As the adoption of Virtualized Radio Access Networks (vRAN) is gaining momentum in 5 G networks, Mobility Network Operators are considering a Centralized RAN (CRAN) architecture that moves the baseband functions to a far-edge cloud in order to gain dimensioning flexibility, resiliency and improved RAN performance. However, there have been limited studies on the benefits of centralization in improving RAN compute utilization, especially in the context of pooling the compute-intensive Distributed Unit (DU) resources. In this paper, we present the first study on the benefits of pooling in improving DU server utilization. Using longitudinal traces from a real-world 5G network, we show that significant Capex and Opex gains of $\mathbf{8 4 \%}$ and $\mathbf{9 4 \%}$, respectively, can be obtained through fine-grained pooling at a granularity of 1 second. We also present an affinitybased and dynamic pooling algorithm that can reduce the pooling overheads while still achieving significant pooling gains.

As virtualization of Radio Access Networks (RAN) gains momentum, understanding the impact of hardware and software disaggregation on resiliency becomes critical to meet the high availability requirements of mobile networks. Our paper presents an analytical model, using continuous time Markov chains, to study the impact of virtualization and disaggregation on RAN availability. Our evaluation, assuming typical parameter value ranges for failure and recovery rates, points to container-ized platform reliability as a constraint on vRAN availability. We also find that with active-passive replication, increasing hardware replication factor beyond 2 may not bring any benefits unless failover times are reduced. We also compare the reliability of centralized and distributed virtualized central units.

Causal consistency has emerged as an attractive middle-ground to architecting cloud storage systems, as it allows for high availability and low latency, while supporting semantics stronger than eventual consistency. However, causally-consistent cloud storage systems have seen limited deployment in practice. A key factor is these systems employ full replication of all the data in all the data centers (DCs), incurring high cost. A simple extension of current causal systems to support partial replication by clustering DCs into rings incurs availability and latency problems. We propose Karma, the first system to enable causal consistency for partitioned data stores while achieving the cost advantages of partial replication without the availability and latency problems of the simple extension. Our evaluation with 64 servers emulating 8 geo-distributed DCs shows that Karma (i) incurs much lower cost than a fully-replicated causal store (obviously due to the lower replication factor); and (ii) offers higher availability and better performance than the above partial-replication extension at similar costs.

It is common for storage systems designed to run on edge datacenters to avoid the high latencies associated with geo-distribution by relying on eventually consistent models to replicate data. Eventual consistency works well for many edge applications because as long as the client interacts with the same replica, the storage system can provide session consistency, a stronger consistency model that has two additional important properties: (i) read-your-writes, where subsequent reads by a client that has updated an object will return the updated value or a newer one; and, (ii) monotonic reads, where if a client has seen a particular value for an object, subsequent reads will return the same value or a newer one. While session consistency does not guarantee that different clients will perceive updates in the same order, it nevertheless presents each individual client with an intuitive view of the world that is consistent with the client's own actions. Unfortunately, these consistency guarantees break down when a client interacts with multiple replicas housed on different datacenters over time, either as a result of application partitioning, or client or code mobility. SessionStore is a datastore for fog/edge computing that ensures session consistency on a top of otherwise eventually consistent replicas. SessionStore enforces session consistency by grouping related data accesses into a session, and using a session-aware reconciliation algorithm to reconcile only the data that is relevant to the session when switching between replicas. This approach reduces data transfer and latency by up to 90% compared to full replica reconciliation.

Many geo-distributed services at web-scale companies still rely on databases (DBs) primarily optimized for single-site performance. At AT&T this is exemplified by services in the network control plane that rely on third-party software that uses DBs like MariaDB and PostgreSQL, which do not provide strict serializability across sites without a significant performance impact. Moreover, it is often impractical for these services to re-purpose their code to use newer DBs optimized for geo-distribution. In this paper, a novel drop-in solution for DB clustering across sites called Metric is presented that can be used by services without changing a single line of code. Metric leverages the single-site performance of an existing service's DB and combines it with a cross-site clustering solution based on an entry-consistent redo log that is specifically tailored for geo-distribution. Detailed correctness arguments are presented and extensive evaluations with various benchmarks show that Metric outperforms other solutions for the access patterns in our production use-cases where service replicas access different tables on different sites. In particular, Metric achieves up to 56% less latency and 5.2x higher throughput than MariaDB and PostgreSQL clustering, and up to 90% less latency and 26x higher throughput than CockroachDB and TiDB, systems that are designed to support geo-distribution.

