pith. sign in

arxiv: 2604.05496 · v1 · submitted 2026-04-07 · 💻 cs.DC · cs.PF· cs.SE

Optimizing OpenFaaS on Kubernetes: Comparative Analysis of Language Runtimes and Cluster Distributions

Ehsan Ataie , Mohammadreza Pooshani , Hossein Aqasizade This is my paper

Pith reviewed 2026-05-10 19:45 UTC · model grok-4.3

classification 💻 cs.DC cs.PFcs.SE
keywords OpenFaaSKubernetes distributionsserverless computingperformance evaluationGo runtimePythonNode.jsconcurrent invocations
0
0 comments X

The pith

Go outperforms Python and Node.js in throughput and CPU efficiency for OpenFaaS functions on Kubernetes, while K3s and Kubeadm lead among the tested distributions.

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

This paper compares how OpenFaaS, an open-source serverless platform, performs when deployed on different Kubernetes distributions and when functions are written in Python, Go, or Node.js. Tests run on CloudLab hardware measure behavior under increasing numbers of simultaneous function calls. The work focuses on practical metrics including throughput, CPU usage, and response delay to identify efficient combinations. A reader would care because these choices directly affect speed, resource costs, and reliability when running serverless applications in production. The results point to clear preferences among the options examined.

Core claim

The paper establishes that Go functions achieve consistently higher throughput and lower CPU usage than equivalent functions in Python or Node.js across the tested setups. Among the Kubernetes distributions, Kubeadm delivers low latency with efficient CPU consumption while K3s provides the highest throughput. These outcomes are derived from controlled experiments that vary the level of concurrent invocations and record usage-level and responsiveness metrics for each combination of runtime and cluster distribution.

What carries the argument

Comparative benchmarking of OpenFaaS function performance across language runtimes and Kubernetes distributions under controlled concurrent invocation loads, tracked through throughput, CPU usage, and delay metrics.

Load-bearing premise

The chosen concurrency levels, simple function implementations, and specific CloudLab hardware configuration produce results representative of typical production serverless workloads without distortion from unmeasured network, storage, or tuning factors.

What would settle it

Re-running the identical benchmark suite on different hardware, with more complex or stateful functions, or under production-like traffic patterns and confirming whether Go and the top distributions retain their measured advantages in throughput, CPU, and latency.

Figures

Figures reproduced from arXiv: 2604.05496 by Ehsan Ataie, Hossein Aqasizade, Mohammadreza Pooshani.

Figure 3
Figure 3. Figure 3: Requests per second for Kubeadm [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Average delay for Kubeadm [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Requests per second for K3s [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 10
Figure 10. Figure 10: Average delay for K0s 4.2 CPU Usage In the previous part, we discussed how Go has the best performance among programming language runtimes, but what about its CPU usage Figures 11 to 14 illustrate the CPU usage of programming language runtimes for K8s distributions. Based on the data, all languages exhibit a progressive increase in CPU usage as the number of concurrent operations grows. The trend indicate… view at source ↗
Figure 7
Figure 7. Figure 7: Requests per second for MicroK8s [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Average delay for MicroK8s [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 11
Figure 11. Figure 11: Kubeadm CPU usage [PITH_FULL_IMAGE:figures/full_fig_p009_11.png] view at source ↗
read the original abstract

