08 Tornado

Information about 08 Tornado

Published on October 5, 2007

Author: Sudiksha

Source: authorstream.com

Content

Tornado: Minimizing Locality and Concurrency in a SMP OS:  Tornado: Minimizing Locality and Concurrency in a SMP OS Slide2:  http://www.eecg.toronto.edu/~okrieg/tmp.pdf Why locality matters::  Why locality matters: Faster processors and more complex controllers -> higher memory latencies Write sharing costs Large secondary caches Large cache lines -> false sharing NUMA effects … Goal:  Goal Minimize read/write and write sharing; -> minimize cache coherence overheads Minimize false sharing Minimize distance between accessing processor and target memory module Do real systems do this?:  Do real systems do this? Yes and no Tornado -> adopt design principles to maximize locality and concurrency Map locality and independency which exists in the OS requests from applications into locality and independence in servicing these requests in the kernel or system servers Approach – re-think who data structures are organized and how operations on them are applied Counter ilustration:  Counter ilustration Shared counter, array counter, padded counter Tornado basics:  Tornado basics Individual resources in individual objects Mechanisms: Clustering objects Protected procedure calls Semi-automatic garbage collection / efficient locking Clustered objects:  Clustered objects Appear as a single object Multiple “reps” assigned to handle object references from one (or more) processors Object = granularity of access Operations, synchronization can be applied only to relevant pieces Will make global policies more difficult (e.g., global paging policy) Implementation should reflect object use Cluster Objects Implementation:  Cluster Objects Implementation Mix of replication and partitioning techniques: Process Obj replicated, Regions distributed and created on demand… Combination of object migration, home rep, and other techniques (think distributed shared memory…) Translation tables to handle implementation Per processor to access local reps Global partitioned table across processors to find rep for given object Default “miss” handler May be quite large, but sparse -> let caching mechanisms help keep around only relevant pieces… Dynamic Memory Allocation:  Dynamic Memory Allocation Local allocation – per “node” For small, less than cache-line data, use separate pool Addresses false sharing issue Avoid interrupt disabling by using efficient locks Protected procedure calls:  Protected procedure calls Jumps into address space of a (server) object Microkernel design Client requests serviced on local processors (translation table) Handoff scheduling # server threads == # client threads Stub generator to generate code based on public interface Reference checking Special MetaPort to handle first use of a PPC Parameter passing Mix of registers, mapped stack or memory regions Cross-processor IPC Optimize so that caller spins in trap Synchronization:  Synchronization They separate locking (for updates) & existence guarantees (deallocations) Encapsulate lock within object (better rep), avoid global locks Avoids contention, limits cache coherence operations on lock access Use spin-then-block locks Garbage Collection:  Garbage Collection Essentially RCU Must ensure all persistent and temporary object references are removed Object/rep keeps track of requests made out to it – counter decremented on completion – so when counter is zero no temp references Since first use of object goes through translation table, can determine which processors have object reps, and can use a token scheme to ensure object ref counter is zero for each processor Finally – safe to dealloc object Evaluation:  Evaluation Use of NUMAchine and simulator NUMAchine – ring of 4 stations, each with 4 processors and a memory module, direct mapped caches Simulator different interconnect and cache coherence protocol First validate simulator is OK then use simulator to gather other data:  First validate simulator is OK then use simulator to gather other data Effects of cluster objects:  Effects of cluster objects Page faults frequent, region deletions aren’t Slide17:  NUMAchine, SimOS and SimOS w/ 4-way assoc cache Compared to other arch/OS, MT and MP mode:  Compared to other arch/OS, MT and MP mode MT MP pagefault fstat thread

