using visualization to understand the behavior of computer systems

48
Using Visualization to Understand the Behavior of Computer Systems Robert P. Bosch Jr. Stanford University May 3, 2001

Upload: kathy

Post on 16-Jan-2016

26 views

Category:

Documents


0 download

DESCRIPTION

Using Visualization to Understand the Behavior of Computer Systems. Robert P. Bosch Jr. Stanford University May 3, 2001. Motivation. Explosion in complexity of computer systems Development of rich data collection tools Complete Machine Simulation: SimOS - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Using Visualization to Understand the Behavior of Computer Systems

Using Visualization to Understand

the Behavior of Computer Systems

Robert P. Bosch Jr.Stanford University

May 3, 2001

Page 2: Using Visualization to Understand the Behavior of Computer Systems

2

Motivation

• Explosion in complexity of computer systems

• Development of rich data collection tools– Complete Machine Simulation: SimOS– Software Monitoring: Performance Co-Pilot

(PCP)– Firmware Instrumentation: FlashPoint– Hardware Monitoring: DCPI

• Challenge: how do we fully exploit the large, detailed data sets these tools can generate?

Page 3: Using Visualization to Understand the Behavior of Computer Systems

3

Data Analysis Challenge

• How do we typically handle this data?– Visual inspection of huge log files– Summarize through statistics and aggregation– Focus on restricted data subsets

• Alternative approach: data visualization– Display large amounts of data at once– Enable interactive exploration of entire data set

• Overview, zoom and filter, details-on-demand

– Use human perception to discover patterns, trends, and interesting information

Page 4: Using Visualization to Understand the Behavior of Computer Systems

4

Computer Systems Visualization:Existing Work

• Pedagogical examples– Processors: DLXView, Pentium Pro Tutorial– Memory: Cache Visualization Tool (CVT)

• Parallel systems performance– AIMS, Pablo, Paradyn, ParaGraph, PARvis,

StormWatch, VAMPIR…

• Other examples– Network performance, file systems, etc.

Page 5: Using Visualization to Understand the Behavior of Computer Systems

5

Computer Systems Visualization:Existing Work

• Demonstrate potential of visualization• Limitations

– Focused on particular system components– Integrated with specific data collection tools– Limited to a fixed set of visual

representations

• Conclusion: rich, flexible data sources require an equally powerful and flexible visualization system

Page 6: Using Visualization to Understand the Behavior of Computer Systems

6

Outline

• Motivation• The Rivet visualization environment

– Architecture– Implementation

• Focused visualization systems– SUIF Explorer– Thor– PipeCleaner– Visible Computer

• Ad hoc analysis and visualization• Contributions

Page 7: Using Visualization to Understand the Behavior of Computer Systems

7

Rivet Architecture: Goals

• Learn once, apply to wide range of problems– Decouple visualization and data collection

• Interactive exploration of large data sets– Couple visualization and analysis

• Rapid prototyping of visualizations• Extensibility

– Allow users to add new components

Page 8: Using Visualization to Understand the Behavior of Computer Systems

8

Rivet Architecture: Approach

• Identify visualization “building blocks”• Mix and match to create visualizations• Users can add new components as

needed• Three basic object types:

– Data management: tuples, tables, and transforms

– Visual representation: primitives and metaphors

– Mapping from data to visual: encodings

Page 9: Using Visualization to Understand the Behavior of Computer Systems

9

Data Model

• Simplified relational model– Tuple: collection of data fields– Table: set of tuples with common data format– Familiar, homogeneous model

• Load data by parsing text files• Directly save/load tables in binary format

Procedure:PID:Page Faults:

Redraw()1717

129...

Tuple Tabl

e

Page 10: Using Visualization to Understand the Behavior of Computer Systems

10

.

.

.

Data Transforms

• Transforms enable users to operate on data• Can be composed to form data networks

– Active: data changes are propagated

• Rivet includes a set of standard transforms– Filter, sort, group, merge, join, aggregate, etc.

• Users may design and incorporate their own– Example: clustering algorithms

.

.

.......

