Using machine learning (ML) to tackle computer systems tasks is gaining popularity. One of the shortcomings of such ML-based approaches is the inability of models to generalize to out-ofdistribution data i.e., data whose distribution is different than the training dataset. We showcase that this issue exists in cloud environments by analyzing various ML models used to improve resource balance in Google’s fleet. We discuss the trade-offs associated with different techniques used to detect out-of-distribution data. Finally, we propose and demonstrate the efficacy of using Bayesian models to detect the model’s confidence in its output when used to improve cloud server resource balance.

As large-scale data processing workloads continue to grow, their carbon footprint raises concerns. Prior research on carbon-aware schedulers has focused on shifting computation to align with the availability of low-carbon energy, but these approaches assume that each task can be executed independently. In contrast, data processing jobs have precedence constraints that complicate decisions, since delaying an upstream "bottleneck" task to a low-carbon period also blocks downstream tasks, impacting makespan. In this paper, we show that carbon-aware scheduling for data processing benefits from knowledge of both time-varying carbon and precedence constraints. Our main contribution is PCAPS, a carbon-aware scheduler that builds on state-of-the-art scoring or probability-based techniques - in doing so, it explicitly relates the structural importance of each task against the time-varying characteristics of carbon intensity. To illustrate gains due to fine-grained task-level scheduling, we also study CAP, a wrapper for any carbon-agnostic scheduler that generalizes the provisioning ideas of PCAPS. Both techniques allow a user-configurable priority between carbon and makespan, and we give basic analytic results to relate the trade-off between these objectives. Our prototype on a 100-node Kubernetes cluster shows that a moderate configuration of PCAPS reduces carbon footprint by up to 32.9% without significantly impacting total efficiency.

Memory allocation constitutes a substantial component of warehouse-scale computation. Optimizing the memory allocator not only reduces the datacenter tax, but also improves application performance, leading to significant cost savings.

The rapid increase in computing demand and corresponding energy consumption have focused attention on computing's impact on the climate and sustainability. Prior work proposes metrics that quantify computing's carbon footprint across several lifecycle phases, including its supply chain, operation, and end-of-life. Industry uses these metrics to optimize the carbon footprint of manufacturing hardware and running computing applications. Unfortunately, prior work on optimizing datacenters' carbon footprint often succumbs to the sunk cost fallacy by considering embodied carbon emissions (a sunk cost) when making operational decisions (i.e., job scheduling and placement), which leads to operational decisions that do not always reduce the total carbon footprint.

Modern cloud applications are prone to high tail latencies since their requests typically follow highly-dispersive distributions. Prior work has proposed both OS- and systemlevel solutions to reduce tail latencies for microsecond-scale workloads through better scheduling. Unfortunately, existing approaches like customized dataplane OSes, require significant OS changes, experience scalability limitations, or do not reach the full performance capabilities hardware offers. We propose LibPreemptible, a preemptive user-level threading library that is flexible, lightweight, and scalable. LibPreemptible is based on three key techniques: 1) a fast and lightweight hardware mechanism for delivery of timed interrupts, 2) a general-purpose user-level scheduling interface, and 3) an API for users to express adaptive scheduling policies tailored to the needs of their applications. Compared to the prior state-of-the-art scheduling system Shinjuku, our system achieves significant tail latency and throughput improvements for various workloads without the need to modify the kernel. We also demonstrate the flexibility of LibPreemptible across scheduling policies for real applications experiencing varying load levels and characteristics.

Resource management for cloud-native microservices has attracted a lot of recent attention. Previous work has shown that machine learning (ML)-driven approaches out-perform traditional techniques, such as autoscaling, in terms of both SLA maintenance and resource efficiency. However, ML-driven approaches also face challenges including lengthy data collection processes and limited scalability. We present Ursa, a lightweight resource management system for cloud-native microservices that addresses these challenges. Ursa uses an analytical model that decomposes the end-to-end SLA into per-service SLA, and maps per-service SLA to individual resource allocations per microservice tier. To speed up the exploration process and avoid prolonged SLA violations, Ursa explores each microservice individually, and swiftly stops exploration if latency exceeds its SLA. We evaluate Ursa on a set of representative and end-to-end microservice topologies, including a social network, media service and video processing pipeline, each consisting of multiple classes and priorities of requests with different SLAs, and compare it against two representative ML-driven systems, Sinan and Firm. Compared to these ML-driven approaches, Ursa provides significant advantages: It shortens the data collection process by more than 128 ×, and its control plane is 43 × faster than ML-driven approaches. At the same time, Ursa does not sacrifice resource efficiency or SLAs. During online deployment, Ursa reduces the SLA violation rate by 9.0% up to 49.9%, and reduces CPU allocation by up to 86.2% compared to ML-driven approaches

The lack of representative, publicly-available cloud services has been a recurring problem in the architecture and systems communities. While open-source benchmarks exist, they do not capture the full complexity of cloud services. Application cloning is a promising way to address this, however, prior work is limited to CPU-/cache-centric, single-node services, operating at user level.

