Toward Operationalizing Rasmussen in Evolving Software Systems
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Robotics Radio. Generated commentary on the latest robotics and control papers.
Rosa: We're looking at the paper: "Toward Operationalizing Rasmussen in Evolving Software Systems".
Dev: Here's the gist of the paper. Software operations increasingly rely on SLOs, traces, deployment specifications, and change events, yet dashboards and thresholding practices often expose share-like operational signals as separate scalar panels or baseline distances <ref:2602.05483#pg2> This paper proposes an automated design for a drift monitor that maps changing software artifacts into a stable compositional monitoring state with lineage-aware policy boundary diagnostics <ref:2602.05483#pg2> Rasmussen’s dynamic safety model frames safety as a moving target where systems operate under competing gradients and local adaptations can produce slow drift toward failure as boundaries of safe operation are approached <ref:2602.05483#pg2> Software operations is a natural target because services evolve continuously, are governed by explicit policies such as SLOs and error-budgets, and emit rich telemetry <ref:2602.05483#pg2> Teams often reason about multiobjective operational state as shares like remaining reliability margin across SLOs or engineering effort across work categories <ref:2602.05483#pg2> Applying Euclidean thresholds to such closed data can conflate safe redistribution with risk accumulation because signals form compositional views where relative structure is meaningful <ref:2602.05483#pg2> Artifact-Derived Model Discovery The monitor pipeline involves a sequence of steps where the artifact stream is mapped to a compositional state and then analyzed for drift <ref:2602.05483#pg2> The extraction operator yields a lightweight model Mt = Extract(At), which can include an SLO inventory, service graph, request-class partition, and policy-constraint set <ref:2602.05483#pg2> The process is defined by the pipeline At → Mt → Φ(Mt, At) → xt → x˜t → ILR(x˜t) → report <ref:2602.05483#pg2> The monitor requires a current part inventory, telemetry-to-part mapping, lineage metadata, policy constraints where available, and health signals to function effectively <ref:2602.05483#pg2> Simplex Geometry and Lineage Stabilization The strictly positive composition with D parts lies in the simplex SD−1 = ∪R D >0: Íi xi = 1 <ref:2602.05483#pg3> Aitchison operations define its geometry, and an isometric log-ratio (ILR) map is used as a coordinate representation of that geometry <ref:2602.05483#pg4> Raw ILR coordinates are not directly comparable when services split or SLOs are revised, necessitating a lineage map pit: 1,..., D(t) → 1,..., K from current parts to stable canonical groups <ref:2602.05483#pg4> The monitor maintains x˜t,k = ∑︁ i:pit (i)=k xt,i, x˜t ∈ SK−1 <ref:2602.05483#pg4> This lineage aggregation is an operational stabilization layer that keeps dashboard-level signals comparable under churn while retaining leaf-level subcompositions for drill-down <ref:2602.05483#pg4> Drift Diagnostics on the Simplex One-step drift in the stabilized state is modeled as x˜t+1 = x˜t ⊕ (beta ⊙ g˜t) ⊕ eta˜t <ref:2602.05483#pg5> In ILR coordinates, this becomes additive: z˜t+1 = z˜t + betau˜t + epsilont, where u˜t = ILR(g˜t) <ref:2602.05483#pg5> Only the product betau˜t is identifiable from observed changes, making the primary directional signal the smoothed direction uˆt = Δz˜t /∥Δz˜t ∥ <ref:2602.05483#pg5> A simple barrier B(x˜t) = −ÍK k=1 log x˜t,k diverges as any part approaches zero Policy boundaries are encoded as inequalities hj (x) ≤ 0, which should be represented as log-ratios for ratio constraints like F /R ≤ τ A complementary diagnostic is step-to-boundary along the smoothed drift direction to seek the smallest λ > 0 such that ILR−1(z˜t+λuˆt) ∉ Ω Instantiations and Sanity Check Results The synthetic sanity check over 300 trajectories of length 100 demonstrated that boundary-relevant drift should be separable from benign redistribution Under injected F /R drift, the boundary monitor detected all crossings with a median lead time of 13 windows Controlled split/merge churn resulted in a stable canonical signal with a median maximum error of 0 This confirms that lineage-aware aggregation preserved the canonical signal when lineage was exact The evaluation tested three falsifiable claims, including H1 that exogenous shocks induce non-zero mean drift directions in balance space Positioning and Conclusion The novelty of this work lies in reducing an evolving software architecture—renames, splits, merges, and policy revisions—to a stable compositional monitoring state with lineage-aware policy boundary diagnostics <ref:2602.05483#pg2> The proposed monitor reports drift direction, step-to-boundary, attribution via top-k balances mapped back to the extracted model, and model-health indicators such as ct and mother t This framework links safety-science drift models with artifact-derived observability for evolving systems <ref:2602.05483#pg2> The goal is to discover an operational model from artifacts, map telemetry to a compositional state, stabilize that state under churn through lineage-aware canonical groups, and report log-ratio coherent, boundary-aware drift diagnostics with model-health gating <ref:2602.05483#pg2> The paper concludes by proposing this method as an automated runtime drift observability under evolving part inventories <ref:2602.05483#pg2> References
1: John Aitchison.
