NFVPerf: Online performance monitoring and bottleneck detection for NFV

Naik, Priyanka; Shaw, Dilip Kumar; Vutukuru, Mythili

doi:10.1109/nfv-sdn.2016.7919491

Cited by 38 publications

(11 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, research has shown that this information might not always be sufficient to reliably identify performance bottlenecks [9]. To alleviate this issue, the approach proposed in [7] suggests to monitor VM-to-VM communication to enable online performance monitoring of network functions. However, port mirroring of all incoming and outgoing packets is required by this solution, which imposes a significant performance overhead in a field where processing times and efficiency are crucial.…”

Section: Background and Related Workmentioning

confidence: 99%

“…Online VNF agnostic Remark SymPerf [5] Offline Code required Arbitrary workload Gym Framework [6] Offline Yes Fully automated, user defined test cases NFVPerf [7] Online Yes Port mirroring required, imposing performance overhead SONATA [8] Online Yes Uses limited information exposed by Linux Kernel, may utilize KOMon reported data KOMon Online Yes Allows monitoring of VNF health and characterization of processing time distribution and are ready to be sent out, the second measurement point monitors outgoing packets and compares payload hashes to the oldest packet still in the ring buffer. If the hashes match, the timestamps are simply subtracted and the processing time is stored as a result.…”

Section: Approachmentioning

confidence: 99%

See 1 more Smart Citation

KOMon — Kernel-based Online Monitoring of VNF Packet Processing Times

Geissler

Lange

Wamser

et al. 2019

2019 International Conference on Networked Systems (NetSys)

View full text Add to dashboard Cite

The ongoing softwarization of networks comes with several advantages like cost efficiency, increased scalability, and better flexibility by migrating functionality from static, application specific hardware appliances to flexible, lightweight software solutions running on COTS hardware. In order to maximize the performance gains promised by the NFV paradigm, several challenges remain to be solved. In this work, we address the accurate acquisition of key performance indicators, specifically the packet processing time, of softwarized network functions. To this end, we present KOMon, a Kernel-based online monitoring tool to reliably measure the packet processing times of network functions through lightweight in-stack monitoring. We show that KOMon reports highly accurate values in different scenarios and discuss the applicability of the proposed mechanism for different use cases.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

Section: Approachmentioning

confidence: 99%

KOMon — Kernel-based Online Monitoring of VNF Packet Processing Times

Geissler

Lange

Wamser

et al. 2019

2019 International Conference on Networked Systems (NetSys)

View full text Add to dashboard Cite

show abstract

“…The cost that each NFV-R associates to the announced prefix, i.e. the IGP cost to reach such an address, represents the vSF state and can be based on a plethora of parameters [16], [17] (e.g., its available capacities, its load etc.). Note however that the used metrics (link and service) should be in the same order of magnitude.…”

Section: Distributed Chaining With Igp Service Augmentationmentioning

confidence: 99%

Let there be Chaining: How to Augment your IGP to Chain your Services

Wion¹,

Bouet²,

Iannone³

et al. 2019

2019 IFIP Networking Conference (IFIP Networking)

View full text Add to dashboard Cite

Ever since Network Functions Virtualization has replaced dedicated appliances, ISPs have been able to add a degree of flexibility in their traffic engineering. However, it also has increased the complexity of the optimization problem, because it is now necessary to place virtual functions and route traffic jointly. Insofar, a logically centralized approach has been taken, where a so-called orchestrator, having full knowledge of the network, the virtual functions, and the traffic, run complex algorithms to find a suitable solution to the problem. The outcome of the algorithms are then translated to network configurations to be pushed to all of the appliances. We argue that there is no need to fully centralize every decision, rather we can leverage existing network intelligence to achieve the same goal. In particular we propose to augment the routing layer with the notion of services, so to rely on the robustness and scalability of Interior Gateway Protocols (IGP). Our solution leverages on existing distributed routing protocols where, in addition, autonomous nodes announce information about the virtual services they provide. Our design is modular and incrementally deployable and has been implemented in what we call a NFV Router. In our evaluation, we show that (i) NFV Routers distributed chaining decisions are close to optimal centrally-computed paths, (ii) on a large scale testbed deployment, NFV Routers efficiently steer traffic through chains and only add a small overhead to control traffic and (iii) our distributed system, because of its local control loop, has a faster reaction to network events than centralized solutions.

show abstract

“…With the adoption of software NFs, there exist several proposals for online performance diagnosis systems. NFVPerf [26] leverages passive traffic monitoring to identify both hardware and software bottlenecks in Virtualized NFs. PerfSight [41] leverages low-level packet processing performance metrics to detect and diagnose performance problems.…”

Section: Performance Evaluation and Diagnosismentioning

confidence: 99%

Automated synthesis of adversarial workloads for network functions

Pedrosa

Iyer

Zaostrovnykh

et al. 2018

Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication

View full text Add to dashboard Cite

Software network functions promise to simplify the deployment of network services and reduce network operation cost. However, they face the challenge of unpredictable performance. Given this performance variability, it is imperative that during deployment, network operators consider the performance of the NF not only for typical but also adversarial workloads. We contribute a tool that helps solve this challenge: it takes as input the LLVM code of a network function and outputs packet sequences that trigger slow execution paths. Under the covers, it combines directed symbolic execution with a sophisticated cache model to look for execution paths that incur many CPU cycles and involve adversarial memory-access patterns. We used our tool on 11 network functions that implement a variety of data structures and discovered workloads that can in some cases triple latency and cut throughput by 19% relative to typical testing workloads.

show abstract

NFVPerf: Online performance monitoring and bottleneck detection for NFV

Cited by 38 publications

References 15 publications

KOMon — Kernel-based Online Monitoring of VNF Packet Processing Times

KOMon — Kernel-based Online Monitoring of VNF Packet Processing Times

Let there be Chaining: How to Augment your IGP to Chain your Services

Automated synthesis of adversarial workloads for network functions

Contact Info

Product

Resources

About