come in handy when you want to look back in time and figure out what happened in the past. Another usefull application of parameters concerns the maintenance of a large number of similar systems. When the parameters of each system are reported regularly to Gnucomo, deviations from the 'standard' system configuration can be easily spotted.

Some properties of parameters are supposed to change regularly. A changed value of such a property will of course not lead to any notification. On the other hand, the change history of these parameters may provide interesting information about the monitored system. This leads to the distiction between static and dynamic properties of parameters. The difference between dynamic and static properties manifests itself mainly in the change history of the parameter's property. Dynamic properties typically have a change record once a day or even a couple of times a day. Change records for static properties are usually months apart. If all properties of a parameter are dynamic, the parameter as a whole is regarded as a dynamic parameter. One of the properties of a dynamic parameter is that it does not need to exist all the time, i.e. it does not have to be listed in every report.

The state of parameters is scanned or probed regularly on a client system and reported to the Gnucomo server. These reports can be created in a variety of ways. For example, filesystems are reported with 'df', installled packages with 'rpm -qa', users by reading /etc/passwd, etc. Many other probing methods may be implemented. Each report from a probe holds the current value of several parameters. Gnucomo will check each property of these parameters against the stored knwon value. If the property's value changed, the actual value in the database is updated and a record is added to the change history of the parameter. When a parameter is listed in the report but that parameter is not in the database or the other way around: a parameter is in the database and is not in the report, this constitutes a notable change in the system.

Whenever the state of the system's parameters changes, Gnucomo may create a notification. Notifications are created when one of the following changes happens:

1. A new parameter is detected. The new parameter is listed in the report but is not in the database.
2. The value of a static property of a parameter has changed.
3. A value of a dynamic property of a parameter goes outside its designated range.
4. A non-dynamic parameter has disappeared. The parameter is in the database but was not listed in the report.

4.3 Web interface

The web interface will used to interact with the user. The interface should be intuitive and easy to understand. More important warnings should directly draw attention. The user must be able to perform settings so that warnings can be rated differently than the original settings.

The interface will do the following things:

• Show a list of warnings that are currently open.
• It must be possible to sort the list on all the fields shown.
• Deliver detailed information (logbook entries) upon request that have led to the warning.
• Undertake certain actions like sending abuse@internet-provider.net e-mails with information.
• Monitor actions on outstanding issues.

4.4 Priority mechanism.

Each notification has a certain priority that requires a different handling of the issue. How each priority will be dealt with is something that can be set per server. The priority mechanism is a simple system of five categories (can be more or less).

4.5 The dataflow diagram.

The main dataflow will be as follows.

4.6 Sending messages to the central gnucomo system.

One of the main tasks is getting all the messages to the database. Ultimately gnucomo will support multiple ways of receiving the data. Basically anything goes, but two mechanisms will be supported in the project to begin with:

1. E-mail. Messages and elements from logfiles will be sent through e-mail.
2. File copy. Using technologies like scp or ftp (not preferred due to the insecure nature) can place files directly in the receiving directory.

4.6.1 Ensuring data integrity.

<TO BE DESCRIBED>

4.6.2 Directories and filenames on the server.

Files will be dealt with as if gnucomo were a user (actually there will be a user gnucomo). The files will be placed in the /home/gnucomo/ directory. Only /home/gnucomo/incoming/dropbox/ can be used to save data in from external systems. All other directories are only available for the user gnucomo. The filename will represent the data that is received. The details are seperated with underscores. When data is sent by e-mail the filename will be written in the first line of the e-mail. A typical filename looks like this:

3_messages_20020807235208_1.asc

the logic behind this is following:

urgency_typeofmessage_timestamp_objectid.typeoffile.

4.6.2.1 Directory for incoming data.

For incoming messages there will be separate directories (/home/gnucomo/incoming/)for:

1. Dropbox. This directory is ready to receive data from all sorts of systems. This directory is world writeable (but not deleteable!): /home/gnucomo/incoming/dropbox/
2. Encrypted messages. In this directory messages will be placed that are still gpg-crypted. In this directory the files await decryption and verification of the signature. This directory is: /home/gnucomo/incoming/crypted/
3. Decrypted with errors. During the decryption exercise anything can go wrong. Decrypting can fail or the signature may show errors. If this happens the original message is moved to a different directory: /home/gnucomo/incoming/cryptfail/
4. Inbox. After successful decryption the decrypted message is moved to the inbox. Data that is not encrypted can be moved here from the dropbox after certain verifications. The directory is: /home/gnucomo/incoming/inbox/
5. Processed. After processing the data the file is stored in the directory processed for further reference for a certain amount of time. The directory is: /home/gnucomo/incoming/processed/
6. Archive. After a certain period the files will be archived. Since not every system will archive everything this directory may also be a symbolic link to /dev/null. The used directory is: /home/gnucomo/archive/

