Engineering infrastructure as the basis of the information system
Often, IT professionals are poorly aware of the totality and complexity of an information system, and especially its environment. This is generally normal at the present time of narrow specialization, but it is still desirable for a real specialist to have at least an idea of ​​what might hinder the implementation of his plan. In the end, understanding that the work of your software package depends on many, many factors, including sometimes unobvious, will not hurt anyone.
The interpretation of the concept of an information system can be viewed, for example, on Wikipedia .
Just in case:
')
How do I see the information system infrastructure? A kind of pyramid, the top of which, the "first layer" - the consumer and the desired, processed and ready-to-use information. Information is perceived as intangible by man. The person is also an element of this infrastructure, and quite significant, like software; However, for some reason it is not customary to attribute it to technical and engineering elements. I will not become.
What is this information? What we see on the monitor screen, we hear from the speakers; that is, on the basis of which we make certain decisions - or consciously delegate the right to make these decisions, again, to computer technology.
Conventionally, the "second layer" of the pyramid - interfaces, devices for display, control, input-output. Why did I decide to bring interfaces into a separate entity? Because the decisions made by him depend on the “quality” of the information provided to the consumer - for quite subjective reasons. Yes, both technically and software-technical interfaces are a completely independent area.
"Communicate" interface systems directly with the third layer, with information processing devices that convert data arrays into a viewable view, with the same personal computers. In modern times, it is sometimes difficult to draw the line between an interface and a “calculator”, for example, smartphones and tablets.
A computer (laptop, tablet, smartphone) in itself, without data for processing, without communication with the source of information, is not a particularly necessary expensive piece of hardware. Formerly, data sources for computing systems were (if left behind the source — human and environmental reality / environment), teletypes, punched cards, punched tapes, then magnetic tapes ... Now the source data is usually taken via networks from other computing devices, mass storage places, through telecommunications. This is the fourth layer, which provides the connection of terminal devices that prepare information for human use with data sources.
The fifth and sixth layers - processing of primary, basic information, and data storage. These two layers can be understood, for example, the entire Internet - as a Saturday entertainment, or a data center, or a separate mainframe tied to a data storage system, a disk array, dozens of Fiber Channel channels through appropriate switches (here has its own mini-hierarchy, which also fit in the above scheme); or just a home NAS server.
But further (below) is just what is usually called engineering (as opposed to information) infrastructure. That which ensures the normal operation of all that is listed above.
It is clear that to implement the method of access to information just described (and operating this information) within the framework of the restrictions adopted, it is necessary to use technical, and not just technical, but high-tech means. Which can work only under certain conditions.
Further obvious:
This is how thick the seventh, engineering, layer of information infrastructure is - its own separate complex infrastructure of several, sometimes many, subsystems.
What is interesting is that this is not the end of the chain at all, as energy is transported further, urban and regional power supply networks, capacity generation, energy extraction ... But this, we will assume, beyond the scope of the topic, I promised to limit the scope of the enterprise.
The goal, in fact, is clear: to provide the consumer with the necessary level of responsibility (or just the right) with information for the final analysis and decision making; as an option - to get some pleasure (do you play computer games? And movies? And music?). The scale of the infrastructure designed to provide the consumer with information is directly dependent on the final importance of the problem being solved.
Information infrastructure. Examples are obvious: on the one hand, a “home network”, including a PC, a laptop and a couple of smart phones as clients, a router as a central hub, and the only communication channel to the provider; on the other hand, an enterprise with branches in half of the world, with a dozen and a half data centers throughout the country. The range is the widest. Accordingly, the implementation of these two infrastructure solutions requires equipment of different levels; manufacturers are aware of this "problem", and position devices in advance: "home wi-fi router", "switch for workgroups", "enterprise-wide server".
The complexity and high cost of solutions for each specific task is determined by the importance of this task - this was determined a long time ago. How valuable to you, as an individual, are your photos, videos, films and books stored on a laptop disk? Are they worth acquiring NAS with a level 10 RAID? Or will it be enough "flash" for 8 gigabytes? Or, for example, two flash drives of 32 gigabytes? Whether expenses for NAS will pay back your moral sufferings from possible loss of unique photos from Cyprus?
On the other hand, how valuable is the information that “lives” in a petabyte storage to a commercial bank? Do I need it, the repository, to mirror the second one in another city, or will there be enough daily backup on tapes? Will the losses from the bank’s downtime for repairs and restoration after the collapse of the storage system are so great that a “mirror” for a few tens of millions of dollars will turn out to be petty expenses?