Dev: The Statistical Analysis of Compositional Data.
Rosa: Chapman & Hall
2: Anatoly A.
Title and authors: Rosa: So we're looking at this paper by Anatoly Krasnovsky called "Toward Operationalizing Rasmussen in Evolving Software Systems." It’s trying to solve a problem where we have all these metrics like SLOs and traces, but the dashboards just throw all these signals into separate panels, and that can actually cause problems when things are changing.
Dev: Exactly. The core idea here is taking those shifting operational signals—things like remaining reliability margin or engineering effort—and turning them into something stable where we can actually track drift toward failure, not just noise from normal business changes.
Taro: I'm curious if this is something that actually works when the software architecture itself is constantly changing, like services splitting or getting renamed. Because that's usually where our simple thresholding breaks down.
Rosa: That’s what it tackles directly with this paper. It proposes an automated design for a drift monitor that maps those changing software artifacts into a stable compositional monitoring state using lineage-aware policy boundary diagnostics.
Dev: It sounds like they are moving away from just looking at single numbers and trying to understand the whole structure of the system through these measurements.
The paper's summary: Rosa: This paper describes a pipeline where you take the artifact stream—the traces, metrics, deployment specifications—and run it through an extraction operator to get a lightweight model at each time step. This model might include an SLO inventory or a service graph, and then they map all that telemetry into a positive composition.
Dev: So instead of just looking at one metric like error budget share across different services, they are creating this whole structure where everything relates back to these defined parts and policies. It’s about getting the data into a shape that respects how the system is actually organized.
Taro: I've seen work on modeling software effort distribution before, but mapping continuous telemetry directly onto a compositional state seems like a big step when you have things like service splits happening while you’re running the analysis.
Rosa: The paper emphasizes that this process involves updating lineage and then using canonical groups to stabilize things through renames or merges. This keeps the dashboard signals comparable even when the underlying architecture is churning.
Dev: So the main takeaway is they are building a system that extracts structure from chaos—the constant architectural change—to create a stable view where we can see if we’re drifting toward a policy boundary, which is what Rasmussen's safety model suggests.
The paper's improvements: Rosa: One of the main things they suggest is moving beyond simple Euclidean thresholds. They use something called an isometric log-ratio map to represent the geometry of this compositional state, which is much better than raw numbers when you have these kinds of relationships.
Dev: And that leads to their stabilization layer where they keep a lineage map and aggregate parts into canonical groups, which is supposed to make sure that even if a service gets renamed or split, the dashboard level signal stays consistent.
Taro: I wonder about the boundary diagnostics part. They talk about step-to-boundary analysis, which sounds like they are trying to figure out how close we are to actually hitting a hard policy limit in this new space.
Rosa: That’s right. They model drift in this stabilized state additively, and only the product of the change and the direction you're moving is what they can reliably identify as a directional signal.
Dev: And they use log-ratios to encode policy constraints, like a ratio constraint such as 'failures divided by requests is less than some threshold,' so that when things go wrong, it shows up in those coordinates too.
Conclusion: Rosa: So we've seen how this framework takes the constant churn of software architecture—renames and splits—and turns it into a stable compositional view using lineage aggregation to keep dashboards comparable.
Dev: The paper's main contribution seems to be linking that compositional structure directly to safety science concepts like drift and boundary diagnostics, giving us a way to see not just that something is wrong, but in what direction it’s moving relative to the policy constraints.
Taro: For someone focused on resilience testing, the synthetic sanity checks are important; they showed that boundary-relevant drift can be separated from benign redistribution even under injected failures.
