{ "schema_version": "0.6.0", "paper_id": "b-perf-black-box-performance-antipatterns", "page_url": "https://naser.github.io/research-publications/papers/b-perf-black-box-performance-antipatterns/", "title": "B-Perf: Black-box Performance Antipattern Detection Using System-level Execution Tracing", "title_variants": [], "authors": [ "Morteza Noferesti", "Mahsa Panahandeh", "Naser Ezzati-Jivan" ], "author_details": [ { "name": "Morteza Noferesti", "orcid": "https://orcid.org/0009-0000-5507-1461", "profile_url": "https://dblp.org/pid/82/10462.html" }, { "name": "Mahsa Panahandeh", "orcid": "https://orcid.org/0000-0002-6369-8982", "profile_url": "https://dblp.org/pid/201/1655.html" }, { "name": "Naser Ezzati-Jivan", "orcid": "https://orcid.org/0000-0003-1435-6297", "profile_url": "https://naser.github.io/" } ], "publication": { "year": 2026, "venue": "ACM/SPEC International Conference on Performance Engineering (ICPE)", "type": "conference paper", "publication_date": "2026-05-03", "pages": "96-107", "publisher": "ACM", "isbn": [ "979-8-4007-2325-4" ], "online_date": "2026-05-03", "print_date": "2026-05-04", "volume": null, "issue": null, "article_number": null, "issn": [], "crossref_type": "proceedings-article" }, "publication_type": "conference paper", "status": "published_with_public_full_text", "canonical_source_url": "https://doi.org/10.1145/3777884.3797014", "source_record_id": "b-perf-black-box-performance-antipattern-detection-using-system-level-execution-tracing-acfda3a610", "identifiers": { "doi": "10.1145/3777884.3797014" }, "abstract": "Performance antipatterns capture recurring behaviours that degrade software efficiency. Black-box approaches aim to detect such issues without modifying the application. This paper presents B-Perf, a system-level black-box method that reconstructs execution, memory, and messaging behaviour from kernel-level traces. By analysing scheduling, allocation, and communication events, B-Perf derives workload-dependent behavioural trends and reports antipattern indicators grounded in resource usage and contention. To handle large trace volumes, the approach follows a pipeline of workload generation, event gathering, trace handling, and antipattern inference. We evaluate B-Perf on three representative antipatterns—One Lane Bridge, Empty Semi Trucks, and Excessive Dynamic Allocation—and apply it to traces from real multi-threaded applications. The results show that system-level events are often sufficient to expose bottlenecks linked to resource contention and system-level interactions. A key limitation is that kernel traces provide limited visibility into fine-grained in-process behaviour. When performance issues are driven by internal logic or function-level interactions, B-Perf may capture only indirect symptoms and may not reveal the full root cause. Within this scope, B-Perf provides practical and efficient black-box detection for antipatterns driven by resource interaction and competition.", "abstract_source": "Author abstract from the official ICPE 2026 preprint, licensed CC BY 4.0; verified against the locally extracted full text.", "abstract_available": true, "scholar_eligibility": { "eligible": true, "basis": "complete-visible-author-abstract", "note": "The page displays the complete verified author abstract." }, "description": "Performance antipatterns capture recurring behaviours that degrade software efficiency. Black-box approaches aim to detect such issues without modifying the application. This paper presents B-Perf, a system-level black-box method that reconstructs execution, memory, and messaging behaviour from kernel-level traces. By analysing scheduling, allocation, and communication events, B-Perf derives workload-dependent behavioural trends and reports antipattern indicators grounded in resource usage and contention. To handle large trace volumes, the approach follows a pipeline of workload generation, event gathering, trace handling, and antipattern inference. We evaluate B-Perf on three representative antipatterns—One Lane Bridge, Empty Semi Trucks, and Excessive Dynamic Allocation—and apply it to traces from real multi-threaded applications. The results show that system-level events are often sufficient to expose bottlenecks linked to resource contention and system-level interactions. A key limitation is that kernel traces provide limited visibility into fine-grained in-process behaviour. When performance issues are driven by internal logic or function-level interactions, B-Perf may capture only indirect symptoms and may not reveal the full root cause. Within this scope, B-Perf provides practical and efficient black-box detection for antipatterns driven by resource interaction and competition.", "evidence_level": "full-text-reviewed", "evidence": { "source_basis": "full-text-reviewed", "coverage": "material paper sections", "summary_origin": "AI-assisted catalog editorial summary", "review_status": "catalog-reviewed; paper-author approval pending", "verified_on": "2026-08-09", "sources": [ { "note": "Private full PDF read: pdf-evidence/notes/b-perf-black-box-performance-antipatterns.md" }, { "note": "B-Perf PDF pp. 4-7: LTTng event model, Trace Compass-style state reconstruction, tracepoints, environment, and workload protocol" }, { "note": "B-Perf PDF pp. 7-10: One Lane Bridge, Excessive Dynamic Allocation, Empty Semi Trucks, Firefox/web validation, and overhead results" }, { "note": "B-Perf PDF pp. 10-11: limitations, threats to validity, and future work" }, { "note": "ICPE 2026 accepted-papers page labels the conference-hosted PDF as the B-Perf Pre-Print; downloaded PDF is byte-identical to the locally reviewed evidence (SHA-256 