Related presentations


Other presentations created by Sudiksha

3 Theodore Roosevelt
22. 10. 2007
0 views

3 Theodore Roosevelt

ramasetu24june200747 69
30. 09. 2007
0 views

ramasetu24june200747 69

ACEI New Orleans 2004
05. 10. 2007
0 views

ACEI New Orleans 2004

Breaking Bad News May 07 ASA
08. 10. 2007
0 views

Breaking Bad News May 07 ASA

DESIGNING A TEMPERATURE SENSOR
12. 10. 2007
0 views

DESIGNING A TEMPERATURE SENSOR

blackhole
07. 10. 2007
0 views

blackhole

quiz
10. 12. 2007
0 views

quiz

BSRUN
19. 10. 2007
0 views

BSRUN

E consultancy slides march 6th
25. 10. 2007
0 views

E consultancy slides march 6th

breakinggridlock0612 01
30. 10. 2007
0 views

breakinggridlock0612 01

EXPLORERS
01. 11. 2007
0 views

EXPLORERS

Japanl
09. 10. 2007
0 views

Japanl

oct16 gfbiedu
25. 10. 2007
0 views

oct16 gfbiedu

dbirday croft
16. 11. 2007
0 views

dbirday croft

KSA V5
23. 11. 2007
0 views

KSA V5

DavidShipman
04. 10. 2007
0 views

DavidShipman

Significance
25. 10. 2007
0 views

Significance

NASA
03. 01. 2008
0 views

NASA

Day 3 Charlotte DUFOUR TIPS
04. 12. 2007
0 views

Day 3 Charlotte DUFOUR TIPS

colangelo
07. 01. 2008
0 views

colangelo

pisa overview
17. 10. 2007
0 views

pisa overview

NG21A 07 Rundle
30. 10. 2007
0 views

NG21A 07 Rundle

McCarthypix
02. 11. 2007
0 views

McCarthypix

frital
24. 10. 2007
0 views

frital

Ammosov RPC IHEP
12. 10. 2007
0 views

Ammosov RPC IHEP

P4 2 Kawagoe
15. 10. 2007
0 views

P4 2 Kawagoe

PE Minerals
16. 02. 2008
0 views

PE Minerals

Internal Analysis Lecture
24. 02. 2008
0 views

Internal Analysis Lecture

OSHAtop102006
26. 02. 2008
0 views

OSHAtop102006

T E of the Machine Gun
27. 02. 2008
0 views

T E of the Machine Gun

secretcodes tcm4 336597
31. 12. 2007
0 views

secretcodes tcm4 336597

1960s
20. 02. 2008
0 views

1960s

Porteous
12. 03. 2008
0 views

Porteous

AG 2002 11 16
24. 10. 2007
0 views

AG 2002 11 16

Md  Ppt
24. 03. 2008
0 views

Md Ppt

science genetics
03. 10. 2007
0 views

science genetics

Grade 9 Heat
03. 04. 2008
0 views

Grade 9 Heat

Final Prelims 2006
16. 04. 2008
0 views

Final Prelims 2006

FOP01 Franchise Opportunity
17. 04. 2008
0 views

FOP01 Franchise Opportunity

pres4
18. 04. 2008
0 views

pres4

LSE
22. 04. 2008
0 views

LSE

B1
30. 10. 2007
0 views

B1

OPC Notes CT
07. 05. 2008
0 views

OPC Notes CT

chris corrigan pres
30. 04. 2008
0 views

chris corrigan pres

Facility Layout Lecture Notes
02. 05. 2008
0 views

Facility Layout Lecture Notes

Sukal Linger Presentation
08. 10. 2007
0 views

Sukal Linger Presentation

access programme 2004
17. 10. 2007
0 views

access programme 2004

SuperValu Presentation2
02. 10. 2007
0 views

SuperValu Presentation2

Larsen
07. 03. 2008
0 views

Larsen

cbm39 269
14. 04. 2008
0 views

cbm39 269

Cuban
23. 12. 2007
0 views

Cuban

ICN2001 Final Report for web
20. 03. 2008
0 views

ICN2001 Final Report for web

map ftaa windows xp
22. 10. 2007
0 views

map ftaa windows xp

The Motorola Phone Comedy
17. 10. 2007
0 views

The Motorola Phone Comedy

PresSchatanIntrod
22. 10. 2007
0 views

PresSchatanIntrod

ERCIMgridasiaRR
16. 10. 2007
0 views

ERCIMgridasiaRR

bpesp
23. 10. 2007
0 views

bpesp

Florida Congress 6 06
22. 10. 2007
0 views

Florida Congress 6 06

WrinkleInTime
24. 10. 2007
0 views

WrinkleInTime

WNVEnterpriseGIS Chicago
21. 10. 2007
0 views

WNVEnterpriseGIS Chicago

pira ing
04. 10. 2007
0 views

pira ing

SolarHeliosphere
11. 03. 2008
0 views

SolarHeliosphere

04ift tomatosalsaPoster combined
04. 03. 2008
0 views

04ift tomatosalsaPoster combined

InstallingPortlets
05. 10. 2007
0 views

InstallingPortlets