Pith. sign in

REVIEW 2 major objections 2 minor 42 references

Reviewed by Pith at T0; open to challenge.

T0 means a machine referee read the full paper against a public rubric. The mark states how deep the mechanical check went, never who wrote it. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.3

A multi-objective optimizer driven by necessary and sufficient conditional probabilities of cross-application interference produces microservice placements that reduce interference and improve response times.

2026-06-25 19:45 UTC pith:DWMU62AA

load-bearing objection The paper applies causal conditional probabilities to interference-aware microservice placement via Optuna but provides no clear evidence that the probabilities are independently derived or hold up outside the profiled traces. the 2 major comments →

arxiv 2606.25922 v1 pith:DWMU62AA submitted 2026-06-24 cs.DC

Interference-Aware Cross-Application Placement: A Multi-Objective Optimization Approach for Microservice Clusters

Iqra Zafar , Christian Medeiros Adriano , Holger Giese This is my paper
classification cs.DC
keywords microservice placementcross-application interferencecausal probabilitiesmulti-objective optimizationcloud resource sharingperformance isolationcontainerized applicationsresponse time optimization
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that interference in shared microservice clusters can be captured in a spatio-temporal data structure whose causal effects are expressed as necessary and sufficient conditional probabilities. These probabilities, together with per-service latency baselines such as 95th-percentile response times, feed a multi-objective optimizer that also accounts for network penalties and isolation constraints. The optimizer yields placements that measurably lower cross-application interference while preserving or improving latency on real multi-application workloads. A sympathetic reader would care because current placement methods either ignore why certain co-locations cause interference or optimize only one objective, leaving operators without explicit levers over the full trade-off space. The approach therefore supplies a concrete mechanism for controlling interference, latency, and communication overhead in one formulation.

Core claim

The paper claims that cross-application profiling can generate traces whose interference effects are formalized as necessary and sufficient conditional probabilities inside a spatio-temporal data structure; these probabilities then drive a multi-objective optimizer (Optuna) that simultaneously minimizes interference, respects latency baselines, network penalties, and isolation requirements, and supports adjustable weighting of the causal metrics, resulting in placements that reduce interference and improve response performance on real workloads.

What carries the argument

Spatio-temporal data structure whose causal effects are expressed as necessary and sufficient conditional probabilities that inform the multi-objective optimizer.

Load-bearing premise

Cross-application profiling can produce traces and probability estimates that remain accurate and generalizable enough to guide placement decisions in live production systems.

What would settle it

Deploy the produced placements on the same real multi-application workloads and measure whether the observed cross-application interference and p95 response times fail to improve over the non-interference-aware baseline.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Placements generated by the optimizer reduce measured cross-application interference on real workloads.
  • The same placements improve response performance as captured by 95th-percentile latency.
  • The formulation can incorporate network penalties and application isolation constraints without changing the core probability structure.
  • Operators can adjust the relative weighting of necessary versus sufficient causal metrics to emphasize different trade-offs.
  • The approach supplies explicit control over interference, latency, and communication overhead in a single optimization run.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the conditional probabilities prove stable across workload changes, the same profiling step could be reused for incremental re-optimization rather than full recomputation.
  • The causal formulation might extend to interference arising from shared storage or cache contention, not only CPU and network resources.
  • Cloud operators could expose the adjustable weights as user-facing knobs, allowing tenants to declare tolerance for interference versus latency.
  • The spatio-temporal structure might serve as input to online monitoring systems that update probabilities from live telemetry instead of offline traces.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 2 minor

Summary. The paper claims to address interference in microservice clusters by constructing a spatio-temporal data structure that captures causal effects of cross-application interference, formalized as necessary and sufficient conditional probabilities estimated via cross-application profiling and trace simulation. These probabilities, along with p95 latency baselines, network penalties, and isolation requirements, drive a multi-objective optimizer (Optuna) with adjustable weighting. The central claim is that this causality-driven approach yields placements that significantly reduce interference and improve response times compared to existing methods, as shown in experiments on real multi-application workloads.