Serverless computing, particularly Function-as-a-Service (FaaS), has revolutionized cloud computing by abstracting infrastructure management and enabling dynamic resource allocation. This paper examines the performance and compatibility of OpenFaaS, an open-source serverless platform, when deployed on various Kubernetes distributions, including Kubeadm, K3s, MicroK8s, and K0s. Moreover, leveraging the CloudLab infrastructure, this study examines the impact of Python, Go, and Node$.$js programming languages on the performance of Kubernetes-enabled OpenFaaS, specifically when these languages are used to develop functions deployed on the platform. The performance is evaluated and analyzed under various levels of concurrent invocations using several usage-level metrics, such as throughput and CPU usage, as well as responsiveness metrics, such as delay. According to our findings, Go consistently outperforms Python and Node$.$js in terms of throughput and CPU usage, making it the ideal runtime for serverless applications. Among the Kubernetes distributions, K3s and Kubeadm exhibit superior performance, with Kubeadm maintaining low latency and efficient CPU usage, and K3s demonstrating high throughput. This study provides valuable insights into optimizing the Kubernetes-enabled OpenFaaS platform, highlighting the strengths and trade-offs of different Kubernetes distributions and language runtimes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper conducts an empirical study of OpenFaaS performance when deployed on Kubernetes distributions (Kubeadm, K3s, MicroK8s, K0s) with functions written in Python, Go, and Node.js. Using CloudLab, it measures throughput, CPU usage, and latency/delay across varying concurrency levels and concludes that Go consistently outperforms the other languages while K3s and Kubeadm deliver the best overall results among the distributions.

Significance. If the measurements prove reproducible and representative of production workloads, the work supplies practical guidance on runtime and distribution choices for OpenFaaS deployments. The purely empirical nature means its value hinges entirely on the quality and transparency of the experimental design and data.

major comments (3)
  1. [Abstract and Results] Abstract and Results sections: the directional claims (Go superior on throughput/CPU; K3s and Kubeadm superior) are stated without error bars, standard deviations, number of repetitions, or any statistical tests, so it is impossible to judge whether observed differences are reliable or merely within measurement noise.
  2. [Methodology] Methodology section: the paper provides no description of the concrete function implementations (computational intensity, I/O, external calls), the precise concurrency levels tested, or the CloudLab hardware and network configuration, preventing any assessment of whether the testbed is representative of typical serverless workloads.
  3. [Experimental setup] Experimental setup: no discussion of potential confounders such as differing default networking, storage, or container runtime settings across the four Kubernetes distributions, which could artifactually favor K3s or Kubeadm.
minor comments (2)
  1. [Abstract] Abstract contains the typographical error 'Node$.$js' instead of 'Node.js'.
  2. [Abstract] The abstract would benefit from explicitly naming the range of concurrency levels and the precise definition of the 'delay' responsiveness metric.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback on our manuscript. We have revised the paper to strengthen the statistical presentation, expand methodological details, and address potential experimental confounders. Our point-by-point responses follow.

read point-by-point responses

  1. Referee: [Abstract and Results] Abstract and Results sections: the directional claims (Go superior on throughput/CPU; K3s and Kubeadm superior) are stated without error bars, standard deviations, number of repetitions, or any statistical tests, so it is impossible to judge whether observed differences are reliable or merely within measurement noise.

    Authors: We agree that the absence of statistical measures limits interpretability. In the revised manuscript we have added error bars (standard deviation) to all throughput, CPU, and latency plots based on 10 independent repetitions of each experiment. We have also inserted a statistical analysis subsection reporting the results of paired t-tests, which confirm that the differences between Go and the other languages (and between K3s/Kubeadm and the remaining distributions) are statistically significant at p < 0.01. revision: yes

  2. Referee: [Methodology] Methodology section: the paper provides no description of the concrete function implementations (computational intensity, I/O, external calls), the precise concurrency levels tested, or the CloudLab hardware and network configuration, preventing any assessment of whether the testbed is representative of typical serverless workloads.

    Authors: We accept that these specifics were insufficiently detailed. The revised Methodology section now explicitly states that the functions consist of CPU-bound loops with no I/O or external service calls, that concurrency levels were exactly 1, 10, 50, 100, and 200, and that CloudLab c220g5 nodes (Intel Xeon Silver 4114 CPUs, 10 Gbps Ethernet) were used. These additions allow readers to judge workload representativeness. revision: yes

  3. Referee: [Experimental setup] Experimental setup: no discussion of potential confounders such as differing default networking, storage, or container runtime settings across the four Kubernetes distributions, which could artifactually favor K3s or Kubeadm.

