EECS 582 Midterm Review Mosharaf Chowdhury EECS 582

EECS 582 Midterm Review Mosharaf Chowdhury EECS 582 – W 16 1

Stats on the 18 Reviewers 2 3 13 Score 3. 5 4. 42 4. 35 4. 29 4. 25 4. 14 4. 05 4 3. 95 3. 9 3. 86 4 3. 48 4. 5 3. 86 5 4. 25 Average Score By Reviewers 4. 24 Number of Reviews by Reviewers 3 2. 5 2 1. 5 1 0. 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Reviews 20 Reviews Reviewers Sorted By Score 21 Reviews EECS 582 – W 16 2

Stats on the 21 Papers We’ve Reviewed Average Score of Papers 5 4 Score 3. 5 3 2. 5 2 1. 5 1 0. 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Papers Sorted by Score EECS 582 – W 16 3

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Stats on the 21 Papers We’ve Reviewed Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 4

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Classics Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 5

Classics • UNIX • (Arguably) the first commodity OS • Simplicity is king when you must support diverse applications • Everything is a file! • System R • The first relational database implementation • Design iteratively and be ready to throw away • Find the “right” metric and do everything when you must deliver performance (i. e. , specialization instead of generalization) EECS 582 – W 16 6

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Storage and File Systems Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 7

Storage and File Systems • RAID • Industry standard for durable, high-performance storage • Parallelize for performance and fault-tolerance • FFS • Improved UNIX’s default file systems • Be aware of and exploit hardware characteristics • JFS • Provides crash recovery to file systems • Log what you’ll do before you do it! EECS 582 – W 16 8

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Kernels Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 9

Kernels • Exokernel • Minimal kernel instead of a full-fledged one • End-to-end argument: only provide services that everyone needs to balance between specialization (performance) and generalization (applicability) • Multikernel • Make communication explicit when you must communicate • Shared-nothing design EECS 582 – W 16 10

Kernels • IX • Separate control and data planes to provide I/O performance • Kernels aren’t inherently slow; it’s about how we do things • Commuter • Interfaces dictate scalable design EECS 582 – W 16 11

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Virtual Memory and RPC Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 12

Virtual Memory and RPC • Memory Coherence • Shared-everything design requires frequent updates • Keeping things coherent is expensive but provides simple programming models • RPC • Makes distributed nature more explicit while keeping the same programming model as a non-distributed system • No shared memory EECS 582 – W 16 13

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Concurrency and Scheduling Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 14

Concurrency and Scheduling • SEDA and Fibers • Thread- and event-based programming models both have their advantages and drawbacks (ease of programming vs. scalability and performance) • It is possible to find a balance between the two • Lottery and Stride Scheduling • Randomized and deterministic proportional scheduling • The key challenge is in determining the proportions (weights) EECS 582 – W 16 15

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Reliability and Fault Tolerance Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 16

Reliability and Fault Tolerance • Eraser • Detecting bugs, specially the non-deterministic ones, is hard • Provide a “tight” coverage to allow manual inspection • Nooks • Failure is inevitable • Isolate it and start again EECS 582 – W 16 17

M em EECS 582 – W 16 rt e. V i R X ES tio n Xe n ra ig M er ks oo N Er as de St ri ry tte Lo rs be Fi A D SE PC R en ce te r IX l ne ut m oh er C e Li v or y om C S ne l er er tik ul M JF S FF D AI R R IX em st ok Ex Sy N U Score Virtual Machines Average Score of Papers 5 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 18

Virtual Machines • Xen and ESX • Virtualization comes in many shapes and forms (e. g. , full virtualization vs paravirtualization) • Choose the one that fits your requirements (e. g. , performance, consolidation, deployability) • Fit for your workload • Live Migration • Keep things running until you must stop • Re. Virt • Quis custodiet ipsos custodes? (Roman poet Juvenal in Satires) • It’s turtles all the way down! EECS 582 – W 16 19