Significance. If the interference probabilities prove accurate, independent of the evaluation traces, and generalizable, the work could offer cloud operators explicit, tunable control over interference-latency-overhead trade-offs in a way that goes beyond latency-only optimization. The multi-objective formulation and use of causal metrics represent a potential advance in microservice placement if the experimental validation demonstrates that the reported gains are attributable to the causal modeling rather than to the specific profiled workloads.

major comments (2)
  1. [Abstract] Abstract: the assertion that 'experimental results on real multi-application workloads show that interference-aware placements significantly reduce cross-application interference and improve response performance' supplies no quantitative metrics, baselines, error bars, statistical tests, or workload descriptions. This absence is load-bearing because the central claim rests entirely on these unreported outcomes.
  2. [Profiling and probability estimation] Profiling and probability estimation section: the manuscript does not describe how the necessary and sufficient conditional probabilities control for confounding variables (load variation, hardware heterogeneity, or non-profiled co-located services) or demonstrate that the estimates remain predictive outside the profiled trace set. Without such controls, performance gains cannot be attributed to the causal placement rather than to the experimental setup.
minor comments (2)
  1. [Experiments] The paper should include a dedicated subsection or table explicitly comparing the proposed method against standard baselines (e.g., latency-only or random placement) with the same workloads.
  2. [Introduction] Clarify whether the spatio-temporal data structure is a novel contribution or an application of an existing formalism; if novel, provide a brief formal definition or pseudocode.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below and will revise the manuscript accordingly to strengthen the presentation of results and methodological details.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the assertion that 'experimental results on real multi-application workloads show that interference-aware placements significantly reduce cross-application interference and improve response performance' supplies no quantitative metrics, baselines, error bars, statistical tests, or workload descriptions. This absence is load-bearing because the central claim rests entirely on these unreported outcomes.

    Authors: We agree that the abstract lacks the quantitative support needed to substantiate the central claim. In the revised version we will expand the abstract to report specific metrics (e.g., percentage reductions in interference events and p95 latency improvements), name the baselines and workloads used, and reference the statistical tests and error bars shown in the experimental section. revision: yes

  2. Referee: [Profiling and probability estimation] Profiling and probability estimation section: the manuscript does not describe how the necessary and sufficient conditional probabilities control for confounding variables (load variation, hardware heterogeneity, or non-profiled co-located services) or demonstrate that the estimates remain predictive outside the profiled trace set. Without such controls, performance gains cannot be attributed to the causal placement rather than to the experimental setup.

    Authors: The manuscript currently describes estimation via cross-application profiling and trace simulation but does not explicitly detail controls for the listed confounders or out-of-sample validation. We will revise the section to add: (1) a description of the controlled profiling environment (fixed load levels, homogeneous hardware, isolated co-location sets) used to isolate the conditional probabilities, and (2) results of predictive checks on held-out trace segments. If the available data do not fully support generalizability claims, we will also note this limitation. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation uses independent profiling estimates to drive optimizer

full rationale

The abstract describes a pipeline in which cross-application profiling generates traces to estimate conditional probabilities of interference; these probabilities, together with separate p95 latency baselines, feed a multi-objective Optuna optimizer whose outputs are then evaluated on real workloads. No equations, self-citations, or definitional steps are supplied that would make the reported interference reductions or latency improvements equivalent to the input estimates by construction. The evaluation step is presented as an independent test of the placements produced by the optimizer, satisfying the requirement for an externally falsifiable outcome.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

Review performed on abstract only; free parameters, axioms, and invented entities cannot be exhaustively listed without the full manuscript. The abstract implies at least one adjustable weighting parameter and the assumption that conditional probabilities capture necessary and sufficient causal effects.

free parameters (1)
  • weighting of necessary and sufficient causal metrics
    Abstract states the approach supports adjustable weighting but provides no values or fitting procedure.
axioms (1)
  • domain assumption Causal effects of interference can be formalized as necessary and sufficient conditional probabilities estimated from profiling
    Central formalization step described in the abstract.
invented entities (1)
  • spatio-temporal data structure for interference no independent evidence
    purpose: Captures causal effects of cross-application interference
    Introduced in the abstract as the foundation for probability estimation; no independent evidence supplied.