4.6.2.2 Directory for outgoing data.

For outgoing messages the directory /home/gnucomo/outgoing will be used. This directory knows a couple of sub directories:

1. Dropbox. This directory is used by the central gnucomo system to place outgoing messages in. The used directory is: /home/gnucomo/outgoing/dropbox/
2. Outbox. After processing the message (including signing and encrypting when applicable) the messages are placed in the outbox: /home/gnucomo/outgoing/outbox/
3. Processed. After processing the data the file is stored in the directory processed for further reference for a certain amount of time. After sending a message a confirmation will be made that is saved as an incoming message. The directory is: /home/gnucomo/outgoing/processed/
4. Archive. After a certain period the files will be archived. Since not every system will archive everything this directory may also be a symbolic link to /dev/null. The used directory is: /home/gnucomo/archive/

4.6.2.3 Overview.

The total directory-structure looks like this:

    /home/gnucomo/
    /home/gnucomo/archive
    /home/gnucomo/incoming
    /home/gnucomo/incoming/crypted
    /home/gnucomo/incoming/cryptfail
    /home/gnucomo/incoming/dropbox
    /home/gnucomo/incoming/inbox
    /home/gnucomo/incoming/processed
    /home/gnucomo/outgoing
    /home/gnucomo/outgoing/dropbox
    /home/gnucomo/outgoing/outbox
    /home/gnucomo/outgoing/processed

4.6.3 Directories on the client-side.

On the client-side the files that need to be transmitted will be placed in a directory system as well. In the future this system may not be in use at all devices (routers, certain MS Windows-machine, IP Telephones, etc.). For those systems a different mechanism will become be described here. Initially we focus on Linux systems that will enter data into the database.

The filename convention will be totally identical to the filename convention on the server, since this the same mechanism.

To facilitate gnucomo client and server on one and the same machine the gnucomo-client should have a different default user. For this purpose the user gcm_client will be created.

4.6.3.1 Directory for incoming data.

For incoming messages there will be separate directories (/home/gnucomo/incoming/)for:

1. Dropbox. This directory is ready to receive data from all sorts of systems. This directory is world writeable (but not deleteable!): /home/gcm_client/incoming/dropbox/
2. Encrypted messages. In this directory messages will be placed that are still gpg-crypted. In this directory the files await decryption and verification of the signature. This directory is: /home/gcm_client/incoming/crypted/
3. Decrypted with errors. During the decryption exercise anything can go wrong. Decrypting can fail or the signature may show errors. If this happens the original message is moved to a different directory: /home/gcm_client/incoming/cryptfail/
4. Inbox. After successful decryption the decrypted message is moved to the inbox. Data that is not encrypted can be moved here from the dropbox after certain verifications. The directory is: /home/gcm_client/incoming/inbox/
5. Processed. After processing the data the file is stored in the directory processed for further reference for a certain amount of time. The directory is: /home/gcm_client/incoming/processed/

4.6.3.2 Directory for outgoing data.

For outgoing messages the directory /home/gcm_client/outgoing will be used. This directory knows a couple of sub directories:

1. Dropbox. This directory is used by the central gnucomo system to place outgoing messages in. The used directory is: /home/gcm_client/outgoing/dropbox/
2. Outbox. After processing the message (including signing and encrypting when applicable) the messages are placed in the outbox: /home/gcm_client/outgoing/outbox/
3. Processed. After processing the data the file is stored in the directory processed for further reference for a certain amount of time. After sending a message a confirmation will be made that is saved as an incoming message. The directory is: /home/gcm_client/outgoing/processed/

4.6.3.3 Overview.

The total directory-structure looks like this:

    /home/gcm_client/
    /home/gcm_client/archive
    /home/gcm_client /incoming
    /home/gcm_client/incoming/crypted
    /home/gcm_client/incoming/cryptfail
    /home/gcm_client/incoming/dropbox
    /home/gcm_client/incoming/inbox
    /home/gcm_client/incoming/processed
    /home/gcm_client/outgoing
    /home/gcm_client/outgoing/dropbox
    /home/gcm_client/outgoing/outbox
    /home/gcm_client/outgoing/processed

4.7 Getting data into the database.

The files in the /home/gnucomo/incoming/inbox/ should be stored in the database. For this purpose there is a table unprocessed_log. The data of the filename as well as the content of the file need to be placed in one record.