Engineering infrastructure. Do I need to buy home an uninterruptible power supply, which will take up scarce space, but can save you from data loss? Are the costs justified - if the electricity in your home was turned off twice in the last five years? And which backup scheme of air conditioners in a data center should be chosen: N + 1 or 2N - considering that each extra air conditioner with a cooling capacity of, for example, 50 kW, will cost half a million?
The questions are quite rhetorical, any infrastructure should correspond to the scale of the task, and be based on the basis of calculations, if not too accurate.
As long as your “pictures” are placed on one CD, you have no problems with infrastructure: your data is always safe (if, of course, you don’t forget to take care of it), accessibility, and in the case of the collapse of your personal information system, this data is quite easy to recover. It is a different matter when the amount of data, to which at least sometimes it is necessary to have access, is tens and hundreds of terabytes; it is easy to imagine a situation in which you (the company) will suffer losses (financial, reputational, moral in the end) if this data is inaccessible, even for a short time.
You can try to determine the necessary scale of the infrastructure, and the necessary composition of its systems and subsystems.
The amount of information provided to the consumer (the one that needs to be stored "at home") can be fully assessed, this is what is used in daily activities and, as a result, "settles" on hard drives - with the exception of information that is random and unproductive (for an enterprise, say, deeply personal stocks of music on personal computers of employees; at home, for example, serials, which after viewing can be erased without regret, but which still occupy disk space). That is, those volumes of data that we operate at the moment, and which may be needed in the future. You can predict the growth of data in the presence of little statistics for previous periods (if you are an enterprise, and you don’t have such statistics, then your admins are either complete lazy or incompetent).
According to the available data volumes, the intensity of their use, and growth forecasts based on statistics, it is quite possible to estimate which technique at which “level” of infrastructure to use.
The nature of the final information can determine the content of the “upper layers” of the infrastructure pyramid,
the technical level of the interface and information delivery tools: a designer web-studio is unlikely to be able to compete effectively in the market, providing its leading employees with Pentium-III-based computers for rendering, and 14-inch monitors with a resolution of 800x600; on the other hand, many bookkeeping use such equipment, and may well still use it for years.
The intensity of data usage and their volumes determine the requirements for data transmission facilities: in the above example with accounting, it will be sufficient to have (for the average office size) a network built on the basis of a category 5 / 5e copper structured cable system and 10/100 Layer 2 switches The core of the bank’s data center network already requires optics for SAN and not only, Layer 3-4 switches with interfaces with a transmission speed of 2-8 (for SAN) and 1-10 (for others) gigabits per second.
The means of information processing are represented by servers of different performance (and, accordingly, power), execution, cost, and even purpose: from the level of the working group in the Midi-Tower building to the monsters of the IBM p795 class. In the “middle” segment, blade servers became deservedly popular (mainly due to the flexibility of solutions). The choice of a particular system depends on the complexity of the tasks to be solved (two big differences: the calculation of the thermal protection of a spacecraft using the finite element method, or “sapper play”) and, accordingly, the required performance.
Data storage is a fairly traditional task; it can be solved in different ways (the base, however, is now the same for all methods - mainly hard drives, if you have in mind operative storage, SSD for tasks that are critical to the speed of data processing, and magnetic tapes for backup and archival copies; generally speaking about information backup, a special conversation; in private life, CD / DVD and flash drives are added). The methods are chosen - oddly enough - depending on the required storage volumes and access speed. This can be a partition on a single HDD in the home computer, a RAID inside the server, a disk “shelf” or their array, or a Hi-End system of three (five, seven) 42U cabinets, one or two of which are “brains” controllers, and the rest - a disk array.
Requirements for engineering infrastructure systems are determined from the characteristics of everything listed above. The main indicator - power consumption, is the basis for further calculations - if we are talking about the company. Why main? Payment for electricity is the lion's share of the cost of operating data centers. The data processing center, where IT equipment with an electrical power of 250 kilowatts is installed, consumes almost 2.2 thousand megawatt-hours for calculations a year, and together with the engineering infrastructure from 3 to 4 thousand megawatt-hours, depending on the efficiency of engineering systems. In money today it means from ten to twelve million rubles. Such potential wastes should not be a surprise, and require prior evaluation. How?