pith-pipeline@v0.9.1-grok · 5768 in / 1338 out tokens · 30333 ms · 2026-06-25T19:45:06.873783+00:00 · methodology

0 comments
read the original abstract

In modern cloud architectures, multiple applications often run within the same clustered environment, sharing underlying resources. This resource sharing can cause interference among applications, leading to degraded latency and reduced system stability. As containerized microservices become increasingly central to cloud-native applications, their performance can suffer from complex interference scenarios related to resource competition. Meanwhile, most existing microservice approaches address interference either by detecting and localizing performance issues or by optimizing latency alone, without explaining why specific co-locations cause cross-application interference, and how this can inform service placement optimization. This work closes that gap by building a spatio-temporal data structure that captures the causal effects of cross-application interference. These causal effects are mathematically formalized as necessary and sufficient conditional probabilities that inform a multi-objective optimizer (Optuna). Cross-application profiling is used to simulate traces and estimate interference probabilities, while per-service latency baselines are provided by performance data, such as 95th-percentile response times (p95). Our approach supports network penalties, application isolation requirements, and adjustable weighting of necessary and sufficient causal metrics. Experimental results on real multi-application workloads show that interference-aware placements significantly reduce cross-application interference and improve response performance. Ultimately, the causality-driven multi-objective formulation gives cloud operators explicit control over interference, latency, and communication overhead when configuring service placements.

Figures

Figures reproduced from arXiv: 2606.25922 by Christian Medeiros Adriano, Holger Giese, Iqra Zafar.

Figure 1
Figure 1. Figure 1: STIG-based MAPE-K loop with interference-aware service placement. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: knowledge graph Reproducibility – To create the type cluster, deploy Istio, and execute all experiments, we offer k6 scripts, Kubernetes manifests (Namespaces, Gateway/HTTPRoutes, Destination￾Rules), and a run script6 . Each run used a different set of random seeds and scenario ordering. To demonstrate the resilience of our proposed method and its benefits over several state-of-the-art approaches, we carry… view at source ↗
Figure 3
Figure 3. Figure 3: Interference Anomaly (Service Failure/Error) vs Request Duration [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 1 canonical work pages

  1. [1]

    Newman,Building microservices: designing fine-grained systems

    S. Newman,Building microservices: designing fine-grained systems. O’Reilly, 2021

    2021

  2. [2]

    iplace: An interference-aware clustering algorithm for microservice placement,

    M. Adeppady, C. Chiasserini, H. Karl, and P. Giaccone, “iplace: An interference-aware clustering algorithm for microservice placement,” in Proc. of the IEEE Int. Conf. on Communications (ICC), 2022

    2022

  3. [3]

    Towards interference-resilient multi-tenant microservices via spatio-temporal models of self-configuration,

    I. Zafar, “Towards interference-resilient multi-tenant microservices via spatio-temporal models of self-configuration,” in2024 IEEE Interna- tional Conference on Autonomic Computing and Self-Organizing Sys- tems Companion (ACSOS-C). IEEE, 2024, pp. 173–175

    2024

  4. [4]

    Understanding performance interference of i/o workload in virtualized cloud environ- ments,

    X. Pu, L. Liu, Y . Mei, S. Sivathanu, Y . Koh, and C. Pu, “Understanding performance interference of i/o workload in virtualized cloud environ- ments,” inIEEE 3rd Int. Conf. on Cloud Computing. Miami, FL, USA: IEEE, 2010, pp. 51–58

    2010

  5. [5]

    Reducing microservices interference and deployment time in resource-constrained cloud systems,

    M. Adeppady, P. Giaccone, H. Karl, and C. F. Chiasserini, “Reducing microservices interference and deployment time in resource-constrained cloud systems,”IEEE Trans. on Network and Service Management, vol. 20, no. 3, pp. 3135–3147, 2023

    2023

  6. [6]

    IRENE: Interference and high availability aware microservice-based applications placement for edge computing,

    P. Souza, J. Nascimento, C. Boeira, ˆA. Vieira, F. Rubin, R. Reis, F. Rossi, and T. Ferreto, “IRENE: Interference and high availability aware microservice-based applications placement for edge computing,” inProc. of the 10th Int. Conf. on Cloud Computing and Services Science (CLOSER). SCITEPRESS, 2020

