John Kubiatowicz

• University of California, Berkeley

Virtualization allows simultaneous execution of multi-tenant workloads on the same platform, either a server or an embedded system. Unfortunately, it is non-trivial to attribute hardware events to multiple virtual tenants, as some system's metrics relate to the whole system (e.g., RAPL energy counters). Virtualized environments have then a rather i...

Multi-tenant virtualized infrastructures allow cloud providers to minimize costs through workload consolidation. One of the largest costs is power consumption, which is challenging to understand in heterogeneous environments. We propose a power modeling methodology that tackles this complexity using a divide-and-conquer approach. Our results outper...

Autonomicity is a golden feature when dealing with a high level of complexity. This complexity can be tackled partitioning huge systems in small autonomous modules, i.e., agents. Each agent then needs to be capable of extracting knowledge from its environment and to learn from it, in order to fulfill its goals: this could not be achieved without pr...

The public cloud is moving to a Platform-as-a-Service model where services such as data management, machine learning or image classification are provided by the cloud operator while applications are written in high-level languages and leverage these services. Managed languages such as Java, Python or Scala are widely used in this setting. However,...

Many distributed workloads in today's data centers are written in managed languages such as Java or Ruby. Examples include big data frameworks such as Hadoop, data stores such as Cassandra or applications such as the SOLR search engine. These workloads typically run across many independent language runtime systems on different nodes. This setup rep...

Many distributed workloads in today's data centers are written in managed languages such as Java or Ruby. Examples include big data frameworks such as Hadoop, data stores such as Cassandra or applications such as the SOLR search engine. These workloads typically run across many independent language runtime systems on different nodes. This setup rep...

Many distributed workloads in today's data centers are written in managed languages such as Java or Ruby. Examples include big data frameworks such as Hadoop, data stores such as Cassandra or applications such as the SOLR search engine. These workloads typically run across many independent language runtime systems on different nodes. This setup rep...

Many distributed workloads in today's data centers are written in managed languages such as Java or Ruby. Examples include big data frameworks such as Hadoop, data stores such as Cassandra or applications such as the SOLR search engine. These workloads typically run across many independent language runtime systems on different nodes. This setup rep...

We address the system-level challenge of supporting a dynamically changing, complex mix of simultaneously running applications with diverse requirements including responsiveness, throughput, and performance guarantees. In our approach, called Adaptive Resource Centric Computing (ARCC), the OS distributes resources to QoS domains called cells, which...

This work compares the overhead of quantum error correction with concatenated and topological quantum error-correcting codes. To perform a numerical analysis, we use the Quantum Resource Estimator Toolbox (QuRE) that we recently developed. We use QuRE to estimate the number of qubits, quantum gates, and amount of time needed to factor a 1024-bit nu...

We introduce PHANTOM [1] a new secure processor that obfuscates its memory access trace. To an adversary who can observe the processor's output pins, all memory access traces are computationally indistinguishable (a property known as obliviousness). We achieve obliviousness through a cryptographic construct known as Oblivious RAM or ORAM. We first...

We describe QuRE, the Quantum Resource Estimator. QuRE is a layout estimation tool that estimates the cost of practical implementations of quantum circuits in a variety of competing physical quantum technologies and with a variety of strategies for fault tolerant encoding. For each specified algorithm, QuRE estimates quantities such as number of ph...

We investigate proactive dynamic load balancing on multicore systems, in which threads are continually migrated to reduce the impact of processor/thread mismatches. Our goal is to enhance the flexibility of the SPMD-style programming model and enable SPMD applications to run efficiently in multiprogrammed environments. We present Juggle, a practica...

This report describes the methodology employed by the Quantum Resource Estimator (QuRE) toolbox to quantify the resources needed to run quantum algorithms on quantum computers with realistic properties. The QuRE toolbox estimates several quantities including the number of physical qubits required to run a specified quantum algorithm, the execution...

Adaptive Resource-Centric Computing (ARCC) enables a simultaneous mix of high-throughput parallel, real-time, and interactive applications through automatic discovery of the correct mix of resource assignments necessary to achieve application requirements. This approach, embodied in the Tessellation manycore operating system, distributes resources...