There are some fields that have to be addressed immediately:

1. Servicecode: The servicecode will be obtained out of the filename the is the type_of_message-field in the filename.
2. Objectid: This data is given in the filename it can be found in the objectid-part of the filename.
3. Logdata: The data is saved as follows: filename (unaltered) <CR>textual data of the file.

The daemon application that delivers the data is called gcm-input. It performs the following steps with no extra functionality:

1. Detect if a file is available.

2. Write the data in the database.

To write the data in the database a database user gcm_input exists. This user has only the right to enter data into the database. There are no deletion, update or select-permissions.

5 The database.

The database is the heart of the system. It will contain all event-data of multiple computers. The intelligence that can be performed on the database will be placed there. To do this as integratedly as possible stored procedures and triggers will be used. To begin with we have selected checks to be performed that will be expanded throughout time.

Since the gnucomo database and files contain sensitive data security measures have to be in place. Several database users will exist that have limited rights to perform a certain task ensuring some protection against unauthorized access. However these mechanisms on it's own will work fine, bad maintenance may still screw-up good security. Good database maintenance is needed. For the gnucomo the protection of the valid authentic nature of the data in the database has our highest priority.

A table-name in this chapter is written in cursive writing.

5.1 Database conventions.

The database will be built according the following conventions:

• All names for tables, indexes and fields will be written in lower case and will be singular. Spaces in names are not allowed.
• If a table contains field referring to other tables the mother-tables are mentioned in alphabetical order with an underscore between the table-names. So object and user will make a third table object_user
• A data access user (being the interface) is not allowed to write to the log-entries and the warnings. To change the state of a warning a stored procedure will do so by having a different entry in a table. This should make it possible to discover who did what at which moment in time.
• The name of two related fields that make the relationship between two tables will be the same to avoid confusion.

5.2 Database design.

In the design we anticipate to deliver an as best as possible database performance. That means that data that needs to be entered occasionally can be heavily indexed to increase performance. However data that is mainly stored will only be indexed marginally to have the best possible performance on data entry. If during one of the checks on data-entry a notification is made, the information related to that notification will be indexed very well to increase retrieval performance. What we will try to avoid is that the user interface will cause full table scan and affect the performance of the overall system dramatically. One of the techniques to increase performance on display is to work with views. So where it is feasible we will use them.

The following model pictures the database as described in the remainder of this chapter.

In general the database must also be maintained well. So daily maintenance scripts should keep the performance good¹⁰.

5.3 Actual design.

In this part of the chapter the tables will be explained and then described with all important elements. Per table a sub-chapter will be created. Each table will have a table design, indexes, relationships and the required data (for those tables where the data itself is relevant in the design or sample data (for those cases where no set data is needed). For the relationships beside a description the subset of the total schema has been incorporated in the document so that it is more clear what exactly is meant. Due to the complex nature of the design those drawings sometimes will seem funny.

For the fieldtypes the types of PostgreSQL will be used. These values can be found in Chapter 3 of the PostgreSQL User Manual (http://www.postgresql.org).

For indexes primary keys are always unique and called primary key in the name. Since unique indexes within PostgreSQL only work on B-Tree indexes (which is default) we will use B-Tree for all indexes. In cases where an exception is made the used type of index will be indicated in the characteristics. A footnote will explain why a different type of index has been selected.

5.3.1 action

This table stores the actions that are recognized by the system. Whenever a situation in the database has been raised (through the table notification) one expects steps to be taken to resolve these issues. All these steps are stored in this table. Each step beginning with detection until a case is closed can be traced back through this table. This will support the accountability (who is responsible for which action) and non-repudiation (being able to trace what exactly happened to each notification). In the table action all recognized actions that can be taken are stored. Several of the taken actions will lead to change in statuscode. When this is the case either the server-side of the interfaces or the gcm_daemon will perform this task. This doesn't apply to all actions though.¹¹. Actions take place through background processes and the interface. Each action known by the system will be delivered by the software. This table mainly is used for retrieval so indexing will be done as much as possible.