Finding nodes which match certain criteria, based on potentially highly dynamic information, is a critical need in many distributed systems, ranging from cloud management, to network service deployments, to emerging IoT applications. With the increasing scale, dynamicity, and richness of data, existing systems, which typically implement a custom solution based around message queues where nodes push status to a central database, are ill-suited for this purpose. In this paper, we present FOCUS, a general and scalable service which easily integrates into existing and emerging systems to provide this fundamental capability. FOCUS utilizes a gossip-based protocol for nodes to organize into groups based on attributes and current value. With this approach, nodes need not synchronize with a central database, and instead the FOCUS service only needs to query the sub-set of nodes which have the potential to positively match a given query. We show FOCUS's flexibility through an operational example of complex querying for Virtual Network Functions instantiation over cloud sites, and illustrate its ease of integration by replacing the push-based approach in OpenStack's placement service. Our evaluation demonstrates a 5-15× reduction in bandwidth consumption and an ability to scale much better than existing approaches.

Network clouds are typically private clouds owned by the network provider, consisting of a large number of geo-distributed sites with heterogeneous capabilities and small capacities. Each of these small clouds often run specialized service chains of Virtual Network Functions (VNFs), which need to meet strict Service Level Objectives (SLOs), especially along the lines of availability (e.g., First responder services). Hence, VNFs in such thinly provisioned clouds may need to be moved (rehomed), both within and across sites, much more frequently than in traditional public clouds (like Amazon's EC2 cloud), in order to meet the performance SLOs, when reacting to various cloud events like hotspots, interference from co-located VMs, failures and upgrades. Rehoming is also required by the infrastructure (platform) providers for various other reasons such as consolidation of resources for saving energy and improving the platform utilization. In this paper, we propose a model-based approach to show that naive strategies for rehoming, applied uniformly across all VNFs of the service chain, are often sub-optimal when considering different metrics like user-perceived service disruption time and the time taken to complete the rehoming action. Our model leverages the transparency between the services and platforms on private clouds (grayness), and provides appropriate rehoming recommendations based on various factors including service characteristics and runtime platform dynamics. We validate our models using a simple, yet ubiquitously deployed service chain, and using out-of-the-box rehoming options provided by Openstack, the most commonly used open-source cloud. Our results show that our graybox approach is able to achieve significant reductions in service disruption times and time taken for the rehoming action.

No abstract available.

Despite much recent progress, Web page latencies over cellular networks remain much higher than those over wired networks. Proxies that execute Web page JavaScript (JS) and push objects needed by the client can reduce latency. However, a key concern is the scalability of the proxy which must execute JS for many concurrent users. In this paper, we propose to scale the proxies, focusing on a design where the proxy's execution is solely to push the needed objects and the client completely executes the page as normal. Such redundant execution is a simple, yet effective approach to cutting network latencies, which dominate page load delays in cellular settings. We develop whittling, a technique to identify and execute in the proxy only the JS code necessary to identify and push the objects required for the client page load, while skipping other code. Whittling is closely related to program slicing, but with the important distinction that it is acceptable to approximate the program slice in the proxy given the client's complete execution. Experiments with top Alexa Web pages show NutShell can sustain, on average, 27\% more user requests per second than a proxy performing fully redundant execution, while preserving, and sometimes enhancing, the latency benefits.

As popular web sites turn to content delivery networks (CDNs) for full-site delivery, there is an opportunity to improve the end-user experience by optimizing the delivery of entire web pages, rather than just individual objects. In particular, this paper explores page-structure-aware strategies for placing objects in CDN cache hierarchies. The key idea is that the objects in a web page that have the largest impact on page latency should be served out of the closest or fastest caches in the hierarchy. We present schemes for identifying these objects and develop mechanisms to ensure that they are served with higher priority by the CDN, while balancing traditional CDN concerns such as optimizing the delivery of popular objects and minimizing bandwidth costs. To establish a baseline for evaluating improvements in page latencies, we collect and analyze publicly visible HTTP headers that reveal the distribution of objects among the various levels of a major CDN's cache hierarchy. Through extensive experiments on 83 real-world web pages, we show that latency reductions of over 100 ms can be obtained for 30% of the popular pages, with even larger reductions for the less popular pages. Using anonymized server logs provided by the CDN, we show the feasibility of reducing capacity and staleness misses of critical objects by 60% with minimal increase in overall miss rates, and bandwidth overheads of under 0.02%.