We summarize the electric passport power consumed by each IT device, add 10-20% “just in case of fire” (according to our measurements during reporting, that is, with intensive calculations, the power consumption of the IT system increased by an average of 9.67 percent compared to with normal, steady daily capacity), if necessary, we add reserve for development, and we obtain the power that IT equipment will consume, that is, approximately 50-70% of the required total capacity (for the entire infrastructure). At the same time, the required power of the systems of guaranteed and uninterrupted power becomes clear, and at the same time the amount of heat that will need to be removed from the IT equipment and from the UPS, that is, you can estimate the power of the air conditioning system . After that, we are determined with the minimum allowable reservation levels, and the basis for draft calculations is ready.
The capacity of IT equipment is added to the capacity of the “engineer” systems, and as a result, the required external power supply capacity and the guaranteed power supply system are determined: emergency diesel generators, or something like that. That is - energy .
This technique, with some amendments, is applicable, in general, for a rough assessment of both the scale of the information infrastructure required for the “business” and the cost of it, taking into account measures to ensure some security. However, it is a separate issue of security, a thing that is multifaceted and diverse in its manifestations, and in extreme cases it can be very expensive. By the way, one of the security measures can be considered an increase in the reliability of both the entire infrastructure and the subsystems and components that make it up.
The reliability of technical systems is quite a multifaceted and entertaining science. But we are only interested in applied aspects; The main question is how to ensure acceptable reliability of the information infrastructure for reasonable money. Insofar as:
Based on the “price” of possible losses in case of failure of any component or system as a whole, one should choose a way to increase reliability and / or level of redundancy. Exclusively for the pleasure of comparing the two extreme options - a home network and a large enterprise. Let's go through the levels of the infrastructure pyramid from the bottom up, from the base to the superstructure, that is, let's start with the energy sector, tabulating the data of subsystems and components. The table would look good and intermediate options, for example, for small and medium-sized enterprises, but here the contrast between very small and large enough delivers stronger. I repeat: the numbers in the table are rather arbitrary, and are given so that you can visually compare costs. In reality, they can “swim” very strongly.
In my opinion, it turned out revealing. The different level of importance of the problem being solved determines different scales, and the scales determine the different cost of the solution.
Note, however, that there is no fundamental difference between the infrastructure, for example, Mobile Telesystems, or VTB-24, and your home network. No one and nothing but neighbors and a wallet will prevent you, if you consider your personal data invaluable, to protect them additionally: install a UPS for each device, and on the balcony an emergency gasoline generator for the most extreme case; connect to two independent providers by installing a separate router / switch on each channel; Add one more level to the existing NAS with a fifth-level RAID, in a mirror; put “in reserve” under the cabinet one more system unit, identical to standing on the table (the so-called “cold reserve”, aha), and insert the BlueRay writing drive into the one on the table, on which to cut another blank with a priceless one weekly; and the recorded blanks once a month to take to the bank cell; and so on. Do you need it?
The work of any sufficiently complex system depends on the normal functioning of some combination of other systems, small and not so much. Sometimes it is useful to remember this, especially when implementing high-tech projects. It is also useful to sometimes get out of the boundaries of your subject area, in order to have an idea of ​​how the systems as a whole function, what they depend on and what they influence.
The interpretation of the concept of an information system can be viewed, for example, on Wikipedia .
Just in case:
')
restrictions and agreements
1. The following is not a dogma, it is exclusively my vision of the situation. Based, however, on a fairly large experience of practical work.
2. In modern times, access to information implies almost exclusively the use of electronic means, and I will talk about them, leaving newspapers, books and classical libraries behind brackets.
3. We confine ourselves to quite insignificant scales, say, from the personal, home, smallest network, to the system of an individual enterprise — to consider the issue within the planet or at least the country can only be theoretically, and I am primarily interested in applied aspects.
4. At the same time I will set aside such an important component of the information system, as the software of the whole complex, I will focus on the “iron” components. This, of course, is wrong. We imply that software is present to some extent in any element of the system.
5. The figures below are quite conditional, as written in the design documentation, “reference”.
2. In modern times, access to information implies almost exclusively the use of electronic means, and I will talk about them, leaving newspapers, books and classical libraries behind brackets.
3. We confine ourselves to quite insignificant scales, say, from the personal, home, smallest network, to the system of an individual enterprise — to consider the issue within the planet or at least the country can only be theoretically, and I am primarily interested in applied aspects.
4. At the same time I will set aside such an important component of the information system, as the software of the whole complex, I will focus on the “iron” components. This, of course, is wrong. We imply that software is present to some extent in any element of the system.