This paper is about the role of the operating system (OS) within computer nodes of network audio systems. While many efforts in the network-audio community focus on low-latency network protocols, here we highlight the importance of the OS for network audio applications. We present Tessellation, an experimental OS tailored to multicore processors. W...

GPUs have become part of most commodity systems. Nonetheless, they are often underutilized when not executing graphics-intensive or special-purpose numerical computations, which are rare in consumer workloads. Emerging architectures, such as integrated CPU/GPU combinations, may create an opportunity to utilize these otherwise unused cycles for offl...

GPUs have become part of most commodity systems. Nonetheless, they are often underutilized when not executing graphics-intensive or special-purpose numerical computations, which are rare in consumer workloads. Emerging architectures, such as integrated CPU/GPU combinations, may create an opportunity to utilize these otherwise unused cycles for offl...

We investigate proactive dynamic load balancing on multicore systems, in which threads are continually migrated to reduce the impact of processor/thread mismatches to enhance the flexibility of the SPMD-style programming model, and enable SPMD applications to run efficiently in multiprogrammed environments. We present Juggle, a practical decentrali...

Tessellation is a manycore OS targeted at the resource management challenges of emerging client devices, in-cluding the need for real-time and QoS guarantees. It is predicated on two central ideas: Space-Time Partition-ing (STP) and Two-Level Scheduling. STP provides per-formance isolation and strong partitioning of resources among interacting soft...

Fault-tolerant services typically make assumptions about the type and maximum number of faults that they can tolerate while providing their correctness guarantees; when such a fault threshold is violated, correctness is lost. We revisit the notion of fault thresholds in the context of long-term archival storage. We observe that fault thresholds are...

Industry needs help from the research community to succeed in its recent dramatic shift to parallel computing. Failure could jeopardize both the IT industry and the portions of the economy that depend on rapidly improving information technology. Jeopardy for the IT industry means opportunity for the research community. If researchers meet the paral...

We optimize the area and latency of Shor's factoring while simultaneously improving fault tolerance through: (1) balancing the use of ancilla generators, (2) aggressive optimization of error correction, and (3) tuning the core adder circuits. Our custom CAD flow produces detailed layouts of the physical components and utilizes simulation to analyze...

As multiprocessors scale to unprecedented numbers of cores in order to sustain performance growth, it is vital that these gains are not nullified by high energy consumption from inter-core communication. With recent advances in 3D Integration CMOS technology, the possibility for realizing hybrid photonic-electronic networks-on-chip warrants investi...

We argue for space-time partitioning (STP) in manycore operating systems. STP divides resources such as cores, cache, and network bandwidth amongst interacting software components. Components are given unrestricted access to their resources and may schedule them in an application-specific fashion, which is critical for good parallel application per...

We analyze circuits for kernels from popular quantum computing applications, characterizing the hardware resources necessary to take ancilla preparation off the critical path. The result is a chip entirely dominated by ancilla generation circuits. To address this issue, we introduce optimized ancilla factories and analyze theirstructure and physica...

Researchers have made great strides in improving the fault tolerance of both centralized and replicated systems against arbitrary (Byzantine) faults. However, there are hard limits to how much can be done with entirely untrusted components; for example, replicated state machines cannot tolerate more than a third of their replica population being By...

Antiquity is a wide-area distributed storage system designed to provide a simple storage service for applications like file systems and back-up. The design assumes that all servers eventually fail and attempts to maintain data despite those failures. Antiquity uses a secure log to maintain data integrity, replicates each log on multiple servers for...

We design and implement a SIMD scheduler that returns the distribution of quantum operations when taking into account communication and layout constraints. We define the ion-motion path interface for trapped ion quantum computing architectures and analyze the design constraints involved in assembling the control circuitry necessary to implement thi...

We design and implement a SIMD scheduler that returns the distribution of quantum operations when taking into account communication and layout constraints. We define the ion-motion path interface for trapped ion quantum computing architectures and analyze the design constraints involved in assembling the control circuitry necessary to implement thi...