5.3.1.1 The fields.

The table below describes the fields:

Fieldname	Fieldtype	Size	Remarks
actionid	Bigserial	8	Autonumber bigint (eight bit). Unique identifyer to refer to when this action is taken.
actionname	Text		Short descriptive name for the type of action.
statuscode	Varchar	3	New status that will be given to a notification when this action takes place. If the action occurs the status in the notification can change because of this. The statuscode will indicate what the new status has to be. If this field is left empty no change to the status will occur.
description	Text		A longer description (without limit) on the action.

5.3.1.2 The indexes.

Indices for this table are:

5.3.1.3 The relationships.

Relationships with other tables:

In the model this looks like this:

5.3.1.4 Default data in table.

The data in this table is standard for the system and part of the design. The user cannot change this or add value to it.

5.3.2 Action_user.

This table stores each actual action taken. Where the action basically stores the type of actions that can be taken, this table says 'at this point in time for this notification this action was taken'. Using this table it is traceable who did what at which moment in time. The table is very important for later use if something goes wrong, but is also relevant for the interface. All steps of a in the process of a notification can be traced using this table. There will be a lot of entries here, but retrieval is more crucial for performance than data entry. So indexing on logic fields is very relevant. Processing might be slower, but that's worth the price.

5.3.2.1 The fields.

The table below describes the fields:

5.3.2.2 The indexes.

The table is indexed on the following fields:

5.3.2.3 The relationships.

Relationships with other tables:

In the model this looks like this:

5.3.2.4 Sample data.

Since no data is delivered automatically a couple of sample records are shown here.

5.3.3 history

The history table records all changes to properties of parameters.

5.3.3.1 The fields

The fields of the history table are listed below:

Each time something about a parameter changes, this is recorded in the change history of the paraneter. When such a change happens, one of three things may occur to a parameter, as stated in the 'change_nature' field:

1. A new parameter is created.
2. The value of one of the properties was altered.
3. The parameter is removed.

5.3.4 log & log_adv.

To store the log-data there are two tables in use that have a one-on-one relationship. The logic behind this is the difference between raw-always needed data and the somewhat processed data to support basic retrieval. The last category isn't always used and when it's used it is redundant. Also the raw log is very important to the integrity of the system. For these reasons the processed data has been physically separated in a second table called log_adv. If needed a view will be available that combines the two tables. Despite the load indexing on log_adv will be done thoroughly.

5.3.5 log.

5.3.5.1 The fields of log.

The fields in log are focussed around the raw data arriving from log-files and output from processes that deliver status information. The log table is the one and only place used to automatically deliver data to the gnucomo-system. This makes the facts very traceable and ensures that there is one point where the data isn't yet fragmented (usefull if new insights are usefull). The log table also provides the integrity of the data as it was delivered. The data will be needed to provide a link to other tables in the system that handle processed derrivates of the original data (added intelligence).

5.3.5.2 The indexes.

The indices of the table:

5.3.5.3 The relationships.

Relationships with other tables:

In the model this looks like this:

5.3.5.4 Log_adv

As soon as new data is detected in the log-table processing takes place. If the data is recognized by the gcm_daemon one or more log_adv records will be written where data is separated. Due to the high diversity multiple types of the log_adv table will be created. All of them will be an inheritance of the log_adv-table.

5.3.5.5 The fields of log_adv.

The fields of the table log_adv are shown below:

5.3.5.6 The indexes.

The table is indexed on the following fields:

5.3.5.7 The relationships.

There is only one relation with this table.

5.3.5.8 Sample data combined from log and log_adv.

The sample data derrived here has been gathered in logs. Since the tablestructure is very long the representation is somewhat different:

5.3.6 log_notification.

In the log_notification the logbook entries that have caused an alert to occur are saved. When this table is used something has been detected. As this is clearly an intermediate table we anticipate to design checks where multiple entries in a log-file can lead to one notification. For forensics indexing will be focussed on retrieval speed.

5.3.6.1 The fields.

The fields are listed below:

5.3.6.2 The indexes.

The table is indexed on the following fields:

5.3.6.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.6.4 Sample data.

Some sample data:

5.3.7 notification.

