Современные информационные технологии/4.Информационная безопасность

 

A.S.Begalin, Master of Science, senior teacher,

Kostanay state university of a name of A.Baytursynov

ONCE AGAIN ON INFORMATION SECURITY

Introduction. Computer Security

Information security - a big and serious subject on which the set of books including the Kazakhstan authors is written. The problem of information security costs and at our university. Here we mainly talk about the attacks from the outside, although we should not forget about internal security. Internal vulnerability of computer systems may affect and external security resources.

Two-faced security

We suggest to allocate two tasks of ensuring information security, because to some extent they can be practiced independently. Or not to be engaged.

The challenge facing all agencies and organizations that have access to the global network, is an ensuring protection against external attacks. Global access to information has the reverse negative side -  a communication channel with the global number of network hackers intruders. Their attacks are not perfect, but we can do all their attack failed.

The challenge before most institutions and organizations large enough (or sufficiently serious ) - is to ensure internal security. It is relevant in that case when a part of employees in certain situations and according to certain resources are regarded as foreign. In other words, when in your network not all is possible for all. In fact, it is actual always, but in the small companies similar severity often neglect. In many cases, it is relatively painless, but we must always think about the likelihood and extent of potential damage [1].

What you need for external protection.

There are two possibilities. First, the most common, is realized if the corporate network is no public (accessible from the Internet) servers, and local machine doesn’t have a global IP- address - the type of technology used by NAT (Network Address Translation). In this case from the outside world to get a direct connection to your LAN is not possible, and they are out of reach for outsiders. But there is always at least one computer which provides communication of all network with the outside world (lock), which must have an external address and is potentially vulnerable. If it can to be "broken", it isn't excluded that after that the burglar will be able to get access and to computers of an internal network. In this case all attention should be focused on protecting the external perimeter of the network, which in the simplest case represented by a single server.

The simplest solution consists of two steps: 1) install and configure properly on the server firewall and 2) constantly to watch vulnerabilities on this computer. If to do it accurately, you can feel safe. Pay attention to paragraph 2, it is the second only in order, but not least (the last but not the least, as the English say). Without executing all work under paragraph 1 may be useless. That so it didn't happen, just and it is worth using the safety scanner (about it a bit later) [2].

If in your network has computers with global IP- addresses, the security is becoming more difficult - attention require all such hosts .

What you need for self-protection

If necessary to ensure internal safety the situation even more becomes complicated. In a limit case all internal computers have to be considered as potentially vulnerable with all that it implies: tracking of their configuration and continuous monitoring of vulnerabilities. Fortunately, this task, the good scanner of safety allows to simplify strongly and substantially to automate.

In addition, there immediately arises for at least another couple of tasks: 1) the creation of competent architecture of an internal network (excluding one-time increase of security, it also minimizes the effort to further security) and 2) the development and observance of the so -called policy (or several politician) security, that is a set of the rules concerning various questions of network functioning (creation and modification of passwords, regulations of access to these or those resources). The first point can be made that is called "for time", and here the second should be engaged constantly, and it appears, as a rule, the heaviest - mainly, in the organizational plan. The most unpleasant that it is impossible to automate its realization substantially.

But, to reach ideally working architecture and a security policy rather difficult, it is possible to construct a network with very high degree of security. In these circumstances key there is besides a continuous monitoring of vulnerabilities of separate computers. In principle, even the network far from an ideal in which each separate computer is impregnable, can be considered as well protected. In other words, the competent solution of a problem of audit of network safety (which it is well automated) allows to compensate substantially defects in those areas where the solution of problems is more labor-consuming or complicated for one reason or another [2].

Vulnerabilities

We use the term "vulnerabilities" in expanded sense, including here actually vulnerabilities (vulnerabilies) and defects (exposures) as both those, and others are found, as a rule, by means of safety scanners. "Classical" vulnerabilities are mistakes in the software (for example, the overflow of the buffer not caught by developers) which can be used by burglars for receiving unauthorized access and achievement of others bad for you and the desired purposes for them. "Defect" (more precisely, but it is longer "susceptibility to foreign influence") is the violation of safety caused by the wrong control or a configuration of the program, that is wine not developers, and administrators. Typical obvious defect - the empty or easily guessed password of access somewhere.

So, we will consider that vulnerabilities (they in computer popular speech - holes) are dangerous mistakes either in the program, or in its configuration. They have the unpleasant property that in the absence of continuous control their number seeks to grow and in all without exception cases sooner or later reaches absolutely unacceptable level. Process, in general, is quite clear: if you regularly do not understand on his desk, then sooner or later there is becoming a mess, which try to tactfully called "labor unrest". Such a  mess" in a computer network is much more dangerous, because it first, it is not visible to the naked eye, and, secondly, the consequences promise to big trouble [3].

"Fertility" of vulnerabilities, of course, has a very specific and understandable reasons. The first - the emergence in a network of the new software with their own "cockroaches" and sporadic changes of settings which aren’t always carried out carefully enough. The second, more fundamental, with which it is impossible to fight organizational - the discovery of new "holes" in the existing software. There is the whole information industry which is engaged in collecting, the publication and the analysis of found vulnerabilities. Regular bulletins (bagtraki) about new vulnerabilities receive information from both software developers and independent experts, professionals and even from hackers who seek fame. Unfortunately, many serious hackers tend not to advertise their vulnerability found wanting as long as possible on the sly use them for their own purposes. In this context it is possible to tell that any base of vulnerabilities by definition is incomplete. This is bad, but this can be partly fight.

 

References

1 Site http://www.atelierweb.com/awft/.

2 Site http://www.firewallleaktester.com/.

3 Computer Press Magazine, № 10, 2006.