We present a computer-aided design flow for quantum circuits, complete with automatic layout and control logic extraction. To motivate automated layout for quantum circuits, we investigate grid-based layouts and show a performance variance of four times as we vary grid structure and initial qubit placement. We then propose two polynomial-time desig...

Researchers have made great strides in improving the fault tolerance of both centralized and replicated systems against arbitrary (Byzantine) faults. However, there are hard limits to how much can be done with entirely untrusted components; for example, replicated state machines cannot tolerate more than a third of their replica population being By...

Antiquity is a wide-area distributed storage system designed to provide a simple storage service for applications like file systems and back-up. The design assumes that all servers eventually fail and attempts to maintain data despite those failures. Antiquity uses a secure log to maintain data integrity, replicates each log on multiple servers for...

We show that the problem of communication in a quantum computer reduces to constructing reliable quantum channels by distributing high-fidelity EPR pairs. We develop analytical models of the latency, bandwidth, error rate and resource utilization of such channels, and show that 100s of qubits must be distributed to accommodate a single data communi...

Without Abstract

We show that the problem of communication in a quantum computer reduces to constructing reliable quantum channels by distributing high-fidelity EPR pairs. We develop analytical models of the latency, bandwidth, error rate and resource utilization of such channels, and show that 100s of qubits must be distributed to accommodate a single data communi...

This paper considers replication strategies for storage systems that aggregate the disks of many nodes spread over the Internet. Maintaining replication in such systems can be prohibitively ex- pensive, since every transient network or host failure could po- tentially lead to copying a server's worth of data over the Internet to maintain replicatio...

Many wide›area storage systems replicate data for durability. A common way of maintaining the replicas is to detect node failures and respond by creating additional copies of objects that were stored on failed nodes and hence suffered a loss of redundancy. Reactive techniques can minimize total bytes sent since they only create replicas as needed;...

Large-scale distributed systems are hard to deploy, and distributed hash tables (DHTs) are no exception. To lower the barriers facing DHT-based applications, we have created a public DHT service called OpenDHT. Designing a DHT that can be widely shared, both among mutually untrusting clients and among a variety of applications, poses two distinct c...

Large-scale distributed systems are hard to deploy, and distributed hash tables (DHTs) are no exception. To lower the barriers facing DHT-based applications, we have created a public DHT service called OpenDHT. Designing a DHT that can be widely shared, both among mutually untrusting clients and among a variety of applications, poses two distinct c...

Recent work has shown that intelligent neighbor selection during construction can significantly enhance the performance of peer-to-peer overlay networks. While its impact on performance has been recognized, few have examined the impact of neighbor selection on network resilience. In this paper, we study the impact with a generalized cost model for...

We develop a method for generating shared, secret, safe primes applicable to use in threshold RSA signature schemes such as the one developed by Shoup. We would like a scheme usable in practical settings, so our protocol is robust and ecien t in asynchronous, hostile environments. We show that the techniques used for robustness need spe- cial care...

Distributed content-addressable storage systems use self-verifying data to protect data integrity and to enable graceful scaling. One feature commonly missing from these systems, however, is the ability to identify the owner of a piece of data in a non-repudiable manner. While a solution that associates a certificate with each block of data is conc...

First-generation peer-to-peer storage systems unnecessarilycouple the unit of client data access to the unit ofinfrastructure data management. Designs that require allpeers to operate on data at a fixed granularity lead to inefficienciessuch as high query load and high per-block storageoverheads. To provide variable granularity accessand support mo...

Over the past 16 months, we have run a public DHT ser-vice called OpenDHT [14] on PlanetLab [2], allowing any networked host to perform puts and gets over an RPC in-terface. We built OpenDHT on Bamboo [13] and shame-lessly adopted other techniques from the literature, including recursive routing, proximity neighbor selection, and server selection,...

Analyzing and controlling large distributed services under a wide range of conditions is difficult. Yet these capabilities are essential to a number of important development and operational tasks such as benchmarking, testing, and system management. To facilitate these tasks, we have built the Application Control and Monitoring Environment (ACME),...