Rosa: It suggests we can build an automated system that reports on drift direction and step-to-boundary metrics, which is really helpful for SRE teams trying to decide when to intervene.
Dev: Overall, this paper proposes a method for discovering an operational model from the artifacts themselves, mapping telemetry into a compositional state, stabilizing that state under architectural churn with lineage awareness, and reporting those coherent log-ratio boundary diagnostics.
Taro: It’s about giving us a way to observe the system's safety margin as it evolves through software changes.
Rosa: That’s the core idea behind this paper "Toward Operationalizing Rasmussen in Evolving Software Systems." We'll be looking at how this stuff plays out next time.
Anatoly A. Krasnovsky
Innopolis University
eess.SY, cs.CY, cs.SY, stat.AP
Submitted: 2026-02-05
Updated: 2026-11-13
Code: https://github.com/a-a-k/drif
License: http://creativecommons.org/licenses/by-sa/4.0/
Importance score: 77/100
The gist: Software operations increasingly rely on SLOs, traces, deployment specifications, and change events, yet dashboards and thresholding practices often expose share-like operational signals as separate…
Key concepts
- Compositional Data Analysis (CoDA)
- This statistical framework is used because operational signals like remaining reliability margin are inherently compositional—parts sum up to a whole. CoDA helps analyze these parts together rather than treating them as independent numbers, which is crucial for accurately assessing system state under change.
- Isometric Log-Ratio (ILR) Map
- The ILR map provides a coordinate system for the compositional space. It allows the monitor to represent complex relationships between different operational parts in a way that preserves geometric distances, making it possible to track drift direction and proximity to policy boundaries effectively.
- Lineage-Aware Stabilization
- This technique addresses architectural churn like service splits or renames. By maintaining a lineage map, the system groups evolving components into stable canonical sets. This ensures that dashboard signals remain comparable even as the underlying software structure changes.
- Boundary-Directed Drift Diagnostics
- Instead of simple thresholds, this method looks for drift along a specific direction toward a defined policy boundary. It calculates 'step-to-boundary' to find the minimum change required to cross a safety limit, providing actionable insights into imminent risks.
Terminology
Summary
Software operations increasingly rely on SLOs, traces, deployment specifications, and change events, yet dashboards and thresholding practices often expose share-like operational signals as separate scalar panels or baseline distances <ref:2602. p. 0 Toward Operationalizing Rasmussen:
Drift Observability on the Simplex for Evolving Systems
Anatoly A. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. The proposed monitor would report drift direction,
step-to-boundary, balance-level attribution, and model-health indicators under architectural churn.Where the paper says it
Where the paper says it
Motivation for Drift Observability
Rasmussen’s dynamic safety model frames safety as a moving target where systems operate under competing gradients and local adaptations can produce slow drift toward failure as boundaries of safe operation are approached <ref:2602.05483#pg2> Software operations is a natural target because services evolve continuously, are governed by explicit policies such as SLOs and error-budgets, and emit rich telemetry <ref:2602. p. 0 Software operations is a natural target. Services evolve continuously, are governed by explicit policies such as SLOs and error
budgets, and emit rich telemetry. Services evolve continuously, are governed by explicit policies such as SLOs and error
budgets, and emit rich telemetry. Yet teams often reason about multiobjective operational state as shares: remaining reliability margin
across SLOs, engineering effort across work categories, or expected
incident impact across services. As shares, these signals form compositional views: relative structure is meaningful, while absolute
scale must be tracked separately. Applying Euclidean thresholds to
such closed data can therefore conflate safe redistribution with risk
accumulation [1, 13, 27, 28].Where the paper says it
Where the paper says it
Where the paper says it
Artifact-Derived Model Discovery
