26 Questions
How is non-volatile memory viewed in the context of computer systems?
As part of the I/O system
What is a benefit of RAID 5 configuration?
It offers better data redundancy
What is the purpose of error-correcting codes in storage systems?
To detect and correct data errors during transmission
How does parity work in data restoration?
By calculating the checksum of data blocks
What is a characteristic of high-performance storage architectures?
They are designed to optimize data transfer rates and storage capacity
What is a key feature of RAID 1 configuration?
It mirrors data to provide real-time duplication
What is the primary performance bottleneck in the storage node?
Disks
What is the approximately cost per IOPS of the storage system?
$7
What is the formula to calculate availability?
MTTF / (MTTF + MTTR)
What is the estimated mean time to fail of the rack?
531 hours
What is the main advantage of using RAID 5 over RAID 1?
Increased storage capacity
What is the purpose of error-correcting codes in storage systems?
To detect and correct data errors
What is the primary benefit of using parity in storage systems?
Data restoration in case of failure
What is the main characteristic of high-performance storage architectures?
High IOPS performance
What is the estimated cost per terabyte of the storage system?
$1,000
What is the main factor contributing to the high cost of the storage system?
Disks
What is the main purpose of seek time in magnetic disks?
To position the arm over the proper track
What is the average rotational latency (ARL) in magnetic disks?
8 ms to 2 ms
What is the primary benefit of RAID systems?
To enhance the availability of data
What is the main characteristic of RAID 0?
It is not redundant but more efficient
What is the primary difference between reliability and availability?
Reliability is the probability of failure, while availability is the probability of unavailability
What is the formula to calculate the average seek time (AST)?
Sum of the time for all possible seeks / Total number of possible seeks
What is the primary benefit of increasing the recording density in magnetic disks?
To increase the storage capacity
What is the formula to calculate the disk access time (DAT)?
Seek time + Rotational latency + Transfer time + Controller time + Queuing delay
What is the primary advantage of using error-correcting codes in magnetic disks?
To detect and correct errors
What is the main purpose of parity in RAID systems?
To restore data in case of failure
Study Notes
Storage
- Non-volatile memory can be viewed as part of the memory hierarchy system or I/O system
- Characteristics of storage include: seek time, rotational latency, and transfer time
- Seek time: 5-12 ms, with actual time being 25-33% of reported time due to locality of disk references
- Rotational latency: average 8 ms to 2 ms, depending on RPM (3,600 to 15,000) and disk diameter (1.0 to 3.5 inches)
- Transfer time depends on transfer size per sector, rotation speed, and recording density
- Magnetic disks evolution has led to increased density, reduced prices, and increased RPM
Magnetic Disks Performance
- Average seek time (AST) is the sum of time for all possible seeks divided by the total number of possible seeks
- Rotational latency (ARL) is half the rotation period, with common values being 5,400 and 7,200 RPM
- Transfer time depends on transfer size, rotation speed, and recording density
- Disk access time (DAT) is the sum of seek time, rotational latency, transfer time, controller time, and queuing delay
RAID Systems
- RAID (Redundant Array of Independent Disks) systems are designed for reliability and availability
- Characteristics: multiple simultaneous accesses, data stripping, and mirroring to increase availability
- Reliability vs availability: more disks mean greater fail probability, but failures do not necessarily lead to unavailability
- RAID levels summary: level 0 (not redundant, but efficient), level 1 (mirroring), and level 3 (striping with parity)
I/O Servers - Clusters
- Performance evaluation: the network switch can be the bottleneck if it cannot support the required bandwidth
- Cost: disks represent almost 70% of the total cost, with a cost per terabyte of almost $1,000 and cost per IOPS of about $7
- Dependability: mean time to failure (MTTF), mean time to repair (MTTR), and mean time between failures (MTBF)
- Availability: MTTF / (MTTF + MTTR)
I/O Servers - Clusters (cont.)
- Dependability: resulting mean time to fail of the rack, with assumptions about MTTF for various components (CPU, memory, enclosure, disks, controllers, switches, power supplies, and fans)
- Calculation of failure rate and MTTF for the rack
This quiz covers the characteristics of non-volatile memory, including seek time, rotational latency, and transfer time, as part of the memory hierarchy and I/O systems.
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free