5. The figures below are quite conditional, as written in the design documentation, “reference”.
Information infrastructure subsystems. Purpose of subsystems, tasks that they solve
How do I see the information system infrastructure? A kind of pyramid, the top of which, the "first layer" - the consumer and the desired, processed and ready-to-use information. Information is perceived as intangible by man. The person is also an element of this infrastructure, and quite significant, like software; However, for some reason it is not customary to attribute it to technical and engineering elements. I will not become.
What is this information? What we see on the monitor screen, we hear from the speakers; that is, on the basis of which we make certain decisions - or consciously delegate the right to make these decisions, again, to computer technology.
Conventionally, the "second layer" of the pyramid - interfaces, devices for display, control, input-output. Why did I decide to bring interfaces into a separate entity? Because the decisions made by him depend on the “quality” of the information provided to the consumer - for quite subjective reasons. Yes, both technically and software-technical interfaces are a completely independent area.
"Communicate" interface systems directly with the third layer, with information processing devices that convert data arrays into a viewable view, with the same personal computers. In modern times, it is sometimes difficult to draw the line between an interface and a “calculator”, for example, smartphones and tablets.
A computer (laptop, tablet, smartphone) in itself, without data for processing, without communication with the source of information, is not a particularly necessary expensive piece of hardware. Formerly, data sources for computing systems were (if left behind the source — human and environmental reality / environment), teletypes, punched cards, punched tapes, then magnetic tapes ... Now the source data is usually taken via networks from other computing devices, mass storage places, through telecommunications. This is the fourth layer, which provides the connection of terminal devices that prepare information for human use with data sources.
The fifth and sixth layers - processing of primary, basic information, and data storage. These two layers can be understood, for example, the entire Internet - as a Saturday entertainment, or a data center, or a separate mainframe tied to a data storage system, a disk array, dozens of Fiber Channel channels through appropriate switches (here has its own mini-hierarchy, which also fit in the above scheme); or just a home NAS server.
But further (below) is just what is usually called engineering (as opposed to information) infrastructure. That which ensures the normal operation of all that is listed above.
It is clear that to implement the method of access to information just described (and operating this information) within the framework of the restrictions adopted, it is necessary to use technical, and not just technical, but high-tech means. Which can work only under certain conditions.
Further obvious:
- First, without electricity, and meeting certain requirements, standards, the work of IT equipment, oddly enough, is impossible - such is the paradox.
- Secondly, the electronic components of IT equipment, in accordance with the laws of physics, work normally, without a large number of failures, can only in a rather limited temperature range, conditionally from -40 to + 50ºC, and 20 ± 2ºC is considered to be “comfortable”. At the same time, IT equipment itself is a source of heat: all consumed electrical energy is converted by components of IT systems into thermal energy.
- Thirdly, due to the technologies used today, there are restrictions on the level of relative humidity of the air: dew may fall out at elevated humidity, which means that a circuit in the electrical circuits is likely; At low humidity, static electricity may accumulate, and the likelihood of electrical components breaking through increases.
- Fourth, taking into account the “second and third”, IT equipment should be protected from unwanted external influences, starting from dust getting into them, and ending with a stone thrown by a bully. A layer of dust makes it difficult to remove heat from the components, and contributes to the accumulation of static; with a stone and everything is clear.
- There is still the problem of compact placement of IT systems and systems that provide them. That is - an architectural subsystem, a dedicated area, or a room, or a building, or structures, where the entire farm is located. This problem is solved in different ways, and often this solution requires a very large part of the funds allocated to the information system. Let's leave it aside, although this is also wrong - like the fact that we do not take into account software.
This is how thick the seventh, engineering, layer of information infrastructure is - its own separate complex infrastructure of several, sometimes many, subsystems.
What is interesting is that this is not the end of the chain at all, as energy is transported further, urban and regional power supply networks, capacity generation, energy extraction ... But this, we will assume, beyond the scope of the topic, I promised to limit the scope of the enterprise.
The purposes for which a specific information system is built; the dependence of infrastructure on the chosen target
The goal, in fact, is clear: to provide the consumer with the necessary level of responsibility (or just the right) with information for the final analysis and decision making; as an option - to get some pleasure (do you play computer games? And movies? And music?). The scale of the infrastructure designed to provide the consumer with information is directly dependent on the final importance of the problem being solved.