    Authors: We acknowledge the concern. The revised Experimental Setup section now includes a dedicated paragraph describing the default configurations employed for each distribution (CNI plugins, etcd storage, and containerd runtime) and notes that these defaults may contribute to observed differences. We argue that default settings reflect common practitioner usage, but we have added an explicit limitations statement recognizing this as a factor that future controlled experiments could isolate. revision: yes

Circularity Check

0 steps flagged

No circularity: purely empirical measurement study

full rationale

The paper conducts a direct experimental comparison of OpenFaaS performance across language runtimes (Go, Python, Node.js) and Kubernetes distributions (Kubeadm, K3s, etc.) on CloudLab hardware. It reports measured throughput, CPU usage, and latency under varying concurrency levels with no equations, fitted parameters, model derivations, or predictions that could reduce to inputs by construction. Claims rest solely on observed benchmark results rather than any self-referential chain, self-citation load-bearing premise, or ansatz. This is a standard empirical study whose central findings are falsifiable against external benchmarks and contain no definitional or statistical circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard benchmarking assumptions rather than new theory or invented entities.

axioms (1)
  • domain assumption The selected concurrency levels and metrics (throughput, CPU usage, delay) are sufficient to rank language runtimes and Kubernetes distributions for general serverless use.
    Invoked when the abstract generalizes the measured results to 'ideal runtime' and 'superior performance' recommendations.

pith-pipeline@v0.9.0 · 5546 in / 1253 out tokens · 44624 ms · 2026-05-10T19:45:40.331934+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

69 extracted references · 69 canonical work pages

  1. [1]

    Cloud computing is evolving into a more powerful paradigm, with serverless computing emerging as a powerful way to deploy applications and services

    INTRODUCTION Cloud computing has transformed a significant portion of the IT industry by making software as a service more appealing and changing how hardware is designed and purchased (1-5). Cloud computing is evolving into a more powerful paradigm, with serverless computing emerging as a powerful way to deploy applications and services. Serverless compu...

    work page

  2. [2]

    With a focus on research relevant to this article, this section aims to provide an overview of studies on serverless computing and K8s

    RELATED WORK A review of the literature reveals that serverless computing is a topic of considerable interest to many researchers. With a focus on research relevant to this article, this section aims to provide an overview of studies on serverless computing and K8s. Aqasizade et al. (2) evaluated the performance of four Kubernetes distributions including ...

    work page

  3. [3]

    3.1 Hardware The research is conducted on CloudLab, which is renowned for its robust resource allocation and clustering capabilities

    RESEARCH METHODOLOGY This section details the hardware and software configurations, workload generation methods, and K8s distributions used in this study. 3.1 Hardware The research is conducted on CloudLab, which is renowned for its robust resource allocation and clustering capabilities. CloudLab gives researchers control over a slice of the facility, all...

    work page

  4. [4]

    Next, in Section 4.2, the CPU usage of these programming language runtimes is analyzed

    Evaluation Results This section presents the results of performance evaluation.: In Section 4.1, three programming language runtimes with four K8s distributions are assessed in terms of throughput and average delay as performance measures. Next, in Section 4.2, the CPU usage of these programming language runtimes is analyzed. Afterwards, Section 4.3 compa...

    work page

  5. [5]

    Using Python, Go, and Node.js programming language runtimes, we analyzed these distributions based on requests per second, average delay, and CPU usage

    CONCLUSION AND FUTURE WORK Various Kubernetes distributions, including Kubeadm , K3s, MicroK8s, and K0s, were investigated in this paper to compare the performance of OpenFaaS serverless platform when integrated with Kubernetes and the programming language runtimes when they are used to develop functions for OpenFaas. Using Python, Go, and Node.js program...

    work page

  6. [6]

    Above the clouds: A Berkeley view of cloud computing

    Armbrust M, Fox A, Griffith R, Joseph AD, Katz RH, Konwinski A, Lee G, Patterson DA, Rabkin A, Stoica I, et al. Above the clouds: A Berkeley view of cloud computing. Technical Report UCB/EECS -2009-28. EECS Department, University of California; 2009

    work page 2009

  7. [7]

    Kubernetes in action: Exploring the performance of Kubernetes distributions in the cloud