Multi-stage serverless applications, i.e., workflows with many computation and I/O stages, are becoming increasingly representative of FaaS platforms. Despite their advantages in terms of fine-grained scalability and modular development, these applications are subject to suboptimal performance, resource inefficiency, and high costs to a larger degree than previous simple serverless functions.

Building large AI fleets to support the rapidly growing DL workloads is an active research topic for modern cloud providers. Generating accurate benchmarks plays an essential role in designing the fast-paced software and hardware solutions in this space. Two fundamental challenges to make this scalable are (i) workload representativeness and (ii) the ability to quickly incorporate changes to the fleet into the benchmarks.

In cellular networks, there is a growing adoption of virtualized radio access networks (vRANs), where operators are replacing the traditional specialized hardware for RAN processing with software running on commodity servers. Today's vRAN deployments lack resilience, since there is no support for vRAN failover or upgrades without long service interruptions. Enabling these features for vRANs is challenging because of their strict real-time latency requirements and black-box nature. Slingshot is a new system that transparently provides resilience for the vRAN's most performance-critical layer: the physical layer (PHY). We design new techniques for realtime workload migration with fast RAN protocol middle-boxes, and realtime RAN failure detection. A key insight in our design is to view the transient disruptions from resilience events to RAN computation state and I/O similarly to regular wireless signal impairments, and leverage the inherent resilience of cellular networks to these events. Experiments with a state-of-the-art 5G vRAN testbed show that Slingshot handles PHY failover with no disruption to video conferencing, and under 110 ms disruption to a TCP connection, and it also enables zero-downtime upgrades.

Many cloud services have Quality-of-Service (QoS) requirements; most requests have to to complete within a given latency constraint. Recently, researchers have begun to investigate whether it is possible to meet QoS while attempting to save power on a per-request basis. Existing work shows that one can indeed hand-tune a request latency predictor offline for a particular cloud application, and consult it at runtime to modulate CPU voltage and frequency, resulting in substantial power savings.In this paper, we propose ReTail, an automated and general solution for request-level power management of latency-critical services with QoS constraints. We present a systematic process to select the features of any given application that best correlate with its request latency. ReTail uses these features to predict latency, and adjust CPU’s power consumption. ReTail’s predictor is trained fully at runtime. We show that unlike previous findings, simple techniques perform better than complex machine learning models, when using the right input features. For a web search engine, ReTail outperforms prior mechanisms based on complex hand-tuned predictors for that application domain. Furthermore, ReTail’s systematic approach also yields superior power savings across a diverse set of cloud applications.

The slowdown of Moore’s Law, combined with advances in 3D stacking of logic and memory, have pushed architects to revisit the concept of processing-in-memory (PIM) to overcome the memory wall bottleneck. This PIM renaissance finds itself in a very different computing landscape from the one twenty years ago, as more and more computation shifts to the cloud. Most PIM architecture papers still focus on best-effort applications, while PIM’s impact on latency-critical cloud applications is not well understood.This paper explores how datacenters can exploit PIM architectures in the context of latency-critical applications. We adopt a general-purpose cloud server with HBM-based, 3D-stacked logic+memory modules, and study the impact of PIM on six diverse interactive cloud applications. We reveal the previously neglected opportunity that PIM presents to these services, and show the importance of properly managing PIM-related resources to meet the QoS targets of interactive services and maximize resource efficiency. Then, we present PIMCloud, a QoS-aware resource manager designed for cloud systems with PIM allowing colocation of multiple latency-critical and best-effort applications. We show that PIMCloud efficiently manages PIM resources: it (1) improves effective machine utilization by up to 70% and 85% (average 24% and 33%) under 2-app and 3-app mixes, compared to the best state-of-the-art manager; (2) helps latency-critical applications meet QoS; and (3) adapts to varying load patterns.

Swarms of autonomous devices are increasing in ubiquity and size, making the need for rethinking their hardware-software system stack critical.

Cloud applications are increasingly shifting from large monolithic services to complex graphs of loosely-coupled microservices. Despite the advantages of modularity and elasticity microservices offer, they also complicate cluster management and performance debugging, as dependencies between tiers introduce backpressure and cascading QoS violations. Prior work on performance debugging for cloud services either relies on empirical techniques, or uses supervised learning to diagnose the root causes of performance issues, which requires significant application instrumentation, and is difficult to deploy in practice.