GroupBy

Transform

Page 11: Using Visualization to Understand the Behavior of Computer Systems

11

Visual Objects and Data Mapping

• Primitives draw individual tuples• Metaphors draw entire tables

– Draws table attributes: axes, labels, etc.– For each tuple: compute bounding box, select primitive

• Encodings map tuple contents to visual properties– Metaphors use spatial encodings

• Map tuple fields to bounding box parameters

– Primitives use attribute encodings• Map tuple fields to retinal properties: color, fill pattern,

size…

• Encodings encapsulate datavisual mapping– Metaphors and primitives are data independent

Page 12: Using Visualization to Understand the Behavior of Computer Systems

12

Display

The Encoding Process: Example

X

Y

Graph Metaphor:

Spatial EncodingsProcedure:

PID:Page Faults:

Redraw()1717

129

Tuple

C

F

S

Rectangle Primitive: Attribute Encodings

0.5

Page 13: Using Visualization to Understand the Behavior of Computer Systems

13

Coordination and Interaction

• Implicit coordination: sharing of objects– Examples:

• Primitives share attribute encodings: brushing• Metaphors share primitives: common appearance• Metaphors share spatial encodings: common axes

– Listener mechanism

• Explicit coordination: events and bindings– Rivet objects can raise events– User can bind actions to these events– Example: details-on-demand

Page 14: Using Visualization to Understand the Behavior of Computer Systems

14

Rivet Implementation

• Goal: balance performance and flexibility– Interactive visualizations of large data sets– Rapid prototyping of visualizations

• C++ and OpenGL for performance– Also provides platform independence

• Scripting language interfaces for flexibility– Simplified Wrapper and Interface Generator (SWIG)– Create visualizations by writing scripts

• Interpreter is not in the main loop– Listener mechanism for high-frequency events– Event bindings for low-frequency user interaction

Page 15: Using Visualization to Understand the Behavior of Computer Systems

15

Sample Rivet Scriptset table [DataVector]set parser [CSVParser -args $table]rparse $parser poletops.csv Import data into

tableCreate point primitive

set primitive [GLPoint]

Create graph metaphor

rwindow .radiosrgeometry .radios W 500 H 675rglob Graph .radios.map.radios.map SetData $table

Encode Longitude as X

set gran [expr 1.0 / 3600.0]set lmin [$table GetMin Longitude]set lmax [$table GetMax Longitude]set long [QUniformRangeMap –args $lmin $lmax $gran].radios.map EncodeAsXPosition \ [QRangeEncoding -args $long Longitude]

Encode # Radios as Color

set hue [list 0.0 0.5]set sat [list 0.0 1.0]set val [list 0.6 1.0]set ramp [IsomorphicColorMap –args $hue $sat $val]set rmin [$table GetMin Radios]set rmax [$table GetMax Radios]set radios [QNumberMap -args $Log $rmin $rmax]$ramp SetDomainMap $radios$primitive EncodeAsColor \ [QColorEncoding -args $ramp Radios]

Encode Latitude as Y

set lmin [$table GetMin Latitude]set lmax [$table GetMax Latitude]set lat [QUniformRangeMap –args $lmin $lmax $gran].radios.map EncodeAsYPosition \ [QRangeEncoding -args $lat Latitude]

Page 16: Using Visualization to Understand the Behavior of Computer Systems

16

Outline

• Motivation• The Rivet visualization environment

– Architecture– Implementation

• Focused visualization systems– SUIF Explorer: Interactive user-directed

parallelization– Thor: Detailed memory profiling on FLASH– PipeCleaner: Superscalar processor pipelines– Visible Computer: System and cluster monitoring

• Ad hoc analysis and visualization• Contributions

Page 17: Using Visualization to Understand the Behavior of Computer Systems

17

SUIF Explorer:Interactive Parallelization

Page 18: Using Visualization to Understand the Behavior of Computer Systems

18

SUIF Explorer: Background

• Goal: enable sequential codes to run fast on MPs

• Compiler parallelizes loops when possible• Otherwise:

– Compiler presents loop analysis results to user– User applies application knowledge to assist compiler

• Data source: SUIF dynamic analyzers– Performance metrics for all loops in program– Coverage, granularity, loop level

• Visualization– Displays data in context of program source code– Focuses on loops most deserving of user attention