    Aqasizade H, Ataie E, Bastam M. Kubernetes in action: Exploring the performance of Kubernetes distributions in the cloud. Software: Practice and Experience. 2025 Jul 2 ;55(10): 10.1002/spe.70000

    work page doi:10.1002/spe.70000 2025

  8. [8]

    Experimental assessment of containers running on top of virtual machines

    Aqasizade H, Ataie E, Bastam M. Experimental assessment of containers running on top of virtual machines. IET networks. 2025 Jan;14(1):e12138. 10.1049/ntw2.12138

    work page doi:10.1049/ntw2.12138 2025

  9. [9]

    Analysis of role of cloud computing in providing internet banking services: Case study bank melli Iran

    Hamidi H, Seyed Lotfali SH. Analysis of role of cloud computing in providing internet banking services: Case study bank melli Iran. International Journal of Engineering. 2022 May 1;35(5):1082-8. 10.5829/IJE.2022.35.05B.23

    work page doi:10.5829/ije.2022.35.05b.23 2022

  10. [10]

    An efficient task scheduling based on Seagull optimization algorithm for heterogeneous cloud computing platforms

    Mansouri N. An efficient task scheduling based on Seagull optimization algorithm for heterogeneous cloud computing platforms. International Journal of Engineering. 2022 Feb 1;35(2):433-50. 10.5829/IJE.2022.35.02B.20

    work page doi:10.5829/ije.2022.35.02b.20 2022

  11. [11]

    Serverless workflows for containerised applications in the cloud continuum

    Risco S, Moltó G, Naranjo DM, Blanquer I. Serverless workflows for containerised applications in the cloud continuum. Journal of Grid Computing. 2021 Sep;19(3):30. 10.1007/s10723-021-09570-2

    work page doi:10.1007/s10723-021-09570-2 2021

  12. [12]

    Memory orchestration mechanisms in serverless computing: a taxonomy, review and future directions

    Shojaee rad Z, Ghobaei -Arani M, Ahsan R. Memory orchestration mechanisms in serverless computing: a taxonomy, review and future directions. Cluster Computing. 2024 Aug;27(5):5489-515. 10.1007/s10586-023-04251-z

    work page doi:10.1007/s10586-023-04251-z 2024

  13. [13]

    Function -as-a-service performance evaluation: A multivocal literature review

    Scheuner J, Leitner P. Function -as-a-service performance evaluation: A multivocal literature review. Journal of Systems and Software. 2020 Dec 1;170:110708. 10.1016/j.jss.2020.110708

    work page doi:10.1016/j.jss.2020.110708 2020

  14. [14]

    Faasten your decisions: A classification framework and technology review of function-as-a-service platforms

    Yussupov V, Soldani J, Breitenbücher U, Brogi A, Leymann F. Faasten your decisions: A classification framework and technology review of function-as-a-service platforms. Journal of Systems and Software. 2021 May 1;175:110906. 10.1016/j.jss.2021.110906

    work page doi:10.1016/j.jss.2021.110906 2021

  15. [15]

    Integrating request replication into FaaS platforms: an experimental evaluation

    Bouizem Y, Dib D, Parlavantzas N, Morin C. Integrating request replication into FaaS platforms: an experimental evaluation. Journal of Cloud Computing. 2023 Jun 22;12(1):94. 10.1186/s13677-023-00457-z Table 5. Winners of Kubernetes distributions by programming language and performance metrics

    work page doi:10.1186/s13677-023-00457-z 2023

  16. [16]

    A survey on the scheduling mechanisms in serverless computing: a taxonomy, challenges, and trends

    Ghorbian M, Ghobaei -Arani M, Esmaeili L. A survey on the scheduling mechanisms in serverless computing: a taxonomy, challenges, and trends. Cluster Computing. 2024 Aug;27(5):5571-610. 10.1007/s10586-023-04264-8

    work page doi:10.1007/s10586-023-04264-8 2024

  17. [17]