Information infrastructure. Examples are obvious: on the one hand, a “home network”, including a PC, a laptop and a couple of smart phones as clients, a router as a central hub, and the only communication channel to the provider; on the other hand, an enterprise with branches in half of the world, with a dozen and a half data centers throughout the country. The range is the widest. Accordingly, the implementation of these two infrastructure solutions requires equipment of different levels; manufacturers are aware of this "problem", and position devices in advance: "home wi-fi router", "switch for workgroups", "enterprise-wide server".
The complexity and high cost of solutions for each specific task is determined by the importance of this task - this was determined a long time ago. How valuable to you, as an individual, are your photos, videos, films and books stored on a laptop disk? Are they worth acquiring NAS with a level 10 RAID? Or will it be enough "flash" for 8 gigabytes? Or, for example, two flash drives of 32 gigabytes? Whether expenses for NAS will pay back your moral sufferings from possible loss of unique photos from Cyprus?
On the other hand, how valuable is the information that “lives” in a petabyte storage to a commercial bank? Do I need it, the repository, to mirror the second one in another city, or will there be enough daily backup on tapes? Will the losses from the bank’s downtime for repairs and restoration after the collapse of the storage system are so great that a “mirror” for a few tens of millions of dollars will turn out to be petty expenses?
Engineering infrastructure. Do I need to buy home an uninterruptible power supply, which will take up scarce space, but can save you from data loss? Are the costs justified - if the electricity in your home was turned off twice in the last five years? And which backup scheme of air conditioners in a data center should be chosen: N + 1 or 2N - considering that each extra air conditioner with a cooling capacity of, for example, 50 kW, will cost half a million?
The questions are quite rhetorical, any infrastructure should correspond to the scale of the task, and be based on the basis of calculations, if not too accurate.
Determining the scope of infrastructure, the composition of subsystems
As long as your “pictures” are placed on one CD, you have no problems with infrastructure: your data is always safe (if, of course, you don’t forget to take care of it), accessibility, and in the case of the collapse of your personal information system, this data is quite easy to recover. It is a different matter when the amount of data, to which at least sometimes it is necessary to have access, is tens and hundreds of terabytes; it is easy to imagine a situation in which you (the company) will suffer losses (financial, reputational, moral in the end) if this data is inaccessible, even for a short time.
You can try to determine the necessary scale of the infrastructure, and the necessary composition of its systems and subsystems.
The amount of information provided to the consumer (the one that needs to be stored "at home") can be fully assessed, this is what is used in daily activities and, as a result, "settles" on hard drives - with the exception of information that is random and unproductive (for an enterprise, say, deeply personal stocks of music on personal computers of employees; at home, for example, serials, which after viewing can be erased without regret, but which still occupy disk space). That is, those volumes of data that we operate at the moment, and which may be needed in the future. You can predict the growth of data in the presence of little statistics for previous periods (if you are an enterprise, and you don’t have such statistics, then your admins are either complete lazy or incompetent).
According to the available data volumes, the intensity of their use, and growth forecasts based on statistics, it is quite possible to estimate which technique at which “level” of infrastructure to use.
The nature of the final information can determine the content of the “upper layers” of the infrastructure pyramid,
the technical level of the interface and information delivery tools: a designer web-studio is unlikely to be able to compete effectively in the market, providing its leading employees with Pentium-III-based computers for rendering, and 14-inch monitors with a resolution of 800x600; on the other hand, many bookkeeping use such equipment, and may well still use it for years.
The intensity of data usage and their volumes determine the requirements for data transmission facilities: in the above example with accounting, it will be sufficient to have (for the average office size) a network built on the basis of a category 5 / 5e copper structured cable system and 10/100 Layer 2 switches The core of the bank’s data center network already requires optics for SAN and not only, Layer 3-4 switches with interfaces with a transmission speed of 2-8 (for SAN) and 1-10 (for others) gigabits per second.
The means of information processing are represented by servers of different performance (and, accordingly, power), execution, cost, and even purpose: from the level of the working group in the Midi-Tower building to the monsters of the IBM p795 class. In the “middle” segment, blade servers became deservedly popular (mainly due to the flexibility of solutions). The choice of a particular system depends on the complexity of the tasks to be solved (two big differences: the calculation of the thermal protection of a spacecraft using the finite element method, or “sapper play”) and, accordingly, the required performance.
