lec13-filesystemsx4

lec13-filesystemsx4 - CS162 Operating Systems and Systems...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
Page 1 CS162 Operating Systems and Systems Programming Lecture 13 File Systems, Naming, Directories, and Caching March 5, 2012 Anthony D. Joseph and Ion Stoica http://inst.eecs.berkeley.edu/~cs162 13.2 3/5/2012 Anthony D. Joseph and Ion Stoica CS162 ©UCB Spring 2012 Review: Magnetic Disk Characteristic • Cylinder: all the tracks under the head at a given point on all surface • Read/write data is a three-stage process: – Seek time: position the head/arm over the proper track (into proper cylinder) – Rotational latency: wait for the desired sector to rotate under the read/write head – Transfer time: transfer a block of bits (sector) under the read-write head • Disk Latency = Queuing Time + Controller time + Seek Time + Rotation Time + Xfer Time • Highest Bandwidth: – transfer large group of blocks sequentially from one track Sector Track Cylinder Head Platter Software Queue (Device Driver) Hardware Controller Media Time (Seek+Rot+Xfer) Request Result 13.3 3/5/2012 Anthony D. Joseph and Ion Stoica CS162 ©UCB Spring 2012 Goals for Today • Finish SSD discussion • Important System Properties • File Systems – Structure, Naming, Directories, Caching Note: Some slides and/or pictures in the following are adapted from slides ©2005 Silberschatz, Galvin, and Gagne adapted from slides ©2005 Silberschatz, Galvin, and Gagne. Many slides generated from my lecture notes by Kubiatowicz. 13.4 3/5/2012 Anthony D. Joseph and Ion Stoica CS162 ©UCB Spring 2012 Review: Solid State Disks (SSDs) • NAND Flash – Sector addressable, but stores 4-64 “sectors” per memory page – No moving parts (no rotate/seek motors) – Very low power and lightweight • Reading data similar to memory read ( 25μs ) – No seek or rotational latency – Transfer time: transfer a block of bits (sector) » Limited by controller and disk interface (SATA: 300-600MB/s) – Disk Latency = Queuing Time + Controller time + Xfer Time – Highest Bandwidth: Sequential OR Random reads
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Page 2 13.5 3/5/2012 Anthony D. Joseph and Ion Stoica CS162 ©UCB Spring 2012 Review: SSD Architecture – Writes • Writing data is complex! (~ 200μs – 1.7ms ) – No seek or rotational latency, Xfer time: transfer a sector • But, can only write empty pages (erase takes ~1.5ms!) – Controller maintains pool of empty pages by coalescing used sectors (read, erase, write), also reserve some % of capacity • Typical steady state behavior when SSD is almost full – One erase every 64 or 128 writes (depending on page size) • Write and erase cycles require “high” voltage – Damages memory cells, limits SSD lifespan – Controller uses ECC, performs wear leveling – OS may provide TRIM information about “deleted” sectors • Result is very workload dependent performance – Disk Latency = Queuing Time + Controller time (Find Free Block) + Xfer Time – Highest BW: Seq. OR Random writes (limited by empty pages) » Sequential easier to implement since can write all data to same pg Rule of thumb: writes 10x more expensive than reads, and erases 10x more expensive than writes 13.6 3/5/2012 Anthony D. Joseph and Ion Stoica CS162 ©UCB Spring 2012 Drive Health: SMART
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 11

lec13-filesystemsx4 - CS162 Operating Systems and Systems...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online