Register now or log in to join your professional community.
First industrial Distributed Systems What is a distributed system? The answer is not clear. In the literature, the term most often appears that a distributed system is a collection of independent computers, giving an impression of its users one compact system logically. This definition captures the main idea, but you should still define the main objectives of this technology and mention the aspect, so that the resulting distributed systems in the industry. These are aspects of computer networks and PLCs. The emergence of computer networks led to rapid technological development. Data transfer over long distances between computers caused the decentralization of calculation tasks and warehouses to store the data. Computer Networks will soon become available in every home. In industry, as in the development of computer systems, an important milestone was the creation of a freely programmable industrial controller (PLC). PLC is a computer with a strict specialization, which is characterized by high reliability and analyze data from multiple sources. Automation of many processes caused the factory even completely unrelated to market New Technologi, began to be peppered with them so sublime on a new and better level. Placing these two elements - computer networks and specialized computational units proved to be an extremely good idea and caused huge traffic associated with the creation of distributed systems. These systems assumptions firstly had to hide from the user the impression that he has before him a complex structure with many distant computers. Secondly, the interface of the system intended to ensure a uniform and consistent interface irrespective of interaction with it. Moreover, such solutions were characterized by high scalability. You could not worry about the fact that making additional computers to compute farms, their architecture, methods of communication and other related aspects. The objectives are being achieved and distributed systems today can be found at almost every step. Distributed system must be Transparent System. This concept defines: • Transparency of access - to harmonize the methods of access to data and hiding the differences in their representation • Transparency settings - users can not determine the physical location of the resource • Transparency trek - the possibility of transferring resources between servers without changing ways of referring to them • Transparency of movement - resources can be transmitted even when using them • Transparency of reproduction - hide from users the fact multiplication of resources • Transparency concurrency - the possibility of concurrent processing does not cause the formation of inconsistencies • Transparency of failure - masking transient failures of individual components of a distributed system • Transparency durability - masking method of storage resource The industry has also a very strict response time to emerging events, reliability and the ability to reconfigure the time of the crash. Security analysis of distributed industrial system is no different from the usual analysis of a distributed system. More attention should be paid to the said in the previous paragraph reliability and response time, but a decent administrator (or auditor), and so must pay attention to these factors. An important component that distinguishes these two types of systems is the nature of their work. Industrial systems are great systems, with a range of specialized equipment operating in difficult conditions. However, from the point of view of computer science - then it is a collection of computers, drivers, programs, input / output devices and communication media.
2 Summary of the dangers systems Every computer system must be properly secured. Each module is exposed to many dangers. Prior to analysis, distributed systems, and industrial, should first look at all the modules individually. Whether we're talking about a single device or a service in a particular computer. Safety is equal to the total security of the most vulnerable element. Only when all services separately, machinery, equipment, buses and communication channels are secured, you can go to look overall. It is the integration of all security components. At this point, the system enters a distributed agent and / or industrial uses. Contrary to appearances, except a few differences securing these systems differs little from securing a centralized system. Distributed systems are usually large structures. Configuration and all implementation can be tricky, because all elements of the installation are distant from one another. The main problems and dangers are communication protocols that links all system modules. Pay attention to the physical location of the various resources, because it is not always easy to define. Industrial systems under the premise of taking part in the control processes a large range of specialized equipment. Distinguished by the respective mechanisms that operate in a particular way in difficult conditions. Breaking the security may result in a very large financial losses, health of many people, or increased unreliability of the equipment. These systems must be protected further into account the increased delays, for example, during the pouring process invalid data. There must be mechanisms to react to external factors such as heat, moisture or dust - after removing the housing unit by the imprudent user may not result in an avalanche of miseries. We should also pay attention to the terrorist attacks, sabotage, riots and strikes.
