Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
29 Cards in this Set
- Front
- Back
|
How fast do magnetic disks rotate? (low end to high end) |
50 to 250 times per second (i.e. 7200 rpm = 120 rps) |
|
|
With respect to a magnetic disk, what is "Positioning Time"? |
Positioning Time = seek time + rotational latency |
|
|
What is "seek time"? |
The time it takes for the arm to move over the cylinder (i.e. the band of the disk) containing the data |
|
|
What is the average latency of a hard drive disk? |
1/2 a revolution (your data is at least under the head or at most, one full rotation around) |
|
|
True or False. Typically, latency is more than seek time. |
False. Typically it takes longer to move the arm. |
|
|
Why are SSDs faster the HDDs? |
No moving parts so no seek time or latency. |
|
|
On a disk drive, where will you find address 0? |
On the outermost ring/cylinder. |
|
|
What is SCSI? |
"Skuzzy", aka Small Computer System Interface. It's a shared bus that connects disks to computers and computers to eachother. |
|
|
What is the difference between SAN (Storage Array Networks) and NAS (Network-Attached Storage)? |
SAN requires SCSI Fibre Channel to connect to it. NAS doesn't (any machine that can connect to a LAN can use NFS, CIFS, or HTTP to access data) |
|
|
How does the OS minimize seek time? |
By storing files in the same cylinders, keeping connected data close together |
|
|
What is wrong with FCFS disk scheduling? |
We jump around all over the disk, which can be enormously expensive in terms of seek time |
|
|
What is wrong with SSTF disk scheduling? |
Starvation of jobs that need some data outside the heavily demanded area. |
|
|
What is the SCAN algorithm for disk scheduling? |
The SCAN algorithm, aka the "elevator algorithm" moves across the disk sequentially to the next requested sector, but could only make stops at requested sectors along the way. Requests below the arm's current location have to wait for it to return. |
|
|
What's the difference between the SCAN and LOOK algorithm? |
SCAN goes all the way to the last sector on the disk, LOOK reverses after reaching the highest sector requested. |
|
|
What is the C-SCAN algorithm? |
The Circular Scan is not like an elevator in that it doesn't reverse direction. When it reaches the end, it starts back at the beginning |
|
|
How does the OS improve latency? |
It generally doesn't since it's difficult for the OS to calculate. The OS focuses on improving seek time. |
|
|
Which disk scheduling algorithm do SSDs use? |
Disk algorithms don't really matter since there's no seek time to minimize. They typically use FCFS. |
|
|
What is the advantage of C-SCAN over SCAN, for disk scheduling? |
With SCAN, slots in the center are serviced twice as often as slots at the edges (i.e. like floor 5 is serviced twice as much by elevator, than floors 10 and 5 (in 10 story building)) With C-SCAN, all slots are visited equally as often |
|
|
True or False: With a small I/O load, the scheduling algorithm doesn't really matter. |
True (slide23) |
|
|
What is low-level formatting (aka physical formatting)? |
Dividing the disk up into sectors (usually done at the factory) that hold data, header info, error correction code. Usually 512 bytes but newer disks are 4K |
|
|
What does it mean to partition? |
Divide the disk up into rings/cylinders, with each "partition" treated as a separate logical disk |
|
|
What is the MBR on a disk? |
The Master Boot Record on a disk holds top level data about the disk, for example the boot code and a partition table concerning the partitions on the disk. It's generally located starting at address 0 on the disk. |
|
|
What is the Swap Space? |
An area either carved out of normal file system or using some special partition, used by virtual memory to pretend to the CPU that there is more RAM |
|
|
How does the OS know which pages are in which Swap Space slot? |
It maintains an in-memory data structure that maps all the processes currently using the page in that slot. |
|
|
What is RAID? |
Redundant Array of Independent Disks. Multiple disk drives that work together to provide data redundancy |
|
|
What is Data Striping? |
Using multiple disks as if they are one. A file can have parts of itself stored sequentially on each. |
|
|
What are some advantages provided by Data Striping? |
Speed - files can be accessed much quicker since multiple arms could be retrieving the file at once. Data redundancy - if one disk fails, the others can be used to reconstruct what was missing. |
|
|
What are "hot-spare" disks? |
Hard disks that are part of a storage array but are only used when another disk in the array fails. |
|
|
What is meant by RAID 1, RAID 2, .... RAID 6? |
https://en.wikipedia.org/wiki/Standard_RAID_levels |