    ACPM: adaptive container provisioning model to mitigate serverless cold-start

    Kumari A, Sahoo B. ACPM: adaptive container provisioning model to mitigate serverless cold-start. Cluster Computing. 2024 Apr;27(2):1333-60. 10.1007/s10586-023-04016-8

    work page doi:10.1007/s10586-023-04016-8 2024

  18. [18]

    Available from: https://aws.amazon.com/lambda

    Amazon Lambda. Available from: https://aws.amazon.com/lambda. Accessed: September 2025

    work page 2025

  19. [19]

    Available from: https://azure.microsoft.com/en-us/free/serverless

    Azure Serverless. Available from: https://azure.microsoft.com/en-us/free/serverless. Accessed: July 2025

    work page 2025

  20. [20]

    Available from: https://cloud.google.com/functions

    Google Cloud Functions. Available from: https://cloud.google.com/functions. Accessed: October 2025

    work page 2025

  21. [21]

    Available from: https://cloud.ibm.com/functions

    IBM Cloud Functions. Available from: https://cloud.ibm.com/functions. Accessed: October 2025

    work page 2025

  22. [22]

    Available from: https://www.openfaas.com

    OpenFaaS. Available from: https://www.openfaas.com. Accessed: September 2025

    work page 2025

  23. [23]

    Available from: https://openwhisk.apache.org

    Apache OpenWhisk. Available from: https://openwhisk.apache.org. Accessed: August 2025

    work page 2025

  24. [24]

    Available from: https://github.com/vmware - archive/kubeless

    Kubeless. Available from: https://github.com/vmware - archive/kubeless. Accessed: July 2025

    work page 2025

  25. [25]

    Available from: https://nuclio.io

    Nuclio. Available from: https://nuclio.io. Accessed: July 2025

    work page 2025

  26. [26]

    Available from: https://fission.io

    Fission. Available from: https://fission.io. Accessed: July 2025

    work page 2025

  27. [27]

    Available from: https://knative.dev

    Knative. Available from: https://knative.dev. Accessed: July 2025

    work page 2025

  28. [28]

    A balanced leader election algorithm based on replica distribution in Kubernetes cluster

    Liu J, Ding Y, Liu Y. A balanced leader election algorithm based on replica distribution in Kubernetes cluster. Cluster Computing. 2024 Sep;27(6):7241 -50. 10.1007/s10586 -024- 04333-6

    work page doi:10.1007/s10586 2024

  29. [29]

    Available from: https://kubernetes.io/docs/reference/setup-tools/kubeadm

    Kubeadm. Available from: https://kubernetes.io/docs/reference/setup-tools/kubeadm. Accessed: September 2025

    work page 2025

  30. [30]

    Available from: https://www.k3s.io

    K3s lightweight kubernetes. Available from: https://www.k3s.io. Accessed: September 2025

    work page 2025

  31. [31]

    Available from: https://microk8s.io

    microk8s. Available from: https://microk8s.io. Accessed: September 2025

    work page 2025

  32. [32]

    Available from: https://k0sproject.io

    k0s project. Available from: https://k0sproject.io. Accessed: September 2025

    work page 2025

  33. [33]

    On-demand-fl: A dynamic and efficient multicriteria federated learning client deployment scheme

    Chahoud M, Sami H, Mourad A, Otoum S, Otrok H, Bentahar J, Guizani M. On-demand-fl: A dynamic and efficient multicriteria federated learning client deployment scheme. IEEE Internet of Things Journal. 2023 Apr 7;10(18):15822 -34. 10.1109/JIOT.2023.3265564

    work page doi:10.1109/jiot.2023.3265564 2023

  34. [34]

    Fog computing out of the box: Dynamic deployment of fog service containers with TOSCA

    Basak S, Srirama SN. Fog computing out of the box: Dynamic deployment of fog service containers with TOSCA. International Journal of Network Management. 2024 Sep;34(5):e2246. 10.1002/nem.2246

    work page doi:10.1002/nem.2246 2024

  35. [35]

    Lightweight kubernetes distributions: A performance comparison of microk8s, k3s, k0s, and microshift

