Abstract
We consider service requests that arrive in an online fashion in Software-Defined Networks (SDNs) with network function virtualization (NFV). Each request is a flow with a high-level specification of routing and processing (by network functions) requirements. Each network function can be performed by a specified subset of servers in the system. The algorithm needs to decide whether to reject the request, or accept it and with a specific routing and processing assignment, under given capacity constraints (solving the path computation and function placement problems). Each served request is assumed to “pay” a pre-specified benefit and the goal is to maximize the total benefit accrued. In this paper we first formalize the problem, and propose a new service model that allows us to cope with requests with unknown duration without preemption. The new service model augments the traditional accept/reject schemes with a new possible response of “stand by.” We also present a new expressive model to describe requests abstractly using a “plan” represented by a directed graph. Our algorithmic result is an online algorithm for path computation and function placement that guarantees, in each time step, throughput of at least a logarithmic fraction of a (very permissive) upper bound on the maximal possible benefit.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Typically, \(k\) is a small constant. In particular, the number of processing stages \((k)\) does not grow as a function of the size \((n)\) of the network.
Our pr-graphs are similar to Merlin’s regular expressions [16], but are more expressive and, in our humble opinion, are more natural to design.
The request model can be extended to deal with nonuniform demands. Namely, a possibly different amount of resource is consumed by the request \(r_{j}\) from each pr-graph element. Formally, \(d_{j}: X_{j}\cup Y_{j} \rightarrow {\mathbb {N}}\) specifies the demand of request \(r_{j}\) from each edge and node of the the pr-graph. This extension can capture varying bandwidth due to compression, asymmetric servers whose processing ability depends on the task at hand, etc.
In Merlin, the input may also contain a “policing” function of capping the maximal bandwidth of a connection. We focus on resource allocation only. Policing may be enforced by an orthogonal entity.
One could also allow for empty events \(\sigma _{j}\) to deal with the case that there is no new request and no active requests wishes to depart.
In case no departure event of a request has occurred, then the corresponding request stays forever.
A shortest-path algorithm suffices for the implementation of Step 2, that is, if the shortest path is light enough then a single path is required (see Line 31 for the invocation of the oracle), otherwise, if it is not light enough then Pj = ∅.
We assume that \(b_{\max }\) is known in advance.
A fractional allocation may serve a request partially, split the assignment among multiple allocations, and change the allocation over time. Formally, a fractional allocation assigns to each request \(r_{j}\) in each time step \(t\) a non-negative linear combination of valid allocations \(\sum _{i}\alpha _{j,t,i} \cdot p_{j,i}\), where (for every \(j,t,i\)) \(\alpha _{j,t,i}\geq 0\), \(p_{j,i}\in {\varGamma }_{j}\), and \(\sum _{i} \alpha _{j,t,i} \leq 1\). Capacity constraints are satisfied in the sense that \(\sum _{j} \sum _{i: e\in p_{j,i}} d_{j}\cdot \alpha _{j,t,i} \leq c_{e}\). The benefit obtained by a fractional allocation is \(\sum _{t} \sum _{j} \sum _{i} b_{j}\cdot \alpha _{j,t,i}\).
This benefit is paid only once, if a request is accepted, as opposed to the benefit per time step paid by SDN requests.
In fact it is proven in [2, Lemmas 4.4, 4.5], for the case of (simple) routing (see first example after Definition 3), that if \(\frac {\max _{j} d_{j}}{\min _{z}c_{z}} \leq \frac {1}{\gamma }\), then the competitive ratio is \({\varOmega }(n^{1/\gamma })\).
References
Aspnes, J., Azar, Y., Fiat, A., Plotkin, S., Waarts, O.: On-line routing of virtual circuits with applications to load balancing and machine scheduling. J. ACM) 44(3), 486–504 (1997)
Awerbuch, B., Azar, Y., Plotkin, S.: Throughput-competitive on-line routing. In: Proc.34th IEEE Annual Symp. on Foundations of Computer Science, pp. 32–40 (1993)
Awerbuch, B., Azar, Y., Plotkin, S., Waarts, O.: Competitive routing of virtual circuits with unknown duration. J. Comput. Syst. Sci. 62(3), 385–397 (2001)
Buchbinder, N., Naor, J.S.: Improved bounds for online routing and packing via a primal-dual approach. In: 47th Ann. IEEE Symp.on Foundations of Computer Science, pp. 293–304 (2006)
Buchbinder, N., Naor, J.S.: Online primal-dual algorithms for covering and packing. Math. Oper. Res. 34(2), 270–286 (2009)
Cohen, R., Lewin-Eytan, L., Naor, J. S., Raz, D.: Near optimal placement of virtual network functions. In: 2015 IEEE Conference on Computer Communications (INFOCOM), pp. 1346–1354 (2015)
Even, G., Medina, M.: Online multi-commodity flow with high demands. In: Proc. 10th Int.Workshop on Approximation and Online Algorithms (WAOA), pp 16–29 (2012)
Even, G, Medina, M., Patt-Shamir, B.: On-line path computation and function placement in sdns. In: Bonakdarpour, B., Petit, F. (eds.) Stabilization, Safety, and Security of Distributed Systems - 18th International Symposium, SSS 2016, Lyon, France, November 7-10, 2016, Proceedings, volume 10083 of Lecture Notes in Computer Science, pp. 131–147 (2016)
Even, G., Medina, M., Schaffrath, G., Schmid, S.: Competitive and deterministic embeddings of virtual networks. Theor. Comput. Sci. 496(0), 184–194 (2013). Distributed Computing and Networking (ICDCN 2012)
Even, G., Rost, M., Schmid, S.: An approximation algorithm for path computation and function placement in sdns. arXiv:1603.09158/ 2016 Appeared in SIROCCO (2016)
Hopcroft, J.E., Motwani, R., Ullman, J.D.: Introduction to Automata Theory, Languages, and Computation), 3rd edn. Pearson, Boston (2006)
Kleinberg, J.M.: Approximation algorithms for disjoint paths problems. PhD thesis Massachusetts Institute of Technology (1996)
Kreutz, D., Ramos, F., Esteves Verissimo, P., Esteve Rothenberg, C., Azodolmolky, S., Uhlig, S.: Software-defined networking: A comprehensive survey. Proc. IEEE 103(1), 14–76 (2015)
Sleator, D.D., Tarjan, R.E.: Amortized efficiency of list update and paging rules. Commun. ACM 28(2), 202–208 (1985)
Soulé, R., Basu, S., Kleinberg, R., Sirer, E.G., Foster, N.: Managing the network with merlin. In: Proc. 12th ACM Workshop on Hot Topics in Networks. ACM, p. 24 (2013)
Soulé, R., Basu, S., Marandi, P.J., Pedone, F., Kleinberg, R., Sirer, E.G., Foster, N.: Merlin: A language for provisioning network resources. In: Proc. 10th ACM Int. Conf. on emerging Networking Experiments and Technologies. ACM, pp. 213–226 (2014)
Acknowledgments
This work was supported in part by the Neptune Consortium, Israel. Preliminary versions of this manuscript appeared in the proceedings of SSS-2016 [8].
Author information
Authors and Affiliations
Corresponding author
Additional information
This article is part of the Topical Collection on Special Issue on Stabilization, Safety, and Security of Distributed Systems (SSS 2016)
Rights and permissions
About this article
Cite this article
Even, G., Medina, M. & Patt-Shamir, B. On-Line Path Computation and Function Placement in SDNs. Theory Comput Syst 63, 306–325 (2019). https://doi.org/10.1007/s00224-018-9863-4
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00224-018-9863-4