Data storage is a fairly traditional task; it can be solved in different ways (the base, however, is now the same for all methods - mainly hard drives, if you have in mind operative storage, SSD for tasks that are critical to the speed of data processing, and magnetic tapes for backup and archival copies; generally speaking about information backup, a special conversation; in private life, CD / DVD and flash drives are added). The methods are chosen - oddly enough - depending on the required storage volumes and access speed. This can be a partition on a single HDD in the home computer, a RAID inside the server, a disk “shelf” or their array, or a Hi-End system of three (five, seven) 42U cabinets, one or two of which are “brains” controllers, and the rest - a disk array.
Requirements for engineering infrastructure systems are determined from the characteristics of everything listed above. The main indicator - power consumption, is the basis for further calculations - if we are talking about the company. Why main? Payment for electricity is the lion's share of the cost of operating data centers. The data processing center, where IT equipment with an electrical power of 250 kilowatts is installed, consumes almost 2.2 thousand megawatt-hours for calculations a year, and together with the engineering infrastructure from 3 to 4 thousand megawatt-hours, depending on the efficiency of engineering systems. In money today it means from ten to twelve million rubles. Such potential wastes should not be a surprise, and require prior evaluation. How?
We summarize the electric passport power consumed by each IT device, add 10-20% “just in case of fire” (according to our measurements during reporting, that is, with intensive calculations, the power consumption of the IT system increased by an average of 9.67 percent compared to with normal, steady daily capacity), if necessary, we add reserve for development, and we obtain the power that IT equipment will consume, that is, approximately 50-70% of the required total capacity (for the entire infrastructure). At the same time, the required power of the systems of guaranteed and uninterrupted power becomes clear, and at the same time the amount of heat that will need to be removed from the IT equipment and from the UPS, that is, you can estimate the power of the air conditioning system . After that, we are determined with the minimum allowable reservation levels, and the basis for draft calculations is ready.
The capacity of IT equipment is added to the capacity of the “engineer” systems, and as a result, the required external power supply capacity and the guaranteed power supply system are determined: emergency diesel generators, or something like that. That is - energy .
This technique, with some amendments, is applicable, in general, for a rough assessment of both the scale of the information infrastructure required for the “business” and the cost of it, taking into account measures to ensure some security. However, it is a separate issue of security, a thing that is multifaceted and diverse in its manifestations, and in extreme cases it can be very expensive. By the way, one of the security measures can be considered an increase in the reliability of both the entire infrastructure and the subsystems and components that make it up.
Reliability issues
The reliability of technical systems is quite a multifaceted and entertaining science. But we are only interested in applied aspects; The main question is how to ensure acceptable reliability of the information infrastructure for reasonable money. Insofar as:
- The main way to increase reliability is to reserve and duplicate components (devices, subsystems, communication channels, etc.).
- An additional way is to use highly reliable and therefore expensive components.
- Both of these ways to improve reliability require certain costs.
Based on the “price” of possible losses in case of failure of any component or system as a whole, one should choose a way to increase reliability and / or level of redundancy. Exclusively for the pleasure of comparing the two extreme options - a home network and a large enterprise. Let's go through the levels of the infrastructure pyramid from the bottom up, from the base to the superstructure, that is, let's start with the energy sector, tabulating the data of subsystems and components. The table would look good and intermediate options, for example, for small and medium-sized enterprises, but here the contrast between very small and large enough delivers stronger. I repeat: the numbers in the table are rather arbitrary, and are given so that you can visually compare costs. In reality, they can “swim” very strongly.
In my opinion, it turned out revealing. The different level of importance of the problem being solved determines different scales, and the scales determine the different cost of the solution.
Note, however, that there is no fundamental difference between the infrastructure, for example, Mobile Telesystems, or VTB-24, and your home network. No one and nothing but neighbors and a wallet will prevent you, if you consider your personal data invaluable, to protect them additionally: install a UPS for each device, and on the balcony an emergency gasoline generator for the most extreme case; connect to two independent providers by installing a separate router / switch on each channel; Add one more level to the existing NAS with a fifth-level RAID, in a mirror; put “in reserve” under the cabinet one more system unit, identical to standing on the table (the so-called “cold reserve”, aha), and insert the BlueRay writing drive into the one on the table, on which to cut another blank with a priceless one weekly; and the recorded blanks once a month to take to the bank cell; and so on. Do you need it?
As a conclusion
The work of any sufficiently complex system depends on the normal functioning of some combination of other systems, small and not so much. Sometimes it is useful to remember this, especially when implementing high-tech projects. It is also useful to sometimes get out of the boundaries of your subject area, in order to have an idea of ​​how the systems as a whole function, what they depend on and what they influence.
Source: https://habr.com/ru/post/268813/
All Articles