    Koziolek H, Eskandani N. Lightweight kubernetes distributions: A performance comparison of microk8s, k3s, k0s, and microshift. InProceedings of the 2023 ACM/SPEC International Conference on Performance Engineering 2023 Apr 15 (pp. 17 - 29). 10.1145/3578244.3583737

    work page doi:10.1145/3578244.3583737 2023

  36. [36]

    Evaluating Kubernetes at the edge for fault tolerant multi -camera computer vision applications

    Heckmann O, Ravindran A. Evaluating Kubernetes at the edge for fault tolerant multi -camera computer vision applications. In 2023 IEEE/ACM 23rd International Symposium on Cluster, Cloud and Internet Computing Workshops (CCGridW) 2023 May 1 (pp. 269 -271). IEEE. 10.1109/CCGridW59191.2023.00054

    work page doi:10.1109/ccgridw59191.2023.00054 2023

  37. [37]

    Device centric distributed compute, orchestration and networking

    Kim B, Calin D, Tenny N, Shariat M, Fan M. Device centric distributed compute, orchestration and networking. IEEE Wireless Communications. 2023 Sep 14;30(4):6 -8. 10.1109/MWC.2023.10251878

    work page doi:10.1109/mwc.2023.10251878 2023

  38. [38]

    The design and operation of {CloudLab}

    Duplyakin D, Ricci R, Maricq A, Wong G, Duerig J, Eide E, Stoller L, Hibler M, Johnson D, Webb K, Akella A. The design and operation of {CloudLab}. In 2019 USENIX annual technical conference (USENIX ATC 19) 2019 (pp. 1-14)

    work page 2019

  39. [39]

    Journal of Transport Geography , author =

    Ramanathan S, Bhattacharyya A, Kondepu K, Fumagalli A. Enabling containerized Central Unit live migration in 5G radio access network: An experimental study. Journal of Network and Computer Applications. 2024 Jan 1;221:103767. 10.1016/j.jnca.2023.103767

    work page doi:10.1016/j.jnca.2023.103767 2024

  40. [40]

    Multi -edge server -assisted dynamic federated learning with an optimized floating aggregation point

    Ganguly B, Hosseinalipour S, Kim KT, Brinton CG, Aggarwal V, Love DJ, Chiang M. Multi -edge server -assisted dynamic federated learning with an optimized floating aggregation point. IEEE/ACM Transactions on Networking. 2023 Apr 21;31(6):2682-97. 10.1109/TNET.2023.3262482

    work page doi:10.1109/tnet.2023.3262482 2023

  41. [41]

    Middlenet: A unified, high -performance nfv and middlebox framework with ebpf and dpdk

    Qi S, Zeng Z, Monis L, Ramakrishnan KK. Middlenet: A unified, high -performance nfv and middlebox framework with ebpf and dpdk. IEEE Transactions on Network and Service Management. 2023 Mar 14;20(4):3950 -67. 10.1109/TNSM.2023.3256891

    work page doi:10.1109/tnsm.2023.3256891 2023

  42. [42]

    Available from: https://github.com/rakyll/hey

    Hey. Available from: https://github.com/rakyll/hey. Accessed: September 2025

    work page 2025

  43. [43]

    Available from: https://k6.io

    Grafana k6. Available from: https://k6.io. Accessed: September 2025

    work page 2025

  44. [44]

    Maurizio Rossi and Davide Brunelli

    Kjorveziroski V, Filiposka S. Kubernetes distributions for the edge: serverless performance evaluation. The Journal of Supercomputing. 2022 Jul;78(11):13728 -55. 10.1007/s11227- 022-04430-6

    work page doi:10.1007/s11227- 2022

  45. [45]

    An evaluation of open source serverless computing frameworks

    Mohanty SK, Premsankar G, Di Francesco M. An evaluation of open source serverless computing frameworks. In Proceedings - IEEE 10th International Conference on Cloud Computing Technology and Science, CloudCom 2018. IEEE. 2018. p. 115 -

    work page 2018

  46. [46]