FE38962FD4ACF91F760AC51907207F2569B99A172CEFE8D2B8D1393B547A6003)" }, { "note": "Local PDF hash verified in pdf-evidence/extraction-manifest.json" } ] }, "summary": { "core_contribution": "B-Perf reconstructs execution, memory, and messaging behavior from Linux kernel-level traces and infers performance-antipattern indicators without requiring application source access or intrusive instrumentation.", "problem": "White-box performance-antipattern detection needs source access or application instrumentation. B-Perf asks whether system-level kernel events can expose execution, memory, and messaging antipatterns for black-box or binary-only targets.", "method": "B-Perf gathers LTTng kernel events, maps them to an abstract event model, reconstructs a Trace Compass-style state system and per-request critical paths, derives execution/memory/messaging projections, and applies trend-based inference over ordered workloads. It reports behavioral indicators rather than claiming strict classification.", "findings": "Controlled paired programs distinguish One Lane Bridge through serialized critical paths and blocking, Excessive Dynamic Allocation through allocation/free churn and unstable memory, and Empty Semi Trucks through many small TCP messages, interrupts, and preemption. A Firefox case rises from roughly 30 seconds for one parallel window to 2.2 minutes for two and 5.7 minutes for three; the method flags execution serialization. CPU tracing overhead is below 0.01%, while the I/O microbenchmark loses about 7-8% throughput.", "limitations": "Kernel traces expose resource interaction but not fine-grained internal logic, algorithms, or purely application-level lock causes. The controlled programs isolate one antipattern at a time, the external validation is limited, the method is Linux/LTTng-focused, and the current pipeline is post-mortem. Indicators can be ambiguous under mixed behavior and background interference.", "future_work": "Validate larger industrial and multi-component systems; support online or near-real-time incremental state reconstruction and reporting; test other operating systems and tracing backends; and refine inference for overlapping antipatterns and mixed workloads." }, "tags": [ "kernel-tracing", "system-tracing", "performance-analysis", "anomaly-detection" ], "keywords": [ "performance antipatterns", "black-box detection", "system-level execution traces", "LTTng", "Trace Compass", "critical paths", "resource contention", "execution serialization", "allocation churn", "small-message overhead" ], "versions": [ { "id": "published-version", "label": "Published version", "relation": "version-of-record", "title": "B-Perf: Black-box Performance Antipattern Detection Using System-level Execution Tracing", "url": "https://doi.org/10.1145/3777884.3797014", "pdf_url": null, "status": "published", "canonical_for_citation": true }, { "id": "official-icpe-preprint", "label": "Official ICPE preprint", "relation": "preprint", "title": "B-Perf: Black-box Performance Antipattern Detection Using System-level Execution Tracing", "url": "https://icpe2026.spec.org/preprint/Black-box_Performance_Antipattern_Detection_Using_System-level_Execution_Tracing.pdf", "pdf_url": "https://icpe2026.spec.org/preprint/Black-box_Performance_Antipattern_Detection_Using_System-level_Execution_Tracing.pdf", "status": "public_full_text", "canonical_for_citation": false } ], "access": { "status": "published_with_public_full_text", "note": "The DOI is the canonical citation target. The external source PDF is linked for readers; no PDF is redistributed in this repository.", "license": "CC BY 4.0" }, "resources": { "code": null, "data": null, "slides": null, "demo": null }, "citation_guidance": { "when_to_cite": "Cite this paper when your work uses or compares black-box detection of execution, memory, and messaging performance-antipattern indicators from kernel-level traces.", "points": [ "black-box detection of execution, memory, and messaging performance-antipattern indicators from kernel-level traces.", "the shared state-system representation that turns scheduling, synchronization, allocation, TCP, interrupt, and preemption events into cross-scope behavior projections.", "the concrete LTTng evaluation of One Lane Bridge, Excessive Dynamic Allocation, and Empty Semi Trucks, including the low CPU-bound tracing overhead and I/O throughput trade-off.", "the distinction between behavioral indication and definitive root-cause classification when kernel evidence is indirect or mixed." ], "canonical_version_id": "published-version" }, "provenance": { "metadata_verified_on": "2026-08-09", "metadata_source": [ "Private full PDF read: pdf-evidence/notes/b-perf-black-box-performance-antipatterns.md", "B-Perf PDF pp. 4-7: LTTng event model, Trace Compass-style state reconstruction, tracepoints, environment, and workload protocol", "B-Perf PDF pp. 7-10: One Lane Bridge, Excessive Dynamic Allocation, Empty Semi Trucks, Firefox/web validation, and overhead results", "B-Perf PDF pp. 10-11: limitations, threats to validity, and future work", "ICPE 2026 accepted-papers page labels the conference-hosted PDF as the B-Perf Pre-Print; downloaded PDF is byte-identical to the locally reviewed evidence (SHA-256 FE38962FD4ACF91F760AC51907207F2569B99A172CEFE8D2B8D1393B547A6003)", "Local PDF hash verified in pdf-evidence/extraction-manifest.json" ], "summary_written_by": "AI-assisted", "summary_verified_by": "full-text-grounded catalog review; author approval pending", "linked_preprint_record": null, "author_order_note": null }, "batch": { "phase": 2, "batch_label": "expanded forty-paper release", "status": "included_in_expanded_catalog", "selected_at": "2026-08-09" } }