RAID Hard Drive Configuration
“RAID which stands for Redundant Array of Inexpensive Disks” is nothing but a configuration where a logical unit of large storage space is created by putting together a number of different hard disk drives. The purpose of doing this is that it is very robust and can be fault free with higher efficiencies that a single or various individual hard disk drives. RAID is a very good and tested technology across platforms and geographies which help the users in a fast access of data with a lot of reliability. RAID is now commonly becoming a part of every network to provide data loss remedies.
Components in RAID
RAID is a combination of physical and logical arrays. The important concepts in RAID are:
Mirroring is an important concept and component in RAID. It means that all the data is replicated as it is on an identical disk drive. If the data is lost or corrupted, the copy of this data can be used thus eliminating the risk of data loss.
Striping is another important component of RAID. More than one hard disk drive is used in parallel to read the data. This is primarily enhances the data read speed. Larger data is divided into smaller blocks and every block is written in a different disk to be read in parallel.
RAID LEVELS
There are several different RAID configurations using one or both the components to get the optimal result required based on the scenario and need. WE can call these combinations as levels however they are just representations and no indicated hierarchies of any sort. Let us look the three important RAID levels:
RAID-0 (Striping)
This level is focused on performance. As defined above, RAID 0 uses Striping to achieve the higher efficiency. In this level, redundancy or data backup is not considered. Data is simply fragmented into smaller standard sized stripes and stored in different disk drives in the array. The minimum number of disk drives required here is two, however, typical there are more number of drives are used.
The advantages of RAID 0 are:
Performance is very high
This is very simple and easy to implement
All the storage capacity is usable because there is no data redundancy
The disadvantages of RAID 0 are:
There is no fault-tolerance
Even if one drive fails, all the data will be useless
This can’t be used in environments where data is critical
Typical usages:
Editing and production of videos
Image processing
Any other application where high bandwidth is required
RAID-1 (Mirroring)
This is used when data redundancy is of primary importance. In RAID 1, we replicate the data using the technique of mirroring. As the data is present on two different drives, even if one drive fails, the data is not lost and can be retrieved from the other drive. As the data has to be written on two different drives, it takes longer to write the data however the read time is same as reading from one single drive.
The advantages of RAID 1 are:
Data is safe
Fault tolerant
The efficiency is as good as a single hard disk drive
Very easy to recover data if one drive fails
This is very simple to implement
The advantages of RAID 1 are:
Amongst all the different RAID configurations, this one is the most inefficient
It required double the storage space thus it is costly
This can only be used for data and the software can’t be installed twice
Usages of RAID 1:
Accounting applications like financial data and payroll system
It is used in any application where high availability is the priority
RAID-10 (A Stripe of Mirrors)
As the name suggests, this is combination of RAID 0 and RAID 1. In this configuration, mirroring is combined with stripping to get the best of both the worlds. It is often called “Mirrored Stripes”. The storage capacity is like RAID 1 where 50% of the memory is utilized due to redundancy however the speed efficiency is like RAID 0 which is very high.
The advantages of RAID 10 are:
Same fault tolerance as of RAID 1
Along with tolerance, high speed is also achieved
It can be rebuilt very quickly
The disadvantages of RAID 10 are:
Costly configuration as it required replication of data and stripes
Very inefficient writing
Overhead is very high and it is very expensive
Usages of RAID 10:
It is primarily used in servers where high tolerance is needed along with higher efficiencies.
CPU Cooling
CPU stands for central processing unit. As it processes most of the data sent to it in the system, it is referred to as central processor. The Central Processing Unit, CPU gets heat up as the processing and functioning generates a lot of heat. This heat is the result of this fast processing of data. It is normal for the processor to generate heat and they are built with some heat tolerance. However there is a threshold for the heat level. If the heat exceeds the tolerance level, there is a huge risk on the CPU’s functionality and it can be damaged. To overcome this problem, there are many cooling systems invented to keep the CPU cool. Here we will look at the two most common types of cooling systems; air and liquid cooling.
AIR Cooling
This is the simplest form of CPU cooling in which basic fans are used. A small fan is attached on the processor inside a computer or server which directly puts air on the CPU and keeps it cooler. There are no extra devices or care needed when this type of cooling is used. The fan comes fit from the manufacturer and there is no extra care needs to be taken. There are vents on the sides and back of the computer which lets the hot air go out thus keeping the CPU cooler.
Advantages of Air Cooling:
This is the cheapest form of cooling system
Relatively low noise as we can’t hear any specific high sounds
No need of extra parts like cables or tubes
Disadvantages of Air Cooling:
Not very effective with the growing speed of the processors
If the fan stops working due to a technical issue, it is sometimes not noticed until the CPU heats up and stops working
The heat is constantly coming out of the system in form of hot air
Liquid Cooling
After the declining effectiveness of air cooling systems and increasing efficiencies of processors, a new cooling system was developed called the basic liquid cooling system. In this technology, a system which is cooled by placing a liquid like water in the heat exchanger which transfers heat from the processor to this cooling device. The liquid constantly rotates through the pipes this making it very effective in transferring heat from processor and keeping the processor very well under threshold. Due to the thermal properties of water, the cooling effect of liquid cooling system is around 3000 times that of a air fan system.
Advantages of liquid cooling system:
Very effective system as it cools down the CPU very quickly and always keeps it under the threshold
Uses the simple ingredients like water
Can be used in the high power costly processors especially in the servers
Disadvantages of liquid cooling system:
A little costly in terms of installation and maintenance
Risk of leakage
Uses of extra parts like motor and pipes which consumes more power than fans
REFERENCES
Doane, Daryl Ann & Franzon, Paul D. “Multichip Module Technologies and Alternatives: The Basics.” Sep. 1993. Springer.
Murphy, Dave. "Maintain Your Water-Cooling Setup". Maximum PC Magazine.
Russel, R. M. "The Cray-1 Computer System". Readings in Computer Architecture. Feb. 2000. Gulf Professional Publishing.
Intel. “How to Apply Thermal Interface Material (TIM)”. Web. Retrieved on April 20. 2016.
Arpaci-Dusseau, Remzi H. & Arpaci-Dusseau, Andrea C. “Operating Systems: Three Easy Pieces [Chapter: RAID].” March. 2014. Arpaci-Dusseau Books.