The main task of gnucomo is detection issues. As soon as something has been detected based on the queries run a notification will be created. This notification will be the warning towards the system that something has been detected. Issues are entered based on immediate detection, periodical detection or manually. In an ideal case no notification will occur, but as time goes by issues will occur. When systems function properly more retrieval than data entry will take place. Also data retrieval will be done in all sorts of ways. One can think of immediate display, but also updates due to actions and ultimately reporting. Indexing must be huge to facilitate all these queries. There will be one single point-of-contact with the gnucomo-system. This point-of-contact will be the XML/SOAP interface. By doing that we can set user-rights (as has been impleted in this table) without communicating that to the user. Also the gnucomo-XML/SOAP-interface will verify if the attempted action is allowed.

5.3.7.1 The fields.

The fields of the notification table are:

5.3.7.2 The indexes.

The indices are:

5.3.7.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.7.4 Sample data.

The following data is an example of a notification placed in the database.

5.3.8 object

The object table contains general information on the objects being monitored. Mostly objects will be computers, but it may also be something else in the future like routers, switches, gateways, PDA's or other devices. The table object will more be used for retrieval, since adding objects will be an occasional process. Anything that for some reason can be indexed ought to be indexed.

5.3.8.1 The fields.

The fields are:

5.3.8.2 The indexes.

The object table is indexed on the following fields:

5.3.8.3 The relationships.

The relationships with other tables are listed below:

In the relationshipmodel this looks like this:

5.3.8.4 Sample data.

There is no preset data and therefor it's an example has been created. The table has quite some fields so the example has the fieldname in the left column and the data in the right column.

5.3.9 object_issue.

This table will store the policy on a certain issue like the priority being recognized and special actions to take. Since values in this table are bound to the object, responses to the same incident can lead to different alert levels based on the importance of the object. Since policies are utilized by the systems continuously all other process will rely on this index, while users will change the values occasionally.

5.3.9.1 The fields.

The fields of object_issue:

5.3.9.2 The indexes.

These are the indices:

5.3.9.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.9.4 Sample data.

Some sample data:

5.3.10 object_priority.

By default a prioritycode has a certain meaning. But default behaviour can be used for most cases, but not all. For some objects a deviation would be usefull. For instance a very important webserver under an attack should maybe alarm more agressive than a printer-server that is used only marginally. This table stores per object how a certain level of priority is being dealt with. The main issue is the question: What policies do apply? If nothing is available default behaviour as defined in the table priority will apply. This table is mostly used for retrieval, so firm indexing is logic.

5.3.10.1 The fields.

The fields are listed below:

5.3.10.2 The indexes.

Indices of the object_priority table:

5.3.10.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.11 object_service

The object service table indicates which services should report itself to the Gnucomo-system. If input fails to show up a notification can be generated.

5.3.11.1 The fields.

The fields are listed below:

5.3.11.2 The indexes.

The table is indexed on the following fields:

5.3.11.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.12 object_system_user

Every object knows users either by the security of the object itself or through a central database (like LDAP for instance). Initially a new object can sent the userlist to gnucomo. Any username that isn't in the userlist is potentially dangerous. To avoid any mis-understandings this user is not a gnucomo-user, but a user on a remote system. So during the processing the read will be done more than the data entry (one account will most likely do multiple actions) and that makes heavy indexing logic.

5.3.12.1 The fields.

The fields of object_system_user are listed below:

5.3.12.2 The indexes.

The table is indexed on the following fields:

5.3.12.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.13 object_user

This table will enable users to get access to the information belonging to an object. Also this table is mainly used for data retrieval and will rely on the indexes.

5.3.13.1 The fields.

The fields of object_user are listed below:

5.3.13.2 The indexes.

The indices of the object_user table:

5.3.13.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.14 parameter

The parameter table stores the operational parameters of a monitored object. The parameters of an object describe the object's resources and configurations. For each object, a large set of parameters can be defined. They range from anything like file systems and installed packages to the system's users.

5.3.14.1 The fields

The fields of the parameter table are listed below:

The combination of objectid, name and class must be unique.

5.3.14.2 Sample data

The table below lists a few examples of parameters

5.3.15 parameter_class

Each parameter is defined to be of a certain class. The class defines which properties a parameter of that class may have.

5.3.15.1 The fields

The fields of the parameter_class table are listed below:

The combination of name and property_name must be unique. Note that the min and max fields are used only for properties of a numerical nature.

