Efficient Cloud Tracing: From Very High Level to Very Low Level

Yves J. Bationo; Naser Ezzati-Jivan; Michel R. Dagenais

doi:10.1109/icce.2018.8326353

2018 · IEEE International Conference on Consumer Electronics (ICCE)

Efficient Cloud Tracing: From Very High Level to Very Low Level

Yves J. Bationo | 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 observability

cloud tracing high-level tracing low-level tracing cross-layer observability OpenStack Nova QEMU KVM Trace Compass live migration

Core contribution: The paper correlates OpenStack Nova, QEMU/KVM, and host/guest kernel evidence to diagnose live-migration behavior across cloud-system layers.

Problem and motivation

A single cloud-management layer cannot explain migration downtime, resource contention, or the host/guest causes of a slow or failed VM migration.

Method and contribution

Build a multi-layer trace from Nova state/log events, QEMU/KVM migration stages, and host/guest LTTng kernel events; synchronize controller, source, and destination traces and inspect the result in Trace Compass.

Findings and evidence

The layered view connects Nova lifecycle states to QEMU migration stages and kernel scheduling, CPU, memory, I/O, and wait evidence. A colocated stress workload exposes slower migration and resource contention that high-level Nova events cannot explain.

Limitations and future directions

Limitations: The evaluation is a small live-migration use-case study with missing hardware/software/run-count metadata. Table III conflicts with the surrounding prose for the middle experiment; integration should preserve the table values and flag the conflict rather than normalize it.

Future work: Add richer Neutron and OpenDaylight/NFV request-flow evidence to cover the network-service path that is insufficiently represented in the current trace.

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 correlating OpenStack Nova lifecycle events with QEMU/KVM live-migration stages and kernel resource evidence.

For correlating OpenStack Nova lifecycle events with QEMU/KVM live-migration stages and kernel resource evidence.
For a multi-layer cloud-tracing workflow that exposes contention hidden by Nova-level logs.
For the explicit distinction between high-level orchestration state and low-level QEMU/kernel evidence when diagnosing migration downtime or failure.

Citation

@inproceedings{ezzatiJivan2018efficientcloud,
  author = {Yves J. Bationo and Naser Ezzati-Jivan and Michel R. Dagenais},
  title = {Efficient Cloud Tracing: From Very High Level to Very Low Level},
  year = {2018},
  booktitle = {IEEE International Conference on Consumer Electronics (ICCE)},
  pages = {1-6},
  publisher = {IEEE},
  doi = {10.1109/icce.2018.8326353},
  url = {https://doi.org/10.1109/icce.2018.8326353}
}

Other citation formats for Word and reference managers

Bationo, Y. J., Ezzati-Jivan, N., & Dagenais, M. R. (2018). Efficient Cloud Tracing: From Very High Level to Very Low Level. In IEEE International Conference on Consumer Electronics (ICCE) (pp. 1-6). https://doi.org/10.1109/icce.2018.8326353

Y. J. Bationo, N. Ezzati-Jivan, and M. R. Dagenais, "Efficient Cloud Tracing: From Very High Level to Very Low Level," in IEEE International Conference on Consumer Electronics (ICCE), pp. 1-6, 2018, doi: 10.1109/icce.2018.8326353

Readable Markdown record · JSON record · Download RIS