    2020

  7. [7]

    C-koordinator: Interference-aware management for large-scale and co- located microservice clusters,

    S. Song, M. Xu, Z. Zhang, C. Gao, F. Zeng, Y . Ding, K. Ye, and C. Xu, “C-koordinator: Interference-aware management for large-scale and co- located microservice clusters,”arXiv preprint, 2025

    2025

  8. [8]

    Community and priority-based microservice placement in dynamic microservice systems,

    Z. Quet al., “Community and priority-based microservice placement in dynamic microservice systems,”Journal of Systems Architecture, 2024, interference-aware community- and priority-based placement

    2024

  9. [9]

    A survey on engineering approaches for self-adaptive systems,

    C. Krupitzer, F. M. Roth, S. VanSyckel, G. Schiele, and C. Becker, “A survey on engineering approaches for self-adaptive systems,”Pervasive and Mobile Computing, vol. 17, pp. 184–206, 2015

    2015

  10. [10]

    Stigs: Spatio-temporal interfer- ence graph simulator for self-configurable multi-tenant cloud systems,

    I. Zafar, C. M. Adriano, and H. Giese, “Stigs: Spatio-temporal interfer- ence graph simulator for self-configurable multi-tenant cloud systems,” inCompanion of the 15th ACM/SPEC International Conference on Performance Engineering, 2024, pp. 52–56

    2024

  11. [11]

    Addressing shared re- source contention in multicore processors via scheduling,

    S. Zhuravlev, S. Blagodurov, and A. Fedorova, “Addressing shared re- source contention in multicore processors via scheduling,”ACM Sigplan Notices, vol. 45, no. 3, pp. 129–142, 2010

    2010

  12. [12]

    Causal sufficiency and actual causation,

    S. Beckers, “Causal sufficiency and actual causation,”Journal of Philo- sophical Logic, vol. 50, no. 6, pp. 1341–1374, 2021

    2021

  13. [13]

    Pearl,Causality

    J. Pearl,Causality. Cambridge university press, 2009

    2009

  14. [14]

    Actual causation and the art of modeling,

    J. Y . Halpern and C. Hitchcock, “Actual causation and the art of modeling,”Metaphysica, vol. 11, no. 1, pp. 25–61, 2010

    2010

  15. [15]

    Causes and explanations in the structural- model approach: Tractable cases,

    T. Eiter and T. Lukasiewicz, “Causes and explanations in the structural- model approach: Tractable cases,”Artificial Intelligence, vol. 116, no. 1–2, pp. 23–74, 2002

    2002

  16. [16]

    Causes and explanations: A structural-model approach. part i: Causes,

    J. Y . Halpern and J. Pearl, “Causes and explanations: A structural-model approach. part i: Causes,”British Journal for the Philosophy of Science, vol. 56, no. 4, pp. 843–887, 2005

    2005

  17. [17]

    Neuro- symbolic causal reasoning for cautious self-adaptation under distribution shifts,

    C. M. Adriano, S. Ghahremani, F. Kaiser, and H. Giese, “Neuro- symbolic causal reasoning for cautious self-adaptation under distribution shifts,” in2025 IEEE International Conference on Autonomic Computing and Self-Organizing Systems (ACSOS). IEEE, 2025, pp. 88–99

    2025

  18. [18]

    Software architecture optimization methods: A systematic literature review,

    A. Aleti, B. Buhnova, L. Grunske, A. Koziolek, and I. Meedeniya, “Software architecture optimization methods: A systematic literature review,”IEEE Trans. on Software Engineering, vol. 39, no. 5, pp. 658– 683, 2012

    2012

  19. [19]

    Anomaly detection and failure root cause analysis in (micro) service-based cloud applications: A survey,

    J. Soldani and A. Brogi, “Anomaly detection and failure root cause analysis in (micro) service-based cloud applications: A survey,”ACM Computing Surveys (CSUR), vol. 55, no. 3, pp. 1–39, 2022

    2022

  20. [20]

    Sleuth: A trace-based root cause analysis system for large-scale microservices with graph neural networks,