5.3.15.2 Sample data

The table below lists a few examples of parameter classes

5.3.16 parameter_notification

The parameter_notification table defines the relationship between parameters and notifications. Whenever a parameter is changed, i.e. the parameter is created, one of its properties changed or a parameter is removed, this may result in a notification. This table provides the link between that notification and the change in parameter that the notification is about. Note that a single notification may be created for a number of changes in parameters.

5.3.16.1 The fields

The fields of the parameter_notification table are listed below:

The combination of notificationid and paramid must be unique.

5.3.17 priority

The priority table contains information on the levels that are recognized by the system. Mainly data retrieval so depending on indexing. It needs to be said that most likely only a couple of states will exist²⁷.

5.3.17.1 The fields.

The fields of the priority table are listed below:

5.3.17.2 The indexes.

The table is indexed on the following fields:

5.3.17.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.18 property

The property table stores the actual values of the properties of operational parameters of a monitored object.

5.3.18.1 The fields

The fields of the property table are listed below:

5.3.18.2 Sample data

The table below lists a few examples of properties

5.3.19 service

The table service indicates the service that can be handled by the system. Out of the servicelist the administrator can indicate what services to expect.

5.3.19.1 The fields.

The fields are listed below:

5.3.19.2 The indexes.

The table is indexed on the following fields:

5.3.19.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.20 status.

The table status contains the possible states that a notification can have. As with the table priority these statuses are limited in number by default but can be expanded. Also here retrieval prevails above data entry and therefor indexing is important.

5.3.20.1 The fields.

The fields of the status table are listed below:

5.3.20.2 The indexes.

The table is indexed on the following fields:

5.3.20.3 The relationships.

Relationships with other tables

In the model this look like this:

5.3.20.4 Default values.

The status values are default for the system and for that reason are predefined.

5.3.21 type_of_issue.

This table will contain a list of all available issues that can be detected. All issues have a suggested priority setting. This is typically data retrieval and good indexing is needed.

5.3.21.1 The fields.

Fields of type_of_issue are:

5.3.21.2 The indexes.

The table is indexed on the following fields:

5.3.21.3 The relationships.

In the model this look like this:

5.3.21.4 Default values.

All checks are entered as code into the system. This table only works for the application only. The user can set specifics in the application only.

5.3.22 unprocessed_log

The user table contains the users that can login the monitoring application. It will also store if the users maintains the system. Mainly used for retrieval so properly indexed.

5.3.22.1 The fields.

The fields are listed below:

5.3.22.2 The indexes.

The indices of the table:

5.3.22.3 The relationships.

Relationships with other tables:

In the model this look like this:

5.3.22.4 Sample values.

To do

5.3.23 user_gnucomo

The user table contains the users that can login the monitoring application. It will also store if the users maintains the system. Mainly used for retrieval so properly indexed.

5.3.23.1 The fields.

The fields are listed in the table below:

5.3.23.2 The indexes

The table is indexed on the following fields:

5.3.23.3 The relationships.

Relationships with other tables

In the model this look like this:

6 Warnings that can be detected.

7 User Interface.

To be determined in the near future.

8 The installation process.

Since the system must make maintenance and security easier to use, the burden of installation should as easy as possible. Where possible the installation script should take away as much work as possible. Where settings need to be done, this should be done through an interface. However at no point we take the user's right to understand and work with system. Configuration-files should be easy to understand and the choice must be there to do installation manually. For the time being, we will use the manual installation procedure outlined below:

Since there is no binary package available for Gnucomo yet, you will need to compile and install Gnucomo from the source code. Before making the Gnucomo binaries, make sure you have the following packages installed:

• postgresql, postgresql-server, postgresql-develop
• AXE
• libxml2, libxml2-develop

To use gnucomo, you need to create a database and a configuration file. To make the database in your PostgreSQL server, log in as a DBA (DataBase Administrator, usually the user 'postgres') and create the database and a user who can use the database. Here is an example:

  createdb gnucomo
  createuser arjen

1. /etc/gnucomo.conf
2. /usr/local/etc/gnucomo.conf

8.1 Supported platforms.

The following two Linux distributions have been selected to be actively supported:

• Debian GNU/Linux (.deb packages)

• RedHat Linux (.rpm packages)

We will try and facilitate as many operating systems client-side and as many unices server-side, but efforts on testing out of the projects will be very minimalistic to ensure that the project keeps delivering new version and new features.

