CE 01000 6 Operating Systems Lecture 20 Disk

CE 01000 -6 Operating Systems Lecture 20 Disk I/O

Overview of lecture l In this lecture we will look at: l l l Disk Structure Disk Scheduling Disk Management Swap-Space Management Disk Reliability/performance Stable-Storage Implementation

Disk Structure l Disk drives are addressed as large 1 -dimensional arrays of logical blocks, where the logical block is the smallest unit of transfer. l The 1 -dimensional array of logical blocks is mapped into the sectors of the disk sequentially. l l Sector 0 is the first sector of the first track on the outermost cylinder. Mapping proceeds in order through that track, then the rest of the tracks in that cylinder, and then through the rest of the cylinders from outermost to innermost.

Disk Scheduling l The operating system is responsible for using hardware efficiently — for the disk drives, this means having a fast access time and disk bandwidth. l Access time has two major components l l Seek time is the time for the disk are to move the heads to the cylinder containing the desired sector. Rotational latency is the additional time waiting for the disk to rotate the desired sector to the disk head.

Disk Scheduling (Cont. ) l l Minimise seek time Seek time seek distance Disk bandwidth is the total number of bytes transferred, divided by the total time between the first request for service and the completion of the last transfer. Several algorithms exist to schedule the servicing of disk I/O requests.

Example request queue l We illustrate them with a request queue (0199 range). 98, 183, 37, 122, 14, 124, 65, 67 Head starts at 53

First Come First Served (FCFS) Below shows total head movement of 640 cylinders.

Shortest Seek Time First (SSTF) l Selects the request with the minimum seek time from the current head position. l SSTF scheduling is a form of SJF scheduling; may cause starvation of some requests – because newly arriving requests may be closer to current position than a request that has been waiting some time l Illustration shows total head movement of 236 cylinders.

SSTF (Cont. )

SCAN l The disk arm starts at one end of the disk, and moves toward the other end, servicing requests that move it in the same direction until it gets to the other end of the disk, where the head movement is reversed and servicing continues. l Sometimes called the elevator algorithm. l Illustration shows total head movement of 236 cylinders. l But service requests can sometimes wait for a while – large variation in waiting times

SCAN (Cont. )

C-SCAN l Circular SCAN (C-SCAN) l Provides a more uniform wait time than SCAN. l The head moves from one end of the disk to the other, servicing requests as it goes. When it reaches the other end, however, it immediately returns to the beginning of the disk, without servicing any requests on the return trip.

C-SCAN (Cont. ) l l Treats the cylinders as a circular list that wraps around from the last cylinder to the first one. Illustration shows total head movement of 383 cylinders.

C-SCAN (Cont. )

C-LOOK l Version of C-SCAN l Arm only goes as far as the last request in each direction, then reverses direction immediately, without first going all the way to the end of the disk. l LOOK is a similar version for SCAN

C-LOOK (Cont. )

Which Disk-Scheduling Algorithm? l SSTF is common and has a natural appeal l SCAN and C-SCAN perform better for systems that place a heavy load on the disk. l Performance depends on the number and types of requests. l Different file-allocation methods give different profiles of use of disk accesses - linked and indexed can be scattered over the disk, whereas contiguous are close together l The disk-scheduling algorithm should be written as a separate module of the operating system, allowing it to be replaced with a different algorithm if necessary. l Either SSTF or LOOK is a reasonable choice for the default algorithm.

Disk Management l Low-level formatting, or physical formatting — Dividing a disk into sectors that the disk controller can read and write. l To use a disk to hold files, the operating system still needs to record its own data structures on the disk. l l Partition the disk into one or more groups of cylinders. Logical formatting or “making a file system”. l Boot l l block initialises system. The bootstrap is stored in ROM. Bootstrap loader program - loads full bootstrap program from disk which in turn then loads the rest of the operating system

Disk Management (Cont. ) l l problem with bad blocks - blocks of disk that are defective and unusable - these need to be identified so that the OS does not attempt to use them - controller maintains a list some OS put aside a number of spare sectors that are hidden from normal OS use, but can be used to replace bad sectors as they are discovered - called sector sparing

Swap-Space Management l Swap-space — Virtual memory uses disk space as an extension of main memory. l Swap-space can be carved out of the normal file system - in a swap file or it can be in a separate disk partition. l Swap-space management examples: l l l Unix 4. 3 BSD allocates swap space to a process when it starts; has space for a text segment (the program) and a data segment. Kernel uses swap maps to track swap-space use. Solaris 2 allocates swap space only when a page is forced out of physical memory, not when the virtual memory page is first created.

Disk Reliability & Performance l Several improvements in disk-use techniques involve the use of multiple disks working cooperatively. l Disk striping uses a group of disks as one storage unit - splits data across several disks - so increases data transfer rate, but does not improve reliability l l bit-level striping - splits bits of each byte across several disks - bit i of each byte goes to disk i block-level striping - blocks of a file across several disks - block k mod n of a file goes to disk k mod n (n = number of disks)

Disk Reliability & Performance (Cont. ) l RAID (Redundant Array of Inexpensive Disks) schemes improve performance and improve the reliability of the storage system by storing redundant data. l l Mirroring or shadowing keeps duplicate of each disk. Block interleaved parity - uses blocks of parity information (error detection and correction bits) much less redundancy.

Stable-Storage Implementation l Storage is stable if information can never be lost l used to keep logs of activities that are involved in atomic transactions so that the transaction can be rolled back if a problem occurs. l To implement stable storage: l l Replicate information on more than one non-volatile storage media with independent failure modes. Update information in a controlled manner to ensure that we can recover the stable data after any failure during data transfer or recovery.

References l Operating System Concepts. Chapter 12.