Modern networking applications replicate data and services widely, leading to a need for location-independent routing---the ability to route queries to objects using names independent of the objects' physical locations. Two important properties of such a routing infrastructure are routing locality and rapid adaptation to arriving and departing node...

over all possible choices. Figure 1 shows the resulting tree. In Tapestry [2], objects are published by placing pointers to them at each node along the path from the publisher to the root, and object location proceeds by checking for pointers along the path from the query source to the root. The stretch is determined by where the publish and search...

In this paper, we describe an ongoing effort to define common APIs for structured peer-to-peer overlays and the key abstractions that can be built on them. In doing so, we hope to facilitate independent innovation in overlay protocols, services, and applications, to allow direct experimental comparisons, and to encourage application development by...

In this paper, we broaden the applicability of peer-topeer storage infrastructures to weakly-connected clients. We present a client-side technique that exploits the commonality between consecutive versions of a file to reduce the bandwidth required to store and retrieve files in a peerto -peer storage infrastructure. We then present a novel data st...

We analyze the characteristics of overlay routing networks generated by selfish nodes playing competitive network construction games. We explore several networking scenarios--- some simplistic, others more realistic---and analyze the resulting Nash equilibrium graphs with respect to topology, performance, and resilience. We find a fundamental trade...

This paper addresses the problem of churn---the continuous process of node arrival and departure---in distributed hash tables (DHTs). We argue that DHTs should perform lookups quickly and consistently under churn rates at least as high as those observed in deployed P2P systems such as Kazaa. We then show through experiments on an emulated network t...

This paper addresses the problem of churn---the continuous process of node arrival and departure---in distributed hash tables (DHTs). We argue that DHTs should perform lookups quickly and consistently under churn rates at least as high as those observed in deployed P2P systems such as Kazaa. We then show through experiments on an emulated network t...

We analyze the characteristics of overlay routing networks generated by selfish nodes playing competitive network construction games. We explore several networking scenarios - some simplistic, others more realistic - and analyze the resulting Nash equilibrium graphs with respect to topology, performance, and resilience. We find a fundamental tradeo...

wireless networking are leading to the availability of more small, networked mobile devices, placing higher stress on existing mobility infrastructures. This problem is exacerbated by the formation of mobile crowds that generate storms of location update traffic as they cross boundaries between base stations. In this paper, we present a novel aggre...

We investigate coding techniques to reduce the energy consumed by on-chip buses in a microprocessor. We explore several simple coding schemes and simulate them using a modified SimpleScalar simulator and SPEC benchmarks. We show an average of 35% savings in transitions on internal buses. To quantify actual power savings, we design a dictionary base...

As quantum computing moves closer to reality the need for basic architectural studies becomes more pressing. Quantum wires, which transport quantum data, will be a fundamental component in all anticipated silicon quantum architectures. Since they cannot consist of a stream of electrons, as in the classical case, quantum wires must fundamentally be...

We present Tapestry, a peer-to-peer overlay routing infrastructure offering efficient, scalable, location-independent routing of messages directly to nearby copies of an object or service using only localized resources. Tapestry supports a generic decentralized object location and routing applications programming interface using a self-repairing, s...

This paper addresses the problem of churn---the continuous process of node arrival and departure---in distributed hash tables (DHTs). We demonstrate through experiment that existing DHT implementations break down at churn levels observed in deployed peer-to-peer systems, contrary to simulation-based results. We present Bamboo, a DHT that handles hi...

We broaden the applicability of peer-to-peer storage infrastructures to weakly-connected clients. We present a client-side technique that exploits the commonality between consecutive versions of a file to reduce the bandwidth required to store and retrieve files in a peer-to-peer storage infrastructure. We then present a novel data structure that a...

We devise an object location scheme that achieves a guaranteed low stretch in a wider and more realistic class of networks than previous schemes. The distinctive feature of our scheme is that it is inherently adaptive to the underlying topology. In particular, the system achieves 1+ε stretch (for arbitrarily fixed ε>0), with a neighbor list size th...