8.2 Installation on the server.

The following steps will be part of a script, that can automatically perform these steps:

1. Create the user gnucomo.

2. Make the directory as described in the chapter Sending messages to the central gnucomo system in the subchapter directories.(server-side).

8.3 Installation on a UNIX-client.

The following steps will be part of a script, that can automatically perform these steps:

1. Create the user gcm_client.

2. Make the directory as described in the chapter Sending messages to the central gnucomo system in the subchapter directories (client-side).

3. Creation of the database user gcm_input. This user has only the right to enter data into the database. There are no deletion, update or select-permissions.

9 Dependency on other free software.

The following list is a set of applications that will be used on to make our application work:

9.1 Related projects

There are a number of projects that can help Gnucomo or perform similar funtions:

• CRM114 - The Controllable Regex Mutilator.
• Hardware Lister (LSHW)
• LogRep - Logfile Extraction and Reporting
• Prelude Hybrid INtrusion detection system.
• Host and service monitoring.

10 Settings on the server machine.

10.1 Required.

The following settings are required to ensure that the functionality is as much as expected.

10.1.1 Timezone in GMT (UTC).

Since all international traffic registers all entries in UTC (Universal Time Coordinate) this system will do so as well. Therefore the clock has to be adjusted to that as well as the system settings.

These settings can found on a RedHat computer in the /etc/sysconfig/clock-file. On Debian this stored in the file /etc/timezone.

10.2 Suggested.

10.2.1 Use NTP.

If a computer is running for some time the clocks tend to be off the correct time. This makes it harder to detect what exactly happened exactly at what moment in time and reduce value of log-entries. Especially considering that ultimately all data is gathered in one central system. To overcome this Network Time Protocol (NTP RFC 13025 March 1992) has been created that explains a protocol to synchronize clocks through the Internet. Many operating systems like Microsoft Windows 2000 Server and FreeBSD, OpenBSD and Linux support this. A ntp-server (or ntpd) can be found at: http://www.eecis.udel.edu/~ntp/ It is strongly recommended that you use this.

11 Settings on the client machine.

11.1 Required.

11.2 Suggested.

11.2.1 Use NTP.

If a computer is running for some time the clocks tend to be off the correct time. This makes it harder to detect what exactly happened exactly at what moment in time and reduce value of log-entries. Especially considering that ultimately all data is gathered in one central system. To overcome this Network Time Protocol (NTP RFC 13025 March 1992) has been created that explains a protocol to synchronize clocks through the Internet. Many operating systems like Microsoft Windows 2000 Server and FreeBSD, OpenBSD and Linux support this. A ntp-server (or ntpd) can be found at: http://www.eecis.udel.edu/~ntp/ It is strongly recommended that you use this.

12 Appendices.

12.1 Appendix A. GNU Public License Version 2, June 1991.

Copyright (C) 1989, 1991 Free Software Foundation, Inc.

59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.

12.1.1 Preamble

The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too.

When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.

To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it.

For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights.

We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software.

Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations.

Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all.

The precise terms and conditions for copying, distribution and modification follow.

GNU GENERAL PUBLIC LICENSE

12.1.2 TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION

1. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) Each licensee is addressed as "you".

   Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does.

2. You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program.

   You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee.

3. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions:

   a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change.

   b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.

   c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.)

   These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

   Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program.

   In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License.

4. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:

   a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,

   b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,

   c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)

   The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable.

   If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code.

5. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.

6. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it.

7. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein.

   You are not responsible for enforcing compliance by third parties to this License.

8. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.

   If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances.

   It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice.

   This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License.

9. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License.

10. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.

    Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

11. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally.

12.1.3 NO WARRANTY

1. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVidE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

2. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

END OF TERMS AND CONDITIONS

12.1.4 How to Apply These Terms to Your New Programs

If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms.

To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found.

<one line to give the program's name and a brief idea of what it does.> Copyright (C) <year> <name of author> This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

Also add information on how to contact you by electronic and paper mail.

If the program is interactive, make it output a short notice like this when it starts in an interactive mode:

Gnomovision version 69, Copyright (C) year name of author

Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.

This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details.

The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `show c'; they could even be mouse-clicks or menu items--whatever suits your program.

You should also get your employer (if you work as a programmer) or your school, if any, to sign a "copyright disclaimer" for the program, if necessary. Here is a sample; alter the names:

Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker.

<signature of Ty Coon>, 1 April 1989 Ty Coon, President of Vice

This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License.