3 The danger of the Internet Today, more and more factories, businesses and government facilities requires access to the Internet. The Internet is the primary communication channel and the advantages of its use are enormous. Unfortunately, the Internet can be added to existing systems in an imprudent manner, which is associated with very important consequences. The biggest mistake is to mix the company's internal network of part of having an Internet connection. In almost every case, there is no explanation for the fact, for example, call monitoring system with Internet access. Administrators often justify the fact that the proper configuration of servers that provide Internet services, settings, firewalls, static routing. The Internet can be a source of thousands of different attacks on corporate systems. Unpatched services, delays in updates, errors in scripts and applications, so-called. 0day vulnerabilities allow penetration of servers and sometimes even take control of them. Seized server can sometimes serve as a starting point for further penetration of the system, reconfiguring routers and devices in the company. Common and very frequently encountered errors in desktop and Web applications, ie errors Cross Site Scripting (XSS), SQL Injection, Cross Site Request Forgery (CSRF), Arbitary File Download (AFD) and many others can result in data leakage, their modification or other unpleasantness. Protection against these factors is, of course, on the side of programmers, webmasters and other people involved in the creation of applications for network services provided in the company. Administration of the server and its corresponding configuration is also not an easy task. Unfortunately, it goes beyond the subject of this material. In addition to these activities, there is one unusual (and sometimes controversial) way of securing a network. It depends on the implementation of the Honeynet network next to the network used in the company. Honeynety are false systems that simulate real network. Their task is to lure the aggressor and pulling his perception of the real systems. The attacker penetrates honeynety like a real network - there encounters various errors, delays, in contact with the same software. Has the opportunity to observe both end devices - the so-called. Honeypots (e.g. FTP) and the intermediate - e.g. Honeywall. Everything is done on specially prepared machines that do not have any useful information or in virtual environments. These systems analyze every move the aggressor, trying to identify and locate him. They are useful in the process of capturing the aggressor and provide very reliable evidence in court cases. 4 Risk of transmission media Transmission medium is a channel in which all of the data flow between components of a distributed system. In the most general way the media can be divided into wired and wireless media. Medium wire creates many problems, particularly in industrial systems. Wiring above all can be damaged, so you should be protected from access by unauthorized persons and against external factors. The damage may paralyze traffic information in the system, especially when devices will be between the key intermediate. It should be used in such a case, additional routers and relay traffic for other paths. Access to the medium wired or intermediary device creates the risk of attacks from the kind of man-in-the-middle. They consist in the fact that the aggressor plugged between two devices by capturing their movement. Forwards the traffic without creating any errors and suspicion. The attacker then has the opportunity to eavesdrop on these data, or even modify them. In order to protect the encryption should be used connections, optical media (optical fiber) network monitoring regarding attaching new hardware. A relatively new danger is that creating electromagnetic interference along the wires. Research is being done, which indicate the possibility of interception of the transmitted data based on the analysis of interference around the cables. Shielding or optical media are the cure for this property. Despite these negative effects associated with wired media should not implement the wireless media in distributed industrial systems, unless it is necessary. Typical threats of wireless media is primarily the risk of eavesdropping. In addition, watch for traffic analysis, scanning, identity theft, unauthorized access, repeat, attack the family man-in-the-middle attacks, flooding, distort packages, generate noise, loss of data as well as physical attacks such as theft of mobile devices. These risks are due to the nature of wireless media, which send out information in the ether, to which access is potentially everyone. Attacks on the media type is dangerous because of the difficulty of determining the place of attack. The basic element of the security policy should be to use cryptography - data encryption, application properly secure authentication methods, use of retransmission protocols, packet numbering, use the session and assign to them the tokens, the use of pseudo-random numbers, marking the time of transmission and methods such as IPSEC. 