Page 19: Using Visualization to Understand the Behavior of Computer Systems

19

SUIF Explorer: Discussion

• Benefits– Combines data with source, instead of loop

IDs– Allows user to filter uninteresting lines of

code– Facilitates comparisons between runs

• Limitations– Loosely coupled with rest of SUIF Explorer– Short loops less prominent than long loops

• Add sortable table of results as a linked view

Page 20: Using Visualization to Understand the Behavior of Computer Systems

20

Thor: Detailed Memory Profiling

Page 21: Using Visualization to Understand the Behavior of Computer Systems

21

Thor: Background

• Memory getting slower relative to CPU• High remote access latencies on NUMA systems• Memory system is often performance bottleneck• Data source: FlashPoint protocol on FLASH

– Collects all cache and TLB misses in firmware– Classified as local/remote, read/write– Attributed to CPU, procedure, data structure

• Visualization– Stacked bar charts showing miss counts– Collection of UI controls for configuring the display

Page 22: Using Visualization to Understand the Behavior of Computer Systems

22

Thor: Discussion

• Both post-mortem and real-time analysis– Live connection to FLASH via socket– Useful for multi-phase applications

• Simple, familiar visualization• Interactive filtering, sorting,

aggregation• Drill down from overview to data of

interest

Page 23: Using Visualization to Understand the Behavior of Computer Systems

23

PipeCleaner: Superscalar Processors

Page 24: Using Visualization to Understand the Behavior of Computer Systems

24

PipeCleaner: Background

• High peak performance• Complex implementation techniques

– Speculation, multiple functional units, out-of-order execution

• Intended to be transparent to the programmer– True for correctness, not necessarily for performance

• Data sources: MXS and MMIX simulators– Detailed superscalar processor pipeline models

• Visualization: three linked views– Overview: occupancy strip charts– Detail: animated pipeline display– Context: program source code

Page 25: Using Visualization to Understand the Behavior of Computer Systems

25

PipeCleaner: Discussion

• Applications of PipeCleaner– Program development– Compiler optimizations– Hardware design– Education – Simulator development

• Possible extensions– Other computer pipelines: graphics

hardware– Physical pipelines: assembly lines, etc.

Page 26: Using Visualization to Understand the Behavior of Computer Systems

26

Visible Computer:System and Cluster Monitoring

Page 27: Using Visualization to Understand the Behavior of Computer Systems

27

Visible Computer: Background

• Real-time analysis of system behavior– Observe behavior of system in its entirety– Explore interesting phenomena in detail

• Data sources: PCP, SimOS– Comprehensive data sources– Collect low-level hardware events

• Cache misses, disk requests, CPU utilization, etc.

– Classify using high-level structures• Process name, user/group, CPU mode, etc.

Page 28: Using Visualization to Understand the Behavior of Computer Systems

28

Visible Computer: Visualization

• Organizes data using nested physical hierarchy

• Provides overviews at each level of detail– Active icons representing system components– User defines data ranges of interest for each

component– Icons activate when components are in range

• Allows users to “remove the cover”, show next level

• Displays detailed charts on demand– Data filtered/colored using high-level classifiers

Page 29: Using Visualization to Understand the Behavior of Computer Systems

29

Visible Computer: Discussion

• Unified interface for computer systems data– Focus-plus-context view of system– Physical layout, virtual classification

• Provides an overview of system behavior

• Draws attention to potential problem areas

• Suggests targets for further exploration

Page 30: Using Visualization to Understand the Behavior of Computer Systems

30

Outline

• Motivation• The Rivet visualization environment

– Architecture– Implementation

• Focused visualization systems– SUIF Explorer– Thor– PipeCleaner– Visible Computer

• Ad hoc analysis and visualization• Contributions

Page 31: Using Visualization to Understand the Behavior of Computer Systems

31

Ad hoc Analysis and Visualization

• SimOS: Complete machine simulator– Full non-intrusive access to HW, OS, SW

state– Flexible data collection mechanism– Deterministic execution

• Combine with Rivet– Flexible data import mechanism– Rapid prototyping of visualizations

