WAN bandwidth remains a constrained resource that is economically infeasible to substantially overprovision. Hence, it is important to allocate capacity according to service priority and based on the incremental value of additional allocation. For example, it may be the highest priority for one service to receive 10Gb/s of bandwidth but upon reaching such an allocation, incremental priority may drop sharply favoring allocation to other services. Motivated by the observation that individual flows with fixed priority may not be the ideal basis for bandwidth allocation, we present the design and implementation of Bandwidth Enforcer (BwE), a global, hierarchical bandwidth allocation infrastructure. BwE supports: i) service-level bandwidth allocation following prioritized bandwidth functions where a service can represent an arbitrary collection of flows, ii)independent allocation and delegation policies according to user-defined hierarchy, all accounting for a global view of bandwidth and failure conditions, iii) multi-path forwarding common in traffic-engineered networks, and iv) a central administrative point to override (perhaps faulty) policy during exceptional conditions. BwE has delivered more service efficient bandwidth utilization and simpler management in production for multiple years.

We present the design, implementation, and evaluation of B4, a private WAN connecting Google's data centers across the planet. B4 has a number of unique characteristics: i) massive bandwidth requirements deployed to a modest number of sites, ii) elastic traffic demand that seeks to maximize average bandwidth, and iii) full control over the edge servers and network, which enables rate limiting and demand measurement at the edge.

Replicating content across a geographically distributed set of servers and redirecting clients to the closest server in terms of latency has emerged as a common paradigm for improving client performance. In this paper, we analyze latencies measured from servers in Google's content distribution network (CDN) to clients all across the Internet to study the effectiveness of latency-based server selection. Our main result is that redirecting every client to the server with least latency does not suffice to optimize client latencies. First, even though most clients are served by a geographically nearby CDN node, a sizeable fraction of experience latencies several tens of milliseconds higher than other in the same region. Second, we find that queueing delays often override the benefits of a client interacting with a nearby server.

We consider the problem of routing in a delay tolerant network (DTN) in the presence of path failures. Previous work on DTN routing has focused on using precisely known network dynamics, which does not account for message losses due to link failures, buffer overruns, path selection errors, unscheduled delays, or other problems. We show how to split, replicate, and erasure code message fragments over multiple delivery paths to optimize the probability of successful message delivery. We provide a formulation of this problem and solve it for two cases: a 0/1 (Bernoulli) path delivery model where messages are either fully lost or delivered, and a Gaussian path delivery model where only a fraction of a message may be delivered. Ideas from the modern portfolio theory literature are borrowed to solve the underlying optimization problem. Our approach is directly relevant to solving similar problems that arise in replica placement in distributed file systems and virtual node placement in DHTs. In three different simulated DTN scenarios covering a wide range of applications, we show the effectiveness of our approach in handling failures.

We formulate the delay-tolerant networking routing problem, where messages are to be moved end-to-end across a connectivity graph that is time-varying but whose dynamics may be known in advance. The problem has the added constraints of finite buffers at each node and the general property that no contemporaneous end-to-end path may ever exist. This situation limits the applicability of traditional routing approaches that tend to treat outages as failures and seek to find an existing end-to-end path. We propose a framework for evaluating routing algorithms in such environments. We then develop several algorithms and use simulations to compare their performance with respect to the amount of knowledge they require about network topology. We find that, as expected, the algorithms using the least knowledge tend to perform poorly. We also find that with limited additional knowledge, far less than complete global knowledge, efficient algorithms can be constructed for routing in such environments. To the best of our knowledge this is the first such investigation of routing issues in DTNs.

We present an algorithm for translating XSLT programs into SQL. Our context is that of virtual XML publishing, in which a single XML view is defined from a relational database, and subsequently queried with XSLT programs. Each XSLT program is translated into a single SQL query and run entirely in the database engine. Our translation works for a large fragment of XSLT, which we define, that includes descendant/ancestor axis, recursive templates, modes, parameters, and aggregates. We put considerable effort in generating correct and efficient SQL queries and describe several optimization techniques to achieve this efficiency. We have tested our system on all 22 SQL queries of the TPC-H database benchmark which we represented in XSLT and then translated back to SQL using our translator.