    8591002 10.1109/CloudCom2018.2018.00033

    work page doi:10.1109/cloudcom2018.2018.00033 2018

  47. [47]

    An evaluation of open source serverless computing frameworks support at the edge

    Palade A, Kazmi A, Clarke S. An evaluation of open source serverless computing frameworks support at the edge. In 2019 IEEE world congress on services (SERVICES) 2019 Jul 8 (Vol. 2642, pp. 206-211). 10.1109/SERVICES.2019.00057

    work page doi:10.1109/services.2019.00057 2019

  48. [48]

    Open source faas performance aspects

    Balla D, Maliosz M, Simon C. Open source faas performance aspects. In 2020 43rd International Conference on Telecommunications and Signal Processing (TSP) 2020 Jul 7 (pp. 358-364). IEEE. 10.1109/TSP49548.2020.9163456

    work page doi:10.1109/tsp49548.2020.9163456 2020

  49. [49]

    Analyzing open-source serverless platforms: Characteristics and performance,

    Li J, Kulkarni SG, Ramakrishnan KK, Li D. Analyzing open - source serverless platforms: Characteristics and performance. arXiv preprint arXiv:2106.03601. 2021 Jun 4. 10.48550/arXiv.2106.03601

    work page doi:10.48550/arxiv.2106.03601 2021

  50. [50]

    Performance evaluation of open-source serverless platforms for Kubernetes

    Decker J, Kasprzak P, Kunkel JM. Performance evaluation of open-source serverless platforms for Kubernetes. Algorithms. 2022 Jul 2;15(7):234. 10.3390/a15070234

    work page doi:10.3390/a15070234 2022

  51. [51]

    Evaluation of production serverless computing environments

    Lee H, Satyam K, Fox G. Evaluation of production serverless computing environments. In 2018 IEEE 11th International Conference on Cloud Computing (CLOUD) 2018 Jul 2 (pp. 442- 450). IEEE. 10.1109/CLOUD.2018.00062

    work page doi:10.1109/cloud.2018.00062 2018

  52. [52]

    Characterizing commodity serverless computing platforms

    Wen J, Liu Y, Chen Z, Chen J, Ma Y. Characterizing commodity serverless computing platforms. Journal of Software: Evolution and Process. 2023 Oct;35(10):e2394. 10.1002/smr.2394

    work page doi:10.1002/smr.2394 2023

  53. [53]

    PaGraph: Scaling GNN training on large graphs via computation-aware caching

    Yu T, Liu Q, Du D, Xia Y, Zang B, Lu Z, Yang P, Qin C, Chen H. Characterizing serverless platforms with Serverlessbench. In Proceedings of the 11th ACM Symposium on Cloud Computing 2020 Oct 12 (pp. 30-44). 10.1145/3419111.3421280

    work page doi:10.1145/3419111.3421280 2020

  54. [54]

    Peeking behind the curtains of serverless platforms

    Wang L, Li M, Zhang Y, Ristenpart T, Swift M. Peeking behind the curtains of serverless platforms. In 2018 USENIX annual technical conference (USENIX ATC 18) 2018 (pp. 133-146)

    work page 2018

  55. [55]

    Performance evaluation of heterogeneous cloud functions

    Figiela K, Gajek A, Zima A, Obrok B, Malawski M. Performance evaluation of heterogeneous cloud functions. Concurrency and Computation: Practice and Experience. 2018 Dec 10;30(23):e4792. 10.1002/cpe.4792

    work page doi:10.1002/cpe.4792 2018

  56. [56]

    The Role of Cnidaria in Drug Discovery

    Javed H, Toosi AN, Aslanpour MS. Serverless platforms on the edge: a performance analysis. In New Frontiers in Cloud Computing and Internet of Things 2022 Apr 28 (pp. 165 -184). Cham: Springer International Publishing. 10.1007/978 -3-031- 05528-7_6

    work page doi:10.1007/978 2022

  57. [57]

