2017 · Concurrency and Computation: Practice and Experience
Multi-scale Navigation of Large Trace Data: A Survey
Evidence basis: full-text-reviewed · Review status: catalog-reviewed; paper-author approval pending
trace-visualization trace-abstraction kernel-tracing performance-analysis system-tracing
trace navigation multi-scale analysis trace visualization content abstraction metric abstraction visual abstraction resource abstraction semantic zoom focus-plus-context Trace Compass Vampir Jumpshot SLOG R-tree quadtree State History Tree
Core contribution: The survey provides a taxonomy and requirements-oriented comparison of techniques for collecting, abstracting, analyzing, visualizing, and navigating large execution traces.
Catalog abstract summary
The survey organizes abstraction and visualization techniques that help analysts navigate long operating-system and kernel traces from overview to event-level evidence.
Source: Institutional accepted-version PDF reviewed; abstract paraphrased for this catalog.
Problem and motivation
Long-running and parallel executions produce traces too large for a single-resolution display; analysts need overview, semantic navigation, and evidence-preserving drill-down.
Method and contribution
Survey trace collection, maintenance, analysis, and visualization through four abstraction families and compare hierarchical, semantic, visual, and resource-oriented navigation mechanisms and data structures.
Findings and evidence
The literature offers complementary techniques rather than one universally dominant abstraction. Linking overview levels to concrete events, states, resources, and metrics is central to usable large-trace analysis.
Limitations and future directions
Limitations: This paper synthesizes prior work and does not validate a new method, dataset, or benchmark. Its coverage and taxonomy are bounded by the selected literature and the paper's OS/kernel-oriented scope.
Future work: Improve bidirectional links between abstraction levels, support issue-to-event drill-down, and advance trace models, data structures, and interactive visualization.
Sources and identifiers
- Published version published
- Institutional full text source_record
When to cite this paper
Cite this paper when your work uses or compares the four-part taxonomy of content/data, metric, visual, and resource abstraction in large-trace navigation.
- For the four-part taxonomy of content/data, metric, visual, and resource abstraction in large-trace navigation.
- For a survey-backed comparison of semantic zoom, focus+context, hierarchical views, and linked drill-down.
- For positioning SHT, SLOG, R-trees, and quadtrees as alternative trace-navigation/data-management structures.
Citation
@article{ezzatiJivan2017multiscale,
author = {Naser Ezzati-Jivan and Michel R. Dagenais},
title = {Multi-scale Navigation of Large Trace Data: A Survey},
year = {2017},
journal = {Concurrency and Computation: Practice and Experience},
volume = {29},
number = {10},
eid = {e4068},
publisher = {Wiley},
issn = {1532-0626, 1532-0634},
doi = {10.1002/cpe.4068},
url = {https://doi.org/10.1002/cpe.4068}
}Other citation formats for Word and reference managers
Ezzati-Jivan, N., & Dagenais, M. R. (2017). Multi-scale Navigation of Large Trace Data: A Survey. Concurrency and Computation: Practice and Experience, 29(10), e4068. https://doi.org/10.1002/cpe.4068N. Ezzati-Jivan and M. R. Dagenais, "Multi-scale Navigation of Large Trace Data: A Survey," Concurrency and Computation: Practice and Experience, vol. 29, no. 10, 2017, doi: 10.1002/cpe.4068