Multilayer Virtualized Systems Analysis with Kernel Tracing

Cedric Biancheri; Naser Ezzati-Jivan; Michel R. Dagenais

doi:10.1109/w-ficloud.2016.18

2016 · IEEE Future Internet of Things and Cloud Workshops (FiCloudW)

Multilayer Virtualized Systems Analysis with Kernel Tracing

Cedric Biancheri | Naser Ezzati-Jivan | Michel R. Dagenais

Evidence basis: full-text-reviewed · Review status: catalog-reviewed; paper-author approval pending

kernel-tracing system-tracing performance-analysis resource-analysis

virtualized systems kernel tracing multilayer analysis virtualization overhead LTTng KVM Trace Compass Fused Virtual Machine view State History Tree vmentry vmexit

Core contribution: The paper fuses synchronized host and virtual-machine kernel traces into a Fused Virtual Machine view that exposes how VM threads, virtual CPUs, hypervisor transitions, and physical CPUs interact.

Problem and motivation

Virtual-machine boundaries obscure direct and indirect interactions between guest activity, hypervisor transitions, and host scheduling or I/O. A process can appear uninterrupted inside a guest while being preempted or delayed on the host.

Method and contribution

Collect synchronized LTTng kernel traces from hosts and VMs, align them using hypercalls and incremental convex-hull synchronization, and fuse them into a state-history representation. The Fused Virtual Machine view maps guest vCPUs onto physical CPUs and exposes cross-layer state.

Findings and evidence

In the three-VM use cases, the fused view reveals host preemption, process migration, vmentry/vmexit transitions, and physical I/O handling that are invisible or ambiguous in a VM-only view.

Limitations and future directions

Limitations: Evaluation is qualitative and use-case based. Containers are explicitly outside scope. The inspected paper does not report hardware, OS/kernel versions, trace volumes, run counts, timing results, or quantitative detection accuracy.

Future work: Reuse existing analysis tools, parallelize fused analysis, and extend the approach to Java, Python, containers, and interactions among VMs on different nodes, as stated in the conclusion.

Sources and identifiers

Published version published
Public full text · PDF public_full_text

When to cite this paper

Cite this paper when your work uses or compares synchronized host/guest kernel tracing that explains VM-level behavior using physical-CPU context.

For synchronized host/guest kernel tracing that explains VM-level behavior using physical-CPU context.
For a fused VM/physical-resource view that reveals host preemption or migration hidden by guest-only control-flow views.
For the methodological distinction between independently parallelizable layer analyses and a correlated fused analysis.

Citation

@inproceedings{ezzatiJivan2016multilayervirtualized,
  author = {Cedric Biancheri and Naser Ezzati-Jivan and Michel R. Dagenais},
  title = {Multilayer Virtualized Systems Analysis with Kernel Tracing},
  year = {2016},
  booktitle = {IEEE Future Internet of Things and Cloud Workshops (FiCloudW)},
  pages = {1-6},
  publisher = {IEEE},
  doi = {10.1109/w-ficloud.2016.18},
  url = {https://doi.org/10.1109/w-ficloud.2016.18}
}

Other citation formats for Word and reference managers

Biancheri, C., Ezzati-Jivan, N., & Dagenais, M. R. (2016). Multilayer Virtualized Systems Analysis with Kernel Tracing. In IEEE Future Internet of Things and Cloud Workshops (FiCloudW) (pp. 1-6). https://doi.org/10.1109/w-ficloud.2016.18

C. Biancheri, N. Ezzati-Jivan, and M. R. Dagenais, "Multilayer Virtualized Systems Analysis with Kernel Tracing," in IEEE Future Internet of Things and Cloud Workshops (FiCloudW), pp. 1-6, 2016, doi: 10.1109/w-ficloud.2016.18

Readable Markdown record · JSON record · Download RIS