The monitor pipeline involves a sequence of steps where the artifact stream is mapped to a compositional state and then analyzed for drift <ref:2602.05483#pg2> The extraction operator yields a lightweight model Mt = Extract(At), which can include an SLO inventory, service graph, request-class partition, and policy-constraint set <ref:2602. p. 0 Prior
trace-based model-discovery and graph-simulation work can supply
parts of this upstream layer [18, 19]; this paper focuses on the driftobservability state built from its output. Concretely, an extraction
operator yields a lightweight model M𝑡 = Extract(A𝑡 ), for example an SLO inventory, service graph, request-class partition, and
policy-constraint set. The monitor pipeline is
A𝑡 → M𝑡 → Φ(M𝑡
, A𝑡 ) → x𝑡 → x˜𝑡 → ILR(x˜𝑡 ) → report. arXiv:2602. At each window, the monitor refreshes M𝑡
, maps telemetry to
positive parts through Φ, updates lineage 𝜋𝑡 and x˜𝑡
, computes balances and boundary distances, and emits either an operational drift
report or a model-health event. This deliberately weak extraction
contract is summarized in Table 1: the monitor needs a current
part inventory, telemetry-to-part mapping, lineage metadata, policy constraints where available, and health signals.Where the paper says it
Where the paper says it
Where the paper says it
Simplex Geometry and Lineage Stabilization
The strictly positive composition with D parts lies in the simplex SD−1 = ∪R D >0: Íi xi = 1 <ref:2602.05483#pg3> Aitchison operations define its geometry, and an isometric log-ratio (ILR) map is used as a coordinate representation of that geometry <ref:2602. p. 1 , 𝑥𝛼
𝐷
) provide the vector-space
structure used for compositional analysis [1, 13, 27]. We use an
isometric log-ratio (ILR) map as a coordinate representation of
that geometry [14]. Raw ILR coordinates are not directly comparable when services
split, SLOs are revised, or request classes are renamed. We maintain
a lineage map 𝜋𝑡
: {1, . Rasmussen’s dynamic safety model motivates drift under competing pressures, but operationalizing it for software is
difficult because relevant state variables—remaining margin, engineering effort, and risk/impact—are often compositional and their
parts evolve. We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. The monitor pipeline is
A𝑡 → M𝑡 → Φ(M𝑡
, A𝑡 ) → x𝑡 → x˜𝑡 → ILR(x˜𝑡 ) → report. arXiv:2602. Lineage
aggregation is an operational stabilization layer, not an isometrypreserving replacement for the leaf-level composition. It keeps
dashboard-level signals comparable under churn while retaining
leaf-level subcompositions for drill-down and sensitivity checks.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Drift Diagnostics on the Simplex
One-step drift in the stabilized state is modeled as x˜t+1 = x˜t ⊕ (beta ⊙ g˜t) ⊕ eta˜t <ref:2602.05483#pg5> In ILR coordinates, this becomes additive: z˜t+1 = z˜t + betau˜t + epsilont, where u˜t = ILR(g˜t) <ref:2602. p. 2 In ILR coordinates z˜𝑡 = ILR(x˜𝑡 ), Eq. (3) becomes additive: z˜𝑡+1 =
z˜𝑡 + 𝛽u˜𝑡 + 𝝐𝑡
, with u˜𝑡 = ILR(g˜𝑡 ). (3) becomes additive: z˜𝑡+1 =
z˜𝑡 + 𝛽u˜𝑡 + 𝝐𝑡
, with u˜𝑡 = ILR(g˜𝑡 ). Only the product 𝛽u˜𝑡
is identifiable from observed changes, so the primary directional signal is
the smoothed direction uˆ𝑡 = Δz˜𝑡 /∥Δz˜𝑡 ∥.Where the paper says it
Where the paper says it
Instantiations and Sanity Check Results
The synthetic sanity check over 300 trajectories of length 100 demonstrated that boundary-relevant drift should be separable from benign redistribution Under injected F /R drift, the boundary monitor detected all crossings with a median lead time of 13 windows Controlled split/merge churn resulted in a stable canonical signal with a median maximum error of 0 This confirms that lineage-aware aggregation preserved the canonical signal when lineage was exact The evaluation tested three falsifiable claims, including H1 that exogenous shocks induce non-zero mean drift directions in balance space
Positioning and Conclusion
The novelty of this work lies in reducing an evolving software architecture—renames, splits, merges, and policy revisions—to a stable compositional monitoring state with lineage-aware policy boundary diagnostics <ref:2602.05483#pg2> The proposed monitor reports drift direction, step-to-boundary, attribution via top-k balances mapped back to the extracted model, and model-health indicators such as ct and mother t This framework links safety-science drift models with artifact-derived observability for evolving systems <ref:2602. p. 2 Policy boundaries are the calibrated safe set; unexplained growth
in policy barriers or persistent boundary-directed balance drift is a
candidate for boundary discovery and post-incident learning, not
