Disc scheduling is an important process in operating systems that determines the order in which disk access requests are serviced. The objective of disc scheduling is to minimize the time it takes to access data on the disk and to minimize the time it takes to complete a disk access request.
Read more: Sistem Operasi – 18 – Queueing Theory, Disk Scheduling, and File Systems.| Time seek | Note from CS162 Lecture 18 (4/2/2020): Queueing Theory, Disk Scheduling & File Systems |
| 3:17 | flash memory |
| 14:00 | queue: simple deterministic world |
| 21:40 | reality: a bursty world |
| 28:28 | background: general use of random distribution |
| 48:19 | a little queuing theory: some result |
| 55:30 | why unbounded response time? |
| 57:37 | a little queuing theory: an example |
| 1:02:14 | optimize I/O performance |
| 1:04:04 | I/O scheduling discussion |
| 1:06:00 | Disk scheduling |
| 1:10:22 | how do we hide I/O latency |
| 1:11:11 | from storage to file system |
| 1:13:27 | I/O & storage layer |
| 1:14:11 | C (language) low level operations |
| 1:14:36 | building a file system |
| 1:16:40 | translating from user to system view |
| 1:18:10 | disk management policy |
| 1:20:54 | what does the file system need |
| 1:21:47 | data structure on disk |
| 1:23:45 | summary |
Poin-poin Penting
- on a stable system, arrival rate equal to departure rate (little’s law)
- if arrival rate greater than departure rate, the system become unstable, since the queue will grow infinitely. however, in reality, the queue capacity is finite, so the response time should be finite. but is has cost: if the queue is full, further arrival will be rejected.
- file system used to translate from user (file name and path) to system view (block/sector of a disk)
Latar Belakang
I/O transaction is slow, need to hide or optimize.
Queuing theory resources