    Y . Gan, G. Liu, X. Zhang, Q. Zhou, J. Wu, and J. Jiang, “Sleuth: A trace-based root cause analysis system for large-scale microservices with graph neural networks,” inProc. of the 28th ACM Int. Conf. on Architectural Support for Programming Languages and Operating Systems, Volume 4, 2023, pp. 324–337

    2023

  21. [21]

    Towards a framework for optimized microservices placement in cloud native environments,

    I. Benamor, M. Zarour, and A. Tari, “Towards a framework for optimized microservices placement in cloud native environments,”Int. Journal of Advanced Computer Science and Applications, vol. 15, no. 7, 2024

    2024

  22. [22]

    Multi-objective reconfiguration of evolving real- time microservices-based environments using energy-based nuclear reaction optimization,

    M. N. Abadeh, “Multi-objective reconfiguration of evolving real- time microservices-based environments using energy-based nuclear reaction optimization,”Systems, 2025, multi-objective nuclear-reaction- based meta-heuristic for real-time microservices. [Online]. Available: https://scimet.iau.ir/University.php?id=157

    2025

  23. [23]

    Optuna: A next-generation hyperparameter optimization framework,

    T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama, “Optuna: A next-generation hyperparameter optimization framework,” inProc. of the 25th ACM SIGKDD Int. Conf. on Knowledge Discovery & Data Mining. ACM, 2019

    2019

  24. [24]

    Edge service placement optimization for location-based service,

    K. Kittikamron, N. Manop, A. Chanakitkarnchok, and K. Rojviboonchai, “Edge service placement optimization for location-based service,” in 20th Int. Joint Conf. on Computer Science and Software Engineering (JCSSE). IEEE, 2023, pp. 488–493

    2023

  25. [25]

    Causal latency modelling for cloud microservices,

    C. Lohse, D. Tsutsumi, A. Ba, P. Harsha, C. Subramanian, M. Straesser, and M. Ruffini, “Causal latency modelling for cloud microservices,” in IEEE 18th Int. Conf. on Cloud Computing. IEEE, 2025, pp. 143–151

    2025

  26. [26]

    Microservices performance testing with causality-enhanced large lan- guage models,

    C. Mascia, R. Pietrantuono, A. Guerriero, L. Giamattei, and S. Russo, “Microservices performance testing with causality-enhanced large lan- guage models,” inProc. of the 2nd ACM Int. Conf. on AI Foundation Models and Software Engineering (FORGE), 2025

    2025

  27. [27]

    Detecting causal structure on cloud application microservices using granger causality models,

    Q. Wang, L. Shwartzet al., “Detecting causal structure on cloud application microservices using granger causality models,” inProc. of the IEEE Int. Conf. on Cloud Computing (CLOUD), 2021

    2021

  28. [28]

    Root cause analysis for microservices based on causal inference: How far are we?

    L. Phamet al., “Root cause analysis for microservices based on causal inference: How far are we?” inProc. of the 39th IEEE/ACM Int. Conf. on Automated Software Engineering (ASE), 2024

    2024

  29. [29]

    A comprehensive survey on root cause analysis in (micro) services: methodologies, challenges, and trends,

    T. Wang and G. Qi, “A comprehensive survey on root cause analysis in (micro) services: methodologies, challenges, and trends,”arXiv preprint arXiv:2408.00803, 2024

    work page arXiv 2024

  30. [30]

    Ser- viceanomaly: An anomaly detection approach in microservices using distributed traces and profiling metrics,

    M. Panahandeh, A. Hamou-Lhadj, M. Hamdaqa, and J. Miller, “Ser- viceanomaly: An anomaly detection approach in microservices using distributed traces and profiling metrics,”Journal of Systems and Soft- ware, vol. 209, p. 111917, 2024

    2024

  31. [31]

    {App–bisect}: Autonomous healing for{microservice-based}apps,

    S. Rajagopalan and H. Jamjoom, “{App–bisect}: Autonomous healing for{microservice-based}apps,” inProc. of the 7th USENIX Workshop on Hot Topics in Cloud Computing (HotCloud 15), 2015

    2015

  32. [32]