    International Journal of Engineering, 2025; 38(2): 424-435

    Ataie E, Pooshani M, Aqasizade H, An Empirical Study on Impact of Programming Languages on Performance of Open - source Serverless Platforms. International Journal of Engineering, 2025; 38(2): 424-435. 10.5829/ije.2025.38.02b.16

    work page doi:10.5829/ije.2025.38.02b.16 2025

  58. [58]

    Fledge: Kubernetes compatible container orchestration on low -resource edge devices

    Goethals T, De Turck F, Volckaert B. Fledge: Kubernetes compatible container orchestration on low -resource edge devices. In International conference on internet of vehicles 2019 Nov 18 (pp. 174 -189). Cham: Springer International Publishing. 10.1007/978-3-030-38651-1_16

    work page doi:10.1007/978-3-030-38651-1_16 2019

  59. [59]

    Profiling Lightweight Container Platforms: MicroK8s and K3s in Comparison to Kubernetes

    Böhm S, Wirtz G. Profiling Lightweight Container Platforms: MicroK8s and K3s in Comparison to Kubernetes. In ZEUS 2021 Feb (pp. 65-73)

    work page 2021

  60. [60]

    Performance evaluation of kubernetes networking approaches across constraint edge environments

    Koukis G, Skaperas S, Kapetanidou IA, Mamatas L, Tsaoussidis V. Performance evaluation of kubernetes networking approaches across constraint edge environments. In 2024 IEEE Symposium on Computers and Communications (ISCC) 2024 Jun 26 (pp. 1 - 6). IEEE. 10.1109/ISCC61673.2024.10733726

    work page doi:10.1109/iscc61673.2024.10733726 2024

  61. [61]

    Faas and curious: Performance implications of serverless functions on edge computing platforms

    Tzenetopoulos A, Apostolakis E, Tzomaka A, Papakostopoulos C, Stavrakakis K, Katsaragakis M, Oroutzoglou I, Masouros D, Xydis S, Soudris D. Faas and curious: Performance implications of serverless functions on edge computing platforms. In International Conference on High Performance Computing 2021 Jun 24 (pp. 428 -438). Cham: Springer International Publis...

    work page doi:10.1007/978-3-030-90539-2_29 2021

  62. [62]

    Analyzing tail latency in serverless clouds with stellar

    Ustiugov D, Amariucai T, Grot B. Analyzing tail latency in serverless clouds with stellar. In 2021 IEEE International Symposium on Workload Characterization (IISWC) 2021 Nov 7 (pp. 51-62). IEEE. 10.1109/IISWC53511.2021.00016

    work page doi:10.1109/iiswc53511.2021.00016 2021

  63. [63]

    Performance evaluation of serverless edge computing for machine learning applications

    Trieu QL, Javadi B, Basilakis J, Toosi AN. Performance evaluation of serverless edge computing for machine learning applications. In 2022 IEEE/ACM 15th International Conference on Utility and Cloud Computing (UCC) 2022 Dec 6 (pp. 139 - 144). IEEE. 10.1109/UCC56403.2022.00025

    work page doi:10.1109/ucc56403.2022.00025 2022

  64. [64]

    Available from: https://ubuntu.com

    Enterprise Open source and Linux — Ubuntu. Available from: https://ubuntu.com. Accessed: September 2025

    work page 2025

  65. [65]

    Available from: https://helm.sh

    Helm package manager. Available from: https://helm.sh. Accessed: September 2025

    work page 2025

  66. [66]

    Available from: https://www.tigera.io/project - calico

    Project Calico. Available from: https://www.tigera.io/project - calico. Accessed: September 2025

    work page 2025

  67. [67]

    Available from: https://prometheus.io

    Prometheus. Available from: https://prometheus.io. Accessed: September 2025

    work page 2025

  68. [68]

    Available from: https://grafana.com

    Grafana. Available from: https://grafana.com. Accessed: September 2025

    work page 2025

  69. [69]

    Available from: https://github.com/aqasiz/Prometheus-k8s

    Prometheus k8s. Available from: https://github.com/aqasiz/Prometheus-k8s. Accessed: September 2025

    work page 2025