5 family Attacks Denial of Service DoS attacks from the family rely on repeated attempt to access a system resource. This test does not need to be successful. The process of the authentication or authorization and generated by the data can be the foundation of a DoS attack. This attack vector is basically a common problem any network and computer system and is widely featured on the Internet and in the specialist literature, so here is nadmieniony only additional aspect for industrial systems. Pouring some information of a distributed system is a very easy method which can effectively undermine the work of the company. Industrial systems are rarely designed to ensure that handle a surprisingly large dose of information because of their nature can predict the amount of data that appear in each node. The result of a DoS attack are increased delays in the victim segment or even completely turning off the attack. In real-time systems and critical devices workplace - increasing the minimum delay can cause large financial losses or other unpleasantness. In addition to the typical DoS attack, the newly observed danger is called. FDoS Financial or DoS. This new danger, which can be observed in companies strongly associated with New Technologies, who decide to Cloud Computing solutions in the style of IaaS (Infrastructure as a Service). IaaS is a collection of computers in the Data Centre are running the virtual machine. These machines are leased. Charges are on the take power and the company is settled in principle for any operation made. These are extremely scalable solution for several years has a great interest. Financial DoS attack is easy to carry out. It consists of multiple retrieval indicated resources in the cloud. Child's play scripts can generate a lot of traffic between the enterprise and the cloud which would imply charging for the use of cloud computing services. You may find that after a few days, the company may even go bankrupt. DoS attacks are the simplest to carry out attacks, and in fact the most difficult to defend. Once an attacker has adequate resources, is able to cut the key nodes of the system for many days. Stringent security policy on firewalls, use the Load Balancerów and implementation modules detect flooding in nearly every node of the system is the key to the success of preventing this attack. Most important, however, is the adaptation of hardware and software to the fact the occurrence of a DoS attack in such a way that its possible consequences were the least severe. 6 Attacks on methods of authentication and authorization Each system must provide authentication and authorization processes. Checking the credibility of documents, files and other objects is called the authentication process. Authentication mechanism verifies what permissions have an object (eg, John Doe) to other objects in the system (eg, valve controllers in heating or files on network drives). Two of these processes can be seen every day on virtually every step. Websites require logins and passwords, ATMs and mobile phones - PIN codes, lady at the post office while receiving parcels - a document with photo and signature. Although these procedures appear to be very complicated, there are many dangers associated with authentication and authorization. Authorization and authentication are two basic elements that protect resources from unauthorized access. The implementation of these mechanisms is quite intuitive administration. Sorry - the implementation of these mechanisms are ill total. This increases the susceptibility of the entire system when a very serious and easy to carry out the attack. Administration in addition does not monitor the actions, thinking that you've done enough. After all, even if the system for something to happen - you will know exactly who is responsible for this. Nothing could be further from the truth! Worse than the lack of security is only a false sense of security. Through such shortcomings can be witnessed most damaging attacks that rely on breaking simple dictionary passwords or logon to different systems at the default setting password access. How to peel a strategy when it comes to verification of the data? Let's examine the following solutions: • password access • magnetic cards • Smart Cards Access is password strings that are usually stored by the users. They are simple in drożeniu, but they have a lot of flaws. The main weakness is the possibility of efficient force attacks (Bruteforce) or a dictionary (Dictionary Based Attacks). When a password is short or very complicated, many times to try to "guess" the correct string in the authentication process. Automating this process can sometimes be complicated (difficult access aggressor terminals or devices in industrial systems and use less hardware than typical in classic information systems), but you can not ignore this type of attack. Passwords should be given access to enter secure cryptographic hash function. The result of their use, is stored in a database password in an implicit way, using the "hash" - that can not be converted into the previously specified password (process of changing the password on a hash is a one-way process). Failure to perform "hashing" passwords can have disastrous consequences. User passwords can be viewed by administrations, illegally used or transferred. No "hashing" increases damage resulting from security breaches. The attacker downloading the database has access directly to the data and can be used to further penetrate the system. It is worth noting that the "hash" does not solve all the problems. Secure hash functions is a relatively narrow group of cryptographic hash functions. Algorithms most of