In this paper, we conduct a detailed study characterizing the performance of multi-tier web applications on commercial cloud platforms and evaluate the potential of techniques to improve the resilience of such applications to performance fluctuations in the cloud. In contrast to prior works that have studied the performance of individual cloud services or that of compute-intensive scientific applications (e.g., map-reduce based), our study focuses on multi-tier web applications. Our work is conducted in the context of four real-world web applications which we instrumented to collect the overall response time and the time spent in each application tier, for each transaction. Our results indicate that cloud applications undergo frequent periods of poor performance that typically (i) are short-lived lasting a few minutes; and (ii) may be attributed to a small subset of application components, though different subsets may be involved at different times. While geo-distributing applications can help mitigate performance variability, coarse-grained approaches that merely choose the best performing data-center (DC) provide only modest benefits. More significant benefits could accrue, however, if combination of cloud services located across multiple datacenters (DCs) are chosen to serve each request.

Today's web page download process is ill suited to cellular networks resulting in high page load times and radio energy usage. While there have been notable prior attempts at tackling the challenge with assistance from proxies (cloud), achieving a responsive and energy efficient browsing experience remains an elusive goal. In this paper, we make a fresh attempt at addressing the challenge by proposing PARCEL. PARCEL splits functionality between the mobile device and the proxy based on their strengths, and in a manner distinct from both traditional browsers and existing cloud-heavy approaches. We conduct extensive evaluations over an operational LTE network using a prototype implementation of PARCEL. Our results show that PARCEL reduces page load times by 49.6%, and radio energy consumption by 65% compared to traditional mobile web browsers. Further, our results show PARCEL continues to perform well under client interactions, owing to its judicious functionality split.

Modern web applications face stringent requirements along many dimensions including latency, scalability, and availability. In response, several geo-distributed cloud data stores have emerged in recent years. Customizing data stores to meet application SLAs is challenging given the scale of applications, and their diverse and dynamic workloads. In this paper, we tackle these challenges in the context of quorum-based systems (e.g. Amazon Dynamo, Cassandra), an important class of cloud storage systems. We present models that optimize percentiles of response time under normal operation and under a data-center (DC) failure. Our models consider factors like the geographic spread of users, DC locations, consistency requirements and inter-DC communication costs. We evaluate our models using real-world traces of three applications: Twitter, Wikipedia and Go Walla on a Cassandra cluster deployed in Amazon EC2. Our results confirm the importance and effectiveness of our models, and highlight the benefits of customizing replication in cloud datastores.

Modern internet applications have resulted in users sharing data with each other in an interactive fashion. These applications have very stringent service level agreements (SLAs) which place tight constraints on the performance of the underlying geo-distributed datastores. Deploying these systems in the cloud to meet such constraints is a challenging task, as application architects have to strike an optimal balance among different contrasting objectives such as maintaining consistency between multiple replicas, minimizing access latency and ensuring high availability. Achieving these objectives requires carefully configuring a number of low-level parameters of the datastores, such as the number of replicas, which DCs contain which data, and the underlying consistency protocol parameters. In this work, we adopt a systematic approach where we develop analytical models that capture the performance of a datastore based on application workload and build a system that can automatically configure the datastore for optimal performance.

Deploying interactive applications in the cloud is a challenge due to the high variability in performance of cloud services. In this paper, we present Dealer — a system that helps geo-distributed, interactive and multi-tier applications meet their stringent requirements on response time despite such variability. Our approach is motivated by the fact that, at any time, only a small number of application components of large multi-tier applications experience poor performance. Dealer continually monitors the performance of individual components and communication latencies between them to build a global view of the application. In serving any given request, Dealer seeks to minimize user response times by picking the best combination of replicas (potentially located across different data centers). While Dealer requires modifications to application code, we show the changes required are modest. Our evaluations on two multi-tier applications using real cloud deployments indicate the 90%ile of response times could be reduced by more than a factor of 6 under natural cloud dynamics. Our results indicate the cost of inter-data-center traffic with Dealer is minor, and that Dealer can in fact be used to reduce the overall operational costs of applications by up to 15% by leveraging the difference in billing plans of cloud instances.

Deploying interactive applications in the cloud is a challenge due to the high variability in performance of cloud services. In this paper, we present Dealer-- a system that helps geo-distributed, interactive and multi-tier applications meet their stringent requirements on response time despite such variability. Our approach is motivated by the fact that, at any time, only a small number of application components of large multi-tier applications experience poor performance. Dealer abstracts application structure as a component graph, with nodes being application components and edges capturing inter-component communication patterns. Dealer continually monitors the performance of individual component replicas and communication latencies between replica pairs. In serving any given user request, Dealer seeks to minimize user response times by picking the best combination of replicas (potentially located across different data-centers). While Dealer does require modifications to application code, we show through integration with two multi-tier applications that the changes required are modest. Our evaluations on two multi-tier applications using real cloud deployments indicate the 90%ile of application response times could be reduced by a factor of 3 under natural cloud dynamics compared to conventional data-center redirection techniques which are agnostic of application structure.