proof of an unknown failure boundary. The monitor emits a compact report: (1) scalar drift level and
trend; (2) boundary imminence via step-to-boundary; (3) attribution via top-𝑘 balances mapped back to the extracted model; and (4)
model-health indicators such as 𝑐𝑡 and 𝑚other
𝑡
. Rasmussen’s model suggests that local optimization under competing pressures can move systems toward unsafe boundaries. We
propose drift observability for evolving software systems: discover
an operational model from artifacts, map telemetry to a compositional state, stabilize that state under churn through lineage-aware
canonical groups, and report log-ratio coherent, boundary-aware
drift diagnostics with model-health gating. Prior SE uses of CoDA address cumulative voting and effort-phase distributions [7–9, 30]. Our target
is different: automated runtime drift observability under evolving
part inventories.Where the paper says it
Where the paper says it
Where the paper says it
References
[1] John Aitchison. 1986. The Statistical Analysis of Compositional Data. p. 3 5 Positioning and conclusion
Unlike ML concept-drift work, this paper links safety-science drift
models [10, 11, 23, 29], CoDA [1, 12–14, 27, 28, 33], and artifactderived observability or dependability analysis for distributed systems [3, 15, 18, 19, 21, 31, 32]. Monitoring compositional data is
studied in statistical process control, for example through ILR-based
control charts [24]; our novelty is not a new CoDA chart, but artifactderived operational state, software lineage, and policy-boundary
diagnostics for evolving systems.Where the paper says it
[2] Anatoly A. Krasnovsky. p. 0 Toward Operationalizing Rasmussen:
Drift Observability on the Simplex for Evolving Systems
Anatoly A. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. Toward Operationalizing Rasmussen:
Drift Observability on the Simplex for Evolving Systems
Anatoly A. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. Toward Operationalizing Rasmussen:
Drift Observability on the Simplex for Evolving Systems
Anatoly A. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. Toward Operationalizing Rasmussen:
Drift Observability on the Simplex for Evolving Systems , ,
The review artifact referenced in this paper is available as Ref. It contains the synthetic trajectory generator, fixed configuration,
replay script, generated summaries, and plots used for the controlled
sanity check in Section 4.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[3] Ali Basiri, Niosha Behnam, Ruud de Rooij, Lorin Hochstein, Luke Kosewski, Justin Reynolds, and Casey Rosenthal. 2016. Chaos Engineering
[4] Betsy Beyer, Chris Jones, Jennifer Petoff, and Niall Richard Murphy. 2016. Site Reliability Engineering: How Google Runs Production Systems
[5] Betsy Beyer, Niall Richard Murphy, David K. Rensin, Kent Kawahara, and Stephen Thorne. p. 4 [5] Betsy Beyer, Niall Richard Murphy, David K. Rensin, Kent Kawahara, and Stephen
Thorne.Where the paper says it
[6] David Challoner, Joanna Wijntjes, David Huska, Matthew Sartwell, Chris Coykendall, Chris Schrier, John Looney, Vivek Rau, Betsy Beyer, Max Luebbe, Alex Perry, and Murali Suriar. 2018. Eliminating Toil
[7] Panagiota Chatzipetrou, Lefteris Angelis, Per Rovegård, and Claes Wohlin. 2010. Prioritization of Issues and Requirements by Cumulative Voting: A Compositional Data Analysis Framework
[8] Panagiota Chatzipetrou, Efi Papatheocharous, Lefteris Angelis, and Andreas S. Andreou. 2012. An investigation of software effort phase distribution using compositional data
[9] Panagiota Chatzipetrou, Efi Papatheocharous, Lefteris Angelis, and Andreas S. Andreou. 2015. A Multivariate Statistical Framework for the Analysis of Software Effort Phase Distribution
[10] Richard Cook and Jens Rasmussen. 2005. “Going Solid”: A Model of System Dynamics and Consequences for Patient Safety <ref:2602. p. 4 004
[10] Richard Cook and Jens Rasmussen. “Going Solid”: A Model of System
Dynamics and Consequences for Patient Safety.Where the paper says it
[11] Sidney Dekker. 2011. Drift into Failure: From Hunting Broken Components to Understanding Complex Systems
[12] Juan José Egozcue and Vera Pawlowsky-Glahn. 2005. Groups of Parts and Their Balances in Compositional Data Analysis
[13] Juan José Egozcue and Vera Pawlowsky-Glahn. 2019. Compositional Data: The Sample Space and Its Structure
[14] Juan José Egozcue, Vera Pawlowsky-Glahn, Gloria Mateu-Figueras, and Carles Barceló-Vidal. 2003. Isometric Logratio Transformations for Compositional Data Analysis