them are well known and there are many tools that can break such a "safe" password. The most common attack on passwords exposed hash function is the use of rainbow tables (rainbow tables). There are huge collections of data that a violation of speeding up the process. The use of parallel processing, the use of multiple processors cores CPU and GPU increasingly accelerates this process. Password less than12 characters may be broken on the fastest computers in a few hours, which is quite disturbing (in home conditions breaking 8-9 character passwords is also feasible within a reasonable time). The cure for this is to use long passwords, adding so. salt (salt), the use of several hash functions (reducing the problem of so-called. collision attacks and birthday). In addition, passwords can be easily overheard. It is well to protect the transmission medium and properly configure all of the intermediate and final. Be careful on the human aspect - the transfer of passwords, saving, saving in various applications. Magnetic cards contain a unique identifier and a user certificate (issued by an algorithm such as RSA). This solution is much safer than the use of passwords. First of all, prevents the formation of "new" identity, because the card is generated by the server (Administration) for the user and he passed. Authentication attempt is also strongly hampered because the attacker needs to get a magnetic card or build it yourself (along with the captured earlier certificates and other data). Authentication should be extended to use additional passwords (eg, passwords PIN). As a result, there is a possibility that the stolen card will not do much damage in the organization. Unfortunately, magnetic cards can be easily intercepted and copied. Employees need to be very careful of the way they store and under no circumstances can lend your cards to other employees. It should also implement appropriate procedures that will be used at the time of loss or destruction of the card. Smart cards are cards that have a microprocessor, RAM, ROM, EEPROM, sometimes even cryptographic coprocessor. In addition to authenticating, are also used, among others, payments or GSM technologies. Their main advantage is the high resistance to copying, modification of physical and wiretapping. Correct implementation of the system can be costly and requires a thorough rethinking of the authentication process - even analysis algorithms used in the same smart card. The problems of human nature are the same as in the case of magnetic cards.
First of all, secure storage of temporary data. They are right always necessary for proper system operation. Sent data (eg files sent by client applications) in the initial phase are received by the server and placed in the space, access to which is to basically every user (default). Despite the use of different methods of data filtering, remember - that the data that could be analyzed by the security system - must first physically in the picture. Such temporary data can be read and executed on the server side before the filter has time to analyze them and remove. An example of an attack on the authentication system may be the process of uploading files to the server (eg, HTTP). Suitable server-side scripts are responsible for the analysis sent us the file and in case of detection of the permitted format, transfer the file from the temporary file to the appropriate location. The attacker, he has access to the system (as for example its plain its user) probably has the ability to run a certain application group, which also generate temporary data. These data go to the corresponding (common) directory, and it is possible to identify the path to that folder. Knowing the path, an attacker is able to locate all the temporary files. Given that, the temporary directory must have a correspondingly low security requirements, an attacker can send malicious code through system discussed above and when the file will be analyzed through different filters - to do it on the server side. The file will be rejected by the filter, however, the malicious code is executed. The above example shows that the authorization policy should be set directly on each object individually. Area to which access has several objects at once) or several different types of objects) - should be as far as possible the most isolated from each other (through sandboxes, virtualization, "Jail" or by mere change of location). Authentication and authorization are very important elements in any system and must be carefully protected. In the distributed industrial systems is very important, because these processes are used very often, for many objects, very distant from one another. There is no possibility of replacing the machine by the people and should rely on their reliability. It should, however, be ensured and properly predict all emergencies that may occur.
It is very very useful to visit this page from InduSoft website and download the Training videos http://www.indusoft.com/Training/Online-Training-Videos
Really it's very excellent learning videoes and nearly it provides all information which u need about this subject
All Sincere wishes with success and progress to u
Join ISA society of Automation.
You can then participate in ISA regional sessions/chapters/workshops and get a ton free material and webinars to attend and learn.
Thanks
Salah.
Thanks Mr Ibrahm Hamdy Brother
.............