We analyze replication of resources by server nodes that act selfishly, using a game-theoretic approach. We refer to this as the selfish caching problem. In our model, nodes incur either cost for replicating resources or cost for access to a remote replica. We show the existence of pure strategy Nash equilibria and investigate the price of anarchy,...

We present the design and implementation of a light-weight decentralized ordering mechanism (SHH) for weakly consistent replication. SHH uses Merkle's cryp-tographic hashes as revision identifiers to ensure order-ing correctness. The collision resistant property of the hash function allows a revision history to be pruned ag-gressively and be secure...

In this paper, we give results relevant to sequential and distributed dynamic data structures for finding nearest neighbors in growth-restricted metrics. Our sequential data structure uses linear space, and requires O(log n) queries in expecation and O(log n) queries for lookups with high probability. This improves the results of Karger and Ruhl [4...

Recent work in P2P overlay networks allow for decentralized object location and routing (DOLR) across networks based on unique IDs. In this paper, we propose an extension to DOLR systems to publish objects using generic feature vectors instead of content-hashed GUIDs, which enables the systems to locate similar objects. We discuss the design of a d...

Advances in quantum devices have brought scalable quantum computation closer to reality. We focus on the system-level issues of how quantum devices can be brought together to form a scalable architecture. In particular, we examine promising silicon-based proposals. We discover that communication of quantum data is a critical resource in such propos...

In this paper, we describe an ongoing effort to define common APIs for structured peer-to-peer overlays and the key abstractions that can be built on them. In doing so, we hope to facilitate independent innovation in overlay protocols, services, and applications, to allow direct experimental comparisons, and to encourage application development by...

Structured peer-to-peer overlays provide a natural infrastructure for resilient routing via efficient fault detection and precomputation of backup paths. These overlays can respond to faults in a few hundred milliseconds by rapidly shifting between alternate routes. In this paper, we present two adaptive mechanisms for structured overlays and illus...

Considerable research effort has recently been devoted to the design of structured peer-to-peer overlays, a term we use to encompass Content-Addressable Networks (CANs), Distributed Hash Tables (DHTs), and Decentralized Object Location and Routing networks (DOLRs). These systems share the property that they consistently map a large set of identifie...

In this paper, we show that two peer-to-peer systems, Pastry [13] and Tapestry [17] can be made,tolerant to certain classes of failures and a limited class of attacks. These systems are said to operate properly if they can nd,the closest node matching a requested ID. The system must also be able to dynamically construct the necessary routing inform...

In this paper, we give sequential and distributed dynamic data structures for finding nearest neighbors in certain growth restricted metrics.

We present several optimizations aimed at improving the object location performance of locality-aware structured peer-to-peer overlays. We present simulation results that demonstrate the effectiveness of these optimizations in Tapestry, and discuss their usage of the overall storage re- sources of the system.

In this paper, we show that two peer-to-peer systems, Pastry [13] and Tapestry [17] can be made tolerant to certain classes of failures and a limited class of attacks. These systems are said to operate properly if they can find the closest node matching a requested ID.

In this paper, we show that two peer-to-peer systems, Pastry [13] and Tapestry [17] can be made tolerant to certain classes of failures and a limited class of attacks. These systems are said to operate properly if they can find the closest node matching a requested ID.

We present Tapestry, a peer-to-peer overlay routing infrastructure offering efficient, scalable, locationindependent routing of messages directly to nearby copies of an object or service using only localized resources. Tapestry supports a generic Decentralized Object Location and Routing (DOLR) API using a self-repairing, soft-state based routing l...

We introduce the hash history mechanism for capturing dependencies among distributed replicas. Hash histories, consisting of a directed graph of version hashes, are independent of the number of active nodes but dependent on the rate and number of modifications. We present the basic hash history scheme and discuss mechanisms for trimming the history...

As quantum computing moves closer to reality the need for basic architectural studies becomes more pressing. Quantum wires, which transport quantum data, will be a fundamental component in all anticipated silicon quantum architectures. In this paper, we introduce a quantum wire architecture based upon quantum teleportation. We compare this teleport...