    Microrca: Root cause localization of performance issues in microservices,

    L. Wu, J. Tordsson, E. Elmroth, and O. Kao, “Microrca: Root cause localization of performance issues in microservices,” inProc. of the IEEE/IFIP Network Operations and Management Symp. (NOMS), 2020

    2020

  33. [33]

    Chain-of-event: Interpretable root cause analysis for microser- vices through automatically learning weighted event causal graph,

    Z. Yao, C. Pei, W. Chen, H. Wang, L. Su, H. Jiang, Z. Xie, X. Nie, and D. Pei, “Chain-of-event: Interpretable root cause analysis for microser- vices through automatically learning weighted event causal graph,” in Companion Proc. of the 32nd ACM Int. Conf. on the Foundations of Software Engineering, 2024, pp. 50–61

    2024

  34. [34]

    Sparserca: Unsupervised root cause analysis in sparse microservice testing traces,

    Z. Yao, H. Ye, C. Pei, G. Cheng, G. Wang, Z. Liu, H. Chen, H. Cui, Z. Li, J. Liet al., “Sparserca: Unsupervised root cause analysis in sparse microservice testing traces,” inProc. of the IEEE 35th Int. Symp. on Software Reliability Engineering (ISSRE). IEEE, 2024, pp. 391–402

    2024

  35. [35]

    {CRISP}: Critical path analysis of{Large-Scale}microser- vice architectures,

    Z. Zhang, M. K. Ramanathan, P. Raj, A. Parwal, T. Sherwood, and M. Chabbi, “{CRISP}: Critical path analysis of{Large-Scale}microser- vice architectures,” inProc. of the USENIX Annual Technical Conf. (USENIX ATC 22), 2022, pp. 655–672

    2022

  36. [36]

    Synthetic time series for anomaly detection in cloud microservices,

    M. Allam, N. Boujnah, N. E. O’Connor, and M. Liu, “Synthetic time series for anomaly detection in cloud microservices,” inInt. Conf. on Machine Learning, Optimization, and Data Science. Springer, 2024, pp. 419–433

    2024

  37. [37]

    A multi-objective genetic algorithm (ga) approach for optimization of surface grinding operations,

    R. Saravanan, P. Asokan, and M. Sachidanandam, “A multi-objective genetic algorithm (ga) approach for optimization of surface grinding operations,”Int. journal of machine tools and manufacture, vol. 42, no. 12, pp. 1327–1334, 2002

    2002

  38. [38]

    Reinforcement learning algorithms: A brief survey,

    A. K. Shakya, G. Pillai, and S. Chakrabarty, “Reinforcement learning algorithms: A brief survey,”Expert Systems with Applications, vol. 231, p. 120495, 2023

    2023

  39. [39]

    A self-stabilizing and auto-provisioning orchestration for microservices in edge-cloud continuum,

    B. Cai, X. Wang, B. Wang, M. Yang, Y . Guo, and Q. Guo, “A self-stabilizing and auto-provisioning orchestration for microservices in edge-cloud continuum,”Computer Networks, vol. 242, p. 110279, 2024

    2024

  40. [40]

    µopt: An efficient optimal autoscaler for microservice applications,

    E. Incerto, R. Pizziol, and M. Tribastone, “µopt: An efficient optimal autoscaler for microservice applications,” in2023 IEEE Int. Conf. on Autonomic Computing and Self-Organizing Systems (ACSOS). IEEE, 2023, pp. 67–76

    2023

  41. [41]

    Autoscaling solutions for cloud applications under dynamic workloads,

    G. Quattrocchi, E. Incerto, R. Pinciroli, C. Trubiani, and L. Baresi, “Autoscaling solutions for cloud applications under dynamic workloads,” IEEE Trans. on Services Computing, vol. 17, no. 3, pp. 804–820, 2024

    2024

  42. [42]

    Autoscaling of microservice resources based on dense connectivity spatio-temporal gnn and q- learning,

    P. Liang, Y . Xun, J. Cai, and H. Yang, “Autoscaling of microservice resources based on dense connectivity spatio-temporal gnn and q- learning,”Future Generation Computer Systems, vol. 174, p. 107909, 2026

    2026