[15] Victor Heorhiadi, Shriram Rajagopalan, Hani Jamjoom, Michael K. Reiter, and Vyas Sekar. p. 4 1023/A:
1023818214614
[15] Victor Heorhiadi, Shriram Rajagopalan, Hani Jamjoom, Michael K. Reiter, and
Vyas Sekar.Where the paper says it
[16] Erik Hollnagel, David D. Woods, and Nancy Leveson (Eds. p. 4 11
[16] Erik Hollnagel, David D. Woods, and Nancy Leveson (Eds. Woods, and Nancy Leveson (Eds. Resilience
Engineering: Concepts and Precepts (1 ed.Where the paper says it
Where the paper says it
[17] Chris Jones, John Wilkes, Niall Murphy, and Cody Smith. 2016. Service Level Objectives
[18] Anatoly A. Krasnovsky. 2025. Evaluating Asynchronous Semantics in TraceDiscovered Resilience Models: A Case Study on the OpenTelemetry Demo
[19] Anatoly A. Krasnovsky. 2026. Model Discovery and Graph Simulation: A Lightweight Gateway to Chaos Engineering
[21] Bowen Li, Xin Peng, Qilin Xiang, Haolin Wang, Tao Xie, Jun Sun, and Xuanzhe Liu. 2022. Enjoy Your Observability: An Industrial Survey of Microservice Tracing and Analysis
[22] Josep A. Martín-Fernández, Carles Barceló-Vidal, and Vera Pawlowsky-Glahn. p. 4 1007/s10664-
021-10063-9
[22] Josep A. Martín-Fernández, Carles Barceló-Vidal, and Vera Pawlowsky-Glahn.Where the paper says it
[23] J. Bradley Morrison and Robert L. p. 4 1
023/A:1023866030544
[23] J. Bradley Morrison and Robert L.Where the paper says it
[24] Thi Thuy Van Nguyen, Cédric Heuchenne, and Kim Phuc Tran. 2022. Anomaly Detection for Compositional Data using VSI MEWMA Control Chart
[25] OpenSLO Community. 2022. OpenSLO: Open Service Level Objective Specification
[26] OpenTelemetry. 2019. OpenTelemetry Specifications
[27] Vera Pawlowsky-Glahn and Juan José Egozcue. 2001. Geometric Approach to Statistical Analysis on the Simplex
[28] Vera Pawlowsky-Glahn, Juan José Egozcue, and Raimon Tolosana-Delgado. 2015. Modelling and Analysis of Compositional Data
[29] Jens Rasmussen. 1997. Risk Management in a Dynamic Society: A Modelling Problem <ref:2602. p. 4 1002/9781119003144
[29] Jens Rasmussen. Risk Management in a Dynamic Society: A Modelling
Problem.Where the paper says it
[30] Kaspars Rinkevics and Richard Torkar. 2013. Equality in Cumulative Voting: A Systematic Review with an Improvement Proposal
[31] Benjamin H. Sigelman, Luiz André Barroso, Mike Burrows, Pat Stephenson, Manoj Plakal, Donald Beaver, Saul Jaspan, and Chandan Shanbhag. p. 4 004
[31] Benjamin H. Sigelman, Luiz André Barroso, Mike Burrows, Pat Stephenson,
Manoj Plakal, Donald Beaver, Saul Jaspan, and Chandan Shanbhag.Where the paper says it
[32] Jacopo Soldani, Javad Khalili, and Antonio Brogi. 2023. Offline Mining of Microservice-Based Architectures (Extended Version)
[33] K. Gerald van den Boogaart and Raimon Tolosana-Delgado. p. 4 1007/s42979-023-01721-4
[33] K. Gerald van den Boogaart and Raimon Tolosana-Delgado. Gerald van den Boogaart and Raimon Tolosana-Delgado. Analyzing
Compositional Data with R.Where the paper says it
Where the paper says it
--- Page 1 ---
Toward Operationalizing Rasmussen: Drift Observability on the Simplex for Evolving Systems
Anatoly A. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment specifications, and change events, yet dashboards and thresholding practices often expose share-like operational signals as separate scalar panels or baseline distances <ref:2602.05483#pg2> This can create false alarms under benign redistribution and miss movement toward policy boundaries <ref:2602. p. 0 Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. This can create false alarms
under benign redistribution and miss movement toward policy
boundaries. Krasnovsky
Innopolis University
Innopolis, Russia
MB3R Lab
Innopolis, Russia
Abstract
Software operations increasingly rely on SLOs, traces, deployment
specifications, and change events, yet dashboards and thresholding
practices often expose share-like operational signals as separate
scalar panels or baseline distances. This can create false alarms
under benign redistribution and miss movement toward policy
boundaries. This can create false alarms
under benign redistribution and miss movement toward policy
boundaries. Rasmussen’s dynamic safety model motivates drift under competing pressures, but operationalizing it for software is
difficult because relevant state variables—remaining margin, engineering effort, and risk/impact—are often compositional and their
parts evolve. This can create false alarms
under benign redistribution and miss movement toward policy
boundaries. Rasmussen’s dynamic safety model motivates drift under competing pressures, but operationalizing it for software is