As quantum computing moves closer to reality the need for basic architectural studies becomes more pressing. Quantum wires, which transport quantum data, will be a fundamental component in all anticipated silicon quantum architectures. In this paper, we introduce a quantum wire architecture based upon quantum teleportation. We compare this teleport...

Recent work in P2P overlay networks allow for decentralized object location and routing (DOLR) across networks based on unique IDs. In this paper, we propose an extension to DOLR systems to publish objects using generic feature vectors instead of content-hashed GUIDs, which enables the systems to locate similar objects. We discuss the design of a d...

We introduce the hash history mechanism for capturing dependencies among distributed replicas. Hash histories, consisting of a directed graph of version hashes, are independent of the number of active nodes but dependent on the rate and number of modifications. We present the basic hash history scheme and discuss mechanisms for trimming the history...

In this paper, we describe an ongoing effort to define common APIs for structured peer-to-peer overlays and the key abstractions that can be built on them. In doing so, we hope to facilitate independent innovation in overlay protocols, services, and applications, to allow direct experimental comparisons, and to encourage application development by...

Considerable research effort has recently been devoted to the design of structured peer-to-peer overlays, a term we use to encompass Content-Addressable Networks (CANs), Distributed Hash Tables (DHTs), and Decentralized Object Location and Routing networks (DOLRs). These systems share the property that they consistently map a large space of identif...

OceanStore is an Internet-scale, persistent data store designed for incremental scalability, secure sharing, and long-term durability. Pond is the OceanStore prototype; it contains many of the features of a complete system including location-independent routing, Byzantine update commitment, push-based update of cached copies through an overlay mult...

OceanStore is an Internet-scale, persistent data store designed for incremental scalability, secure sharing, and long-term durability. Pond is the OceanStore prototype; it contains many of the features of a complete system including location-independent routing, Byzantine update commitment, push-based update of cached copies through an overlay mult...

This article focuses on chaos in a Peer-to-Peer (P2P) architecture. The P2P revolution promises freedom from boundaries, censorship and centralized control. P2P proponents claim that the vast untapped resource of personal computers owned by ordinary people can be combined together to build something greater and more reliable than the sum of its par...

As quantum computing moves closer to reality the need for basic architectural studies becomes more pressing. Quantum wires, which transport quantum data, is a fundamental component in all anticipated silicon quantum architectures. We introduce a quantum wire architecture based upon quantum teleportation. We compare this teleportation channel with t...

Structured peer-to-peer overlays provide a natural infrastructure for resilient routing via efficient fault detection and precomputation of backup paths. These overlays can respond to faults in a few hundred milliseconds by rapidly shifting between alternate routes. In this paper, we present two adaptive mechanisms for structured overlays and illus...

Quantum computation has become an intriguing technology with which to attack difficult problems and to enhance system security. Quantum algorithms, however, have been analyzed under idealized assumptions without important physical constraints in mind. In this paper, we analyze two key constraints: the short spatial distance of quantum interactions...

OceanStore is an Internet-scale, persistent data store designed for incremental scalability, secure sharing, and long-term durability. Pond is the OceanStore prototype; it contains many of the features of a complete system including location-independent routing, Byzantine update commitment, push-based update of cached copies through an overlay mult...

In this paper, we present Mobile Tapestry, an extension to the Tapestry overlay network protocol, that enables scalable, fault-tolerant, and timely delivery of network messages, including multimedia streams, to and from rapidly moving nodes. Mobile Tapestry efficiently supports individual mobile nodes and, by using an approach we call hierarchical...

As new and interesting peer-to-peer applications combine with advancements in networking technology, they are reaching millions of users across the globe. Numerous studies have shown, however, that loss of connectivity is common on the wide-area network, due to hardware and software failures, and network misconfigurations. Despite the natural redun...

As new and interesting peer-to-peer applications combine with advancements in networking technology, they are reaching millions of users across the globe. Numerous studies have shown, however, that loss of connectivity is common on the wide-area network, due to hardware and software failures, and network misconfigurations. Despite the natural redun...