Robert Grandl

OpenNF: enabling innovation in network function control

Network functions virtualization (NFV) together with software-defined networking (SDN) has the potential to help operators satisfy tight service level agreements, accurately monitor and manipulate network traffic, and minimize operating expenses. However, in scenarios that require packet processing to be redistributed across a collection of network function (NF) instances, simultaneously achieving all three goals requires a framework that provides efficient, coordinated control of both internal NF state and network forwarding state. To this end, we design a control plane called OpenNF. We use carefully designed APIs and a clever combination of events and forwarding updates to address race conditions, bound overhead, and accommodate a variety of NFs. Our evaluation shows that OpenNF offers efficient state control without compromising flexibility, and requires modest additions to NFs.

Quickr: Lazily Approximating Complex AdHoc Queries in BigData Clusters

We present a system that approximates the answer to complex ad-hoc queries in big-data clusters by injecting samplers on-the-fly and without requiring pre-existing samples. Improvements can be substantial when big-data queries take multiple passes over data and when samplers execute early in the query plan. We present a new, universe, sampler which is able to sample multiple join inputs. By incorporating samplers natively into a cost-based query optimizer, we automatically generate plans with appropriate samplers at appropriate locations. We devise an accuracy analysis method using which we ensure that query plans with samplers will not miss groups and that aggregate values are within a small ratio of their true value. An implementation on a cluster with tens of thousands of machines shows that queries in the TPC-DS benchmark use a median of 2X fewer resources. In contrast, approaches that construct input samples even when given 10X the size of the input to store samples improve only 22% of the queries, i.e., a median speed up of 0X.

Design and implementation of a framework for software-defined middlebox networking

Middleboxes (MBs) are used widely to ensure security (e.g., intrusion detection systems), improve performance (e.g., WAN optimizers), and provide other novel network functionality [4, 6]. Recently, researchers have proposed several new architectures for MB deployment, including Stratos [2], CoMb [4], and APLOMB [6]. These frameworks all advocate dynamic deployment of software-based MBs with the goal of increasing flexibility, improving efficiency, and reducing management overhead. However, approaches for controlling the behavior of MBs (i.e., how MBs examine and modify network traffic) remain limited. Today, configuration policies and parameters are manipulated using narrow, MB-specific configuration interfaces, while internal algorithms and state are completely inaccessible and unmodifiable. This apparent lack of finegrained control over MBs and their state precludes correct and performant implementation of control scenarios that involve re-allocating live flows across MBs: e.g., server migration, scale up/down of MBs to meet cost-performance trade-offs, recovery from network or MB failures, etc. Several key requirements must be satisfied to effectively support the above scenarios. To illustrate these requirements, we consider a scenario where MB instances are added and removed based on current network load [2] (Figure 1). When scaling up, some in-progress flows may need to be moved to a new MB instance to reduce the load on the original instance. To preserve the correctness and fidelity of MB operations, the new instance must receive the internal MB state associated with the moved flows, while the old instance still has the internal state associated with the remaining flows. For some MBs (e.g., an intrusion prevention

FCP: a flexible transport framework for accommodating diversity

Transport protocols must accommodate diverse application and network requirements. As a result, TCP has evolved over time with new congestion control algorithms such as support for generalized AIMD, background flows, and multipath. On the other hand, explicit congestion control algorithms have been shown to be more efficient. However, they are inherently more rigid because they rely on in-network components. Therefore, it is not clear whether they can be made flexible enough to support diverse application requirements. This paper presents a flexible framework for network resource allocation, called FCP, that accommodates diversity by exposing a simple abstraction for resource allocation. FCP incorporates novel primitives for end-point flexibility (aggregation and preloading) into a single framework and makes economics-based congestion control practical by explicitly handling load variations and by decoupling it from actual billing. We show that FCP allows evolution by accommodating diversity and ensuring coexistence, while being as efficient as existing explicit congestion control algorithms.

Harmony: coordinating network, compute, and storage in software-defined clouds

The progress of a big data job is often a function of storage, networking and processing. Hence, for efficient job execution, it is important to collectively optimize all three components. Prior proposals [1], in contrast, have focused on mainly on one or two of the three components. This narrow focus constraints the extent to which these proposals can support efficient operation of big data applications.

Supporting network evolution and incremental deployment with XIA

eXpressive Internet Architecture (XIA) [1] is an architecture that natively supports multiple communication types and allows networks to evolve their abstractions and functionality to accommodate new styles of communication over time. XIA embeds an elegant mechanism for handling unforeseen communication types for legacy routers. In this demonstration, we show that XIA overcomes three key barriers in network evolution (outlined below) by (1) allowing end-hosts and applications to start using new communication types (e.g., service and content) before the network supports them, (2) ensuring that upgrading a subset of routers to support new functionalities immediately benefits applications, and (3) using the same mechanisms we employ for 1 and 2 to incrementally deploy XIA in IP networks.