• Result: powerful analysis framework

Page 32: Using Visualization to Understand the Behavior of Computer Systems

32

Change HW model or SW

Simulation & Visualization Cycle

Configure simulated machine and

software

Perform simulation

Visualize results

Change annotations

Change visualization

Page 33: Using Visualization to Understand the Behavior of Computer Systems

33

Case Study: Argus

• Parallel, multithreaded graphics library– NURBS rendering application – Implemented using fork & shared memory

region

• Performance limitations on SGI Origin– Linear speedup up to 26 processors– Rapid performance falloff beyond 26 CPUs

• Imported into SimOS environment– Same performance characteristics observed

• Expectation: memory system is the problem

Page 34: Using Visualization to Understand the Behavior of Computer Systems

34

Memory Visualization

Page 35: Using Visualization to Understand the Behavior of Computer Systems

35

Memory Visualization

1. Histograms of memory stall time vs. physical/virtual address

Page 36: Using Visualization to Understand the Behavior of Computer Systems

36

Memory Visualization

2. Memory stall time incurred by each line of source code

Page 37: Using Visualization to Understand the Behavior of Computer Systems

37

Memory Visualization

3. Local/remote stall time and idle time for each process

Page 38: Using Visualization to Understand the Behavior of Computer Systems

38

Memory Visualization

Large amounts of unexpected idle timeVery little memory stall time in process view

Page 39: Using Visualization to Understand the Behavior of Computer Systems

39

Visualization of Process Data

Processes go idle in kernel pfault/vfault calls

Page 40: Using Visualization to Understand the Behavior of Computer Systems

40

Process Scheduling & Kernel Locks

CPU schedulingkernel lock

If the lock is unavailable, the process is descheduledIf the lock is available, it is immediately granted

Page 41: Using Visualization to Understand the Behavior of Computer Systems

41

Process Scheduling & Kernel Locks

kernel lock is held

process is descheduled

Kernel lock is heavily contended

Page 42: Using Visualization to Understand the Behavior of Computer Systems

42

Results With Processes Pinned

Lots of idle time and kernel lock contention still remain

Page 43: Using Visualization to Understand the Behavior of Computer Systems

43

Results Using sproc

• Minimal kernel time• No idle time at all• Completes nearly

twice as fast as original version

• 95% parallel efficiency

Page 44: Using Visualization to Understand the Behavior of Computer Systems

44

Argus: Conclusion

• Five simulation & visualization iterations• Initial suspicions were totally incorrect• Flexibility of Rivet & SimOS enabled us

to follow leads• Visualizations led us to the bottleneck

– Single scheduling event out of over 50,000 events

Page 45: Using Visualization to Understand the Behavior of Computer Systems

45

Contributions

• Rivet computer systems visualization environment– Rapid prototyping support

• Quick ‘rough cut’ visualization of data• Incremental development of sophisticated data displays

– Modular architecture• Mix and match basic building blocks to create visualizations• Enables coordination through object sharing

– Data transforms as first-class citizens in viz environment• Unifies the analysis and visualization process

• Collection of computer systems visualizations– Emphasize interactive data exploration– Use relatively conventional visual representations– Provide extensive support for filtering, sorting, brushing

Page 46: Using Visualization to Understand the Behavior of Computer Systems

46

Acknowledgments

• Orals committee: Joel Ferziger, Mendel Rosenblum, Pat Hanrahan, Mark Horowitz, Monica Lam

• Mendel• Visualization collaborators

– PipeCleaner: Donald Knuth– Visible Computer: John Gerth– Argus: Gordon Stoll– Thor: Jeff Gibson– SUIF Explorer: Shih-Wei Liao

Page 47: Using Visualization to Understand the Behavior of Computer Systems

47

Acknowledgments

• Groupmates: Rivet, SimOS, FLASH, Graphics

• Studio 354– Steve Herrod and John Heinlein– The foosball table– Chris Stolte and Diane Tang

• Staff: – John, Charlie, Thoi, Kevin– Ada, Heather, Chris, …

• Funding agencies: ONR, [D?]ARPA, ASCI• Robert Bosch Corporation

Page 48: Using Visualization to Understand the Behavior of Computer Systems

48

Acknowledgments

• Rains 7A: Steve, Hoa, Marco• Friends• Family• Ming