4 27 11 CS 61C Great Ideas in Computer Architecture Machine Structures Redundant Arrays of Inexpensive Disks Instructors Randy H Katz David A PaHerson hHp inst eecs Berkeley edu cs61c fa10 4 27 11 Spring 2011 Lecture 27 1 4 27 11 Spring 2011 Lecture 27 Agenda Agenda RAID Administrivia Course Summary Randy Cal Culture Dave Course EvaluaTon 4 27 11 2 Spring 2011 Lecture 27 3 RAID Administrivia Course Summary Randy Cal Culture Dave Course EvaluaTon 4 27 11 Spring 2011 Lecture 27 EvoluTon of the Disk Drive 4 Arrays of Small Disks Can smaller disks be used to close gap in performance between disks and CPUs ConvenTonal 4 disk designs 3 5 5 25 10 IBM 3390K 1986 Low End 14 High End Disk Array 1 disk design 4 27 11 IBM RAMAC 305 1956 Spring 2011 Lecture 27 Apple SCSI 1986 3 5 5 4 27 11 Spring 2011 Lecture 27 6 1 4 27 11 RAID Redundant Arrays of Inexpensive Disks Replace Small Number of Large Disks with Large Number of Small Disks 1988 Disks Capacity Volume Power Data Rate I O Rate MTTF Cost IBM 3390K 20 GBytes 97 cu e 3 KW 15 MB s 600 I Os s 250 KHrs 250K IBM 3 5 0061 320 MBytes 0 1 cu e 11 W 1 5 MB s 55 I Os s 50 KHrs 2K x70 23 GBytes 11 cu e 9X 3X 1 KW 120 MB s 8X 3900 IOs s 6X Hrs 150K Disk Arrays have potenTal for large data and I O rates high MB per cu e high MB per KW but what about reliability 4 27 11 Spring 2011 Lecture 27 7 Redundant Arrays of Inexpensive Disks RAID 1 Disk Mirroring Shadowing recovery group Each disk is fully duplicated onto its mirror Very high availability can be achieved Bandwidth sacri ce on write Logical write two physical writes Reads may be opTmized Most expensive soluTon 100 capacity overhead 4 27 11 Spring 2011 Lecture 27 9 Files are striped across mulTple disks Redundancy yields high data availability Availability service sTll provided to user even if some components failed Disks will sTll fail Contents reconstructed from data redundantly stored in the array Capacity penalty to store redundant info Bandwidth penalty to update redundant info 4 27 11 Example small read D0 D5 large write D12 D15 4 27 11 D0 D1 D2 D3 P D4 D5 D6 D7 P D8 D9 D10 D11 P D12 D13 D14 D15 P D16 D17 D18 D19 P D20 D21 D22 D23 P Disk Columns 27 Spring 2011 Lecture Increasing Logical Disk Address Stripe 11 8 Redundant Array of Inexpensive Disks RAID 3 Parity Disk 10010011 11001101 10010011 logical record Striped physical records P 1 0 1 0 0 0 1 1 P contains sum of other disks per stripe mod 2 parity If disk fails subtract P from sum of other disks to nd missing informaTon 4 27 11 Spring 2011 Lecture 27 Redundant Arrays of Inexpensive Disks RAID 4 High I O Rate Parity Insides of 5 disks Spring 2011 Lecture 27 1 1 0 0 1 1 0 1 1 0 1 0 0 0 1 1 1 1 0 0 1 1 0 1 10 InspiraTon for RAID 5 RAID 4 works well for small reads Small writes write to one disk OpTon 1 read other data disks create new sum and write to Parity Disk OpTon 2 since P has old sum compare old data to new data add the di erence to P Small writes are limited by Parity Disk Write to D0 D5 both also write to P disk 4 27 11 D0 D1 D2 D3 P D4 D5 D6 D7 P Spring 2011 Lecture 27 12 2 4 27 11 RAID 5 High I O Rate Interleaved Parity Problems of Disk Arrays Small Writes RAID 5 Small Write Algorithm D0 Independent writes possible because of interleaved parity D1 D2 D3 P D4 D5 D6 P D7 D8 D9 P D10 D11 Increasing Logical Disk Addresses 1 Logical Write 2 Physical Reads 2 Physical Writes D0 D0 D1 D2 D3 old data 1 Read new data P old 2 Read parity XOR D12 P D13 D14 D15 P D16 D17 D18 D19 D20 D21 D22 D23 P XOR Example write to D0 D5 uses disks 0 1 3 4 4 27 11 3 Write Disk C olumns Spring 2 011 Lecture 27 D0 13 4 27 11 Tech Report Read Round the World D1 4 Write D2 D3 Spring 2011 Lecture 27 P 14 RAID I December 1987 RAID I 1989 Consisted of a Sun 4 280 workstaTon with 128 MB of DRAM four dual string SCSI controllers 28 5 25 inch SCSI disks and specialized disk striping soeware 4 27 11 Spring 2011 Lecture 27 15 4 27 11 Spring 2011 Lecture 27 16 RAID II RAID II 1990 1993 Early Network AHached Storage NAS System running a Log Structured File System LFS Impact 25 Billion year in 2002 Over 150 Billion in RAID device sold since 1990 2002 200 RAID companies at the peak Soeware RAID a standard component of modern OSs 4 27 11 Spring 2011 Lecture 27 17 4 27 11 Spring 2011 Lecture 27 18 3 4 27 11 RAID Summary Agenda Logical to physical block mapping parity striping read modify write processing Embedded caches and orchestraTng data staging between network interfaces parity hardware and le server interfaces Failed disk replacement hot spares background copies and backup Embedded log structured le systems compression on the y Soeware complexity dominates hardware 4 27 11 Spring 2011 Lecture 27 19 RAID Administrivia Course Summary Randy Cal Culture Dave Course EvaluaTon 4 27 11 Administrivia I felt the midterm was 9 Far too di cult 43 Somewhat harder than it should have been 46 Fair 1 A liHle too easy 1 Far too easy Designed for 90 minutes you will have 3 hours Comprehensive parTcularly problem areas on midterm but focused on course since midterm lecture lab hws and projects are fair game 8 inch x 11 inch crib sheet like midterm Spring 2011 Lecture 27 21 4 27 11 Some Survey Results Spring 2011 Lecture 27 22 Some Survey Results How much Tme per week to you spend on average in 61C including lecture discussion and labs 18 10 hours per week 26 11 12 hours per week Berkeley guidelines 3 hours unit 13 13 14 hours per week EECS classes oeen guidelines 24 15 16 hours per week 10 17 20 hours per week 10 21 hours per week Rated as Enjoyed and learned a lot Project 4 Processor Design in Logisim Part 2 73 Part 1 60 Logisim Labs 58 C memory management lab 55 TLP and DLP Labs 52 Project 3 Matrix MulTply Performance Improvement Part 2 36 Part 1 35 4 27 11 4 27 11 Spring 2011 Lecture 27 20 Some Survey Results Final Review Mon 5 2 5 8PM 2050 VLSB Final Exam Mon 5 9 11 30 2 30PM 100 Haas Pavilion 4 27 11 Spring 2011 Lecture 27 23 Spring 2011 Lecture 27 24 4 4 …
View Full Document
Unlocking...