difficult because relevant state variables—remaining margin, engineering effort, and risk/impact—are often compositional and their
parts evolve. Rasmussen’s dynamic safety model motivates drift under competing pressures, but operationalizing it for software is
difficult because relevant state variables—remaining margin, engineering effort, and risk/impact—are often compositional and their
parts evolve. We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. The proposed monitor would report drift direction,
step-to-boundary, balance-level attribution, and model-health indicators under architectural churn.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Pruning redundant reasoning tokens
Software operations increasingly rely on SLOs, traces, deployment specifications, and change events <ref:2602.05483#pg2> This can create false alarms under benign redistribution and miss movement toward policy boundaries <ref:2602.05483#pg2> Rasmussen’s dynamic safety model motivates drift under competing pressures, but operationalizing it for software is difficult because relevant state variables—remaining margin, engineering effort, and risk/impact—are often compositional and their parts evolve <ref:2602.05483#pg2>
How the training signal is built
We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable compositional monitoring state <ref:2602.05483#pg2> The monitor extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware canonical groups <ref:2602. p. 0 We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. The proposed monitor would report drift direction,
step-to-boundary, balance-level attribution, and model-health indicators under architectural churn.Where the paper says it
Where the gains come from
The novelty is not CoDA itself or a new trace parser but reducing an evolving software architecture—renames, splits, merges, and policy revisions—to a stable compositional monitoring state with lineage-aware policyboundary diagnostics <ref:2602.05483#pg2> The proposed monitor reports drift direction, step-to-boundary, balance-level attribution, and model-health indicators under architectural churn <ref:2602. p. 0 We formulate an automated, artifact-derived driftmonitor design that maps changing software artifacts into a stable
compositional monitoring state: it extracts a current part inventory and policy constraints, maps telemetry to a positive composition, stabilizes splits, merges, and renames through lineage-aware
canonical groups, and analyzes boundary-directed drift in log-ratio
coordinates. The proposed monitor would report drift direction,
step-to-boundary, balance-level attribution, and model-health indicators under architectural churn.Where the paper says it
From artifacts to a compositional state Let At denote the artifact/telemetry stream at time t <ref:2602.05483#pg3> Concretely, an extraction operator yields a lightweight model Mt = Extract(At), for example an SLO inventory, service graph, request-class partition, and policy-constraint set <ref:2602. p. 0 Prior
trace-based model-discovery and graph-simulation work can supply
parts of this upstream layer [18, 19]; this paper focuses on the driftobservability state built from its output. Concretely, an extraction
operator yields a lightweight model M𝑡 = Extract(A𝑡 ), for example an SLO inventory, service graph, request-class partition, and
policy-constraint set. The monitor pipeline is
A𝑡 → M𝑡 → Φ(M𝑡
, A𝑡 ) → x𝑡 → x˜𝑡 → ILR(x˜𝑡 ) → report. arXiv:2602.Where the paper says it
Where the paper says it
--- Page 2 ---
,, Krasnovsky
Figure 1: A minimal effort-share example x = (F,R,O) for feature work, reliability work, and operations/toil. In (A), Euclidean monitoring can alarm although F /R is unchanged. p. 1 , , Krasnovsky
Figure 1: A minimal effort-share example x = (𝐹, 𝑅,𝑂) for feature work, reliability work, and operations/toil. In (A), Euclidean
monitoring can alarm although 𝐹 /𝑅 is unchanged. In (A), Euclidean
monitoring can alarm although 𝐹 /𝑅 is unchanged. In (B), the point remains Euclidean-near to baseline although 𝐹 /𝑅 > 1.Where the paper says it
Where the paper says it
At each window, the monitor refreshes Mt, maps telemetry to positive parts through Φ, updates lineage pit and x˜t, computes balances and boundary distances <ref:2602.05483#pg2> Missingness, low extraction confidence ct, or large mass in “other” triggers rebaselining/model refinement rather than a trusted drift alarm <ref:2602. p. 1 This deliberately weak extraction
contract is summarized in Table 1: the monitor needs a current
part inventory, telemetry-to-part mapping, lineage metadata, policy constraints where available, and health signals. Missingness,
low extraction confidence 𝑐𝑡
, or large mass in “other” triggers rebaselining/model refinement rather than a trusted drift alarm.Where the paper says it
A strictly positive composition with D parts lies in SD−1 = ∪R D >0: Íi xi = 1 We use an isometric log-ratio (ILR) map as a coordinate representation of that geometry A balance basis is a particular orthonormal ILR basis induced by a sequential binary partition chosen for operational interpretability Raw ILR coordinates are not directly comparable when services split, SLOs are revised, or request classes are renamed We maintain a lineage map pit: 1,..., D(t) → 1,...
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Automated operational state representation: The system can map changing software artifacts into a stable compositional monitoring state by extracting a current part inventory and policy constraints, as described in Section 2.1.
-
Lineage-aware stabilization: The AI system can stabilize renames, splits, and merges through lineage aware canonical groups to maintain comparable dashboard-level signals under architectural churn, following the method outlined in Section 2.2.
-
Boundary-directed drift diagnostics: The system can report drift direction, step-to-boundary (seeking the smallest lambda such that ILR minus one is not in Omega), and model health indicators like missingness or other mass to drive SRE actions, as detailed in Section 3.2.
-
Policy boundary encoding: The system can encode ratio constraints into log-ratios, such as F/R less than a threshold, to define policy boundaries and diagnose when unexplained growth in these barriers indicates a candidate for boundary discovery, per Section 3.2.
-
Synthetic sanity checking: The improved AI system can be validated against controlled synthetic trajectories that test necessary conditions like separating boundary-relevant drift from benign redistribution and verifying that lineage-aware aggregation preserves the canonical signal during churn, as described in Section 4.2.
Abstract
Rasmussen's qualitative drift model explains how successive, locally successful adaptations under efficiency pressures can move a system toward unacceptable operation. Brooker and Hochstein applied this perspective conceptually to software operations without specifying a quantitative procedure for observing drift. We take a step in this direction and propose a runtime observer that quantitatively operationalizes the model's observable part for evolving software. The proposed observer is intended to give engineers evidence for pausing or revising adaptation before declared operating limits are crossed. It tracks how activity or resource use is distributed across components and types of work. Measurement history quantifies accumulated change and recent movement toward or away from current limits. Assuming that direction and limits remain unchanged, the proposed observer calculates the first constraint ahead and the amount of further change required to reach it. Measurements remain comparable across supported component changes, while policy revisions are distinguished from changes in execution. Preliminary runtime studies demonstrated these distinctions under annotated software changes and revised retry limits. A separate multivariate exponentially weighted moving average (MEWMA) experiment detected CPU redistribution earlier with log-ratio coordinates than with raw shares, motivating investigation of representation-dependent observation. Further work will ground constraints in service outcomes, investigate runtime-model acquisition, and test whether decisions informed by observed drift reduce violations at acceptable useful-work loss and intervention cost.
Sources
- Evaluating Asynchronous Semantics in Trace-Discovered Resilience Models: A Case Study on the OpenTelemetry Demo
- Anomaly Detection for Compositional Data using VSI MEWMA control chart
Related papers
- Industrial Load Modeling and Optimization for Market-Based Interaction with Power Systems
- Exact and Approximate Convex Reformulation of Linear Stochastic Optimal Control with Chance Constraints
- Certified Set Convergence for Piecewise Affine Systems
- Symbolic Control for Stochastic Systems via Finite Parity Games
- Multiobjective optimization-based design and dispatch of islanded, hybrid microgrids for remote, off-grid communities in sub-Saharan Africa
- Physiology-informed layered sensing for intelligent human-exoskeleton interaction