The ongoing shift of cloud services from monolithic designs to mi- croservices creates high demand for efficient and high performance datacenter networking stacks, optimized for fine-grained work- loads. Commodity networking systems based on software stacks and peripheral NICs introduce high overheads when it comes to delivering small messages. We present Dagger, a hardware acceleration fabric for cloud RPCs based on FPGAs, where the accelerator is closely-coupled with the host processor over a configurable memory interconnect. The three key design principle of Dagger are: (1) offloading the entire RPC stack to an FPGA-based NIC, (2) leveraging memory interconnects instead of PCIe buses as the interface with the host CPU, and (3) making the acceleration fabric reconfigurable, so it can accommodate the diverse needs of microservices. We show that the combination of these principles significantly improves the efficiency and performance of cloud RPC systems while preserving their generality. Dagger achieves 1.3 − 3.8× higher per-core RPC throughput compared to both highly-optimized software stacks, and systems using specialized RDMA adapters. It also scales up to 84 Mrps with 8 threads on 4 CPU cores, while maintaining state-of- the-art µs-scale tail latency. We also demonstrate that large third- party applications, like memcached and MICA KVS, can be easily ported on Dagger with minimal changes to their codebase, bringing their median and tail KVS access latency down to 2.8 − 3.5 us and 5.4 − 7.8 us, respectively. Finally, we show that Dagger is beneficial for multi-tier end-to-end microservices with different threading models by evaluating it using an 8-tier application implementing a flight check-in service.

Cloud applications are increasingly shifting from large monolithic services, to large numbers of loosely-coupled, specialized microservices. Despite their advantages in terms of facilitating development, deployment, modularity, and isolation, microservices complicate resource management, as dependencies between them introduce backpressure effects and cascading QoS violations.

Cloud services are increasingly adopting new programming models, such as microservices and serverless compute. While these frameworks offer several advantages, such as better modularity, ease of maintenance and deployment, they also introduce new hardware and software challenges.

Serverless computing is becoming increasingly popular due to its ease of programming, fast elasticity, and fine-grained billing. However, the serverless provider still needs to provision, manage, and pay the IaaS provider for the virtual machines (VMs) hosting its platform. This ties the cost of the serverless platform to the cost of the underlying VMs. One way to significantly reduce cost is to use spare resources, which cloud providers rent at a massive discount. Harvest VMs offer such cheap resources: they grow and shrink to harvest all the unallocated CPU cores in their host servers, but may be evicted to make room for more expensive VMs. Thus, using Harvest VMs to run the serverless platform comes with two main challenges that must be carefully managed: VM evictions and dynamically varying resources in each VM.

Infrastructure-as-a-Service cloud providers sell virtual machines that are only specified in terms of number of CPU cores, amount of memory, and I/O throughput. Performance-critical aspects such as cache sizes and memory latency are missing or reported in ways that make them hard to compare across cloud providers. It is difficult for users to adapt their application’s behavior to the available resources. In this work, we aim to increase the visibility that cloud users have into shared resources on public clouds. Specifically, we present CacheInspector , a lightweight runtime that determines the performance and allocated capacity of shared caches on multi-tenant public clouds. We validate CacheInspector ’s accuracy in a controlled environment, and use it to study the characteristics and variability of cache resources in the cloud, across time, instances, availability regions, and cloud providers. We show that CacheInspector ’s output allows cloud users to tailor their application’s behavior, including their output quality, to avoid suboptimal performance when resources are scarce.

Multi-tenancy for latency-critical applications leads to resource interference and unpredictable performance. Core reconfiguration opens up more opportunities for application colocation, as it allows the hardware to adjust to the dynamic performance and power needs of a specific mix of co-scheduled services. However, reconfigurability also introduces challenges, as even for a small number of reconfigurable cores, exploring the design space becomes more time- and resource-demanding.We present CuttleSys, a runtime for reconfigurable multicores that leverages scalable and lightweight data mining to quickly identify suitable core and cache configurations for a set of co-scheduled applications. The runtime combines collaborative filtering to infer the behavior of each job on every core and cache configuration, with Dynamically Dimensioned Search to efficiently explore the configuration space. We evaluate CuttleSys on multicores with tens of reconfigurable cores and show up to 2.46× and 1.55× performance improvements compared to core-level gating and oracle-like asymmetric multicores respectively, under stringent power constraints.

Multi-tenancy in modern datacenters is currently limited to a single latency-critical, interactive service, running alongside one or more low-priority, best-effort jobs. This limits the efficiency gains from multi-tenancy, especially as an increasing number of cloud applications are shifting from batch jobs to services with strict latency requirements. We present PARTIES, a QoS-aware resource manager that enables an arbitrary number of interactive, latency-critical services to share a physical node without QoS violations. PARTIES leverages a set of hardware and software resource partitioning mechanisms to adjust allocations dynamically at runtime, in a way that meets the QoS requirements of each co-scheduled workload, and maximizes throughput for the machine. We evaluate PARTIES on state-of-the-art server platforms across a set of diverse interactive services. Our results show that PARTIES improves throughput under QoS by 61% on average, compared to existing resource managers, and that the rate of improvement increases with the number of co-scheduled applications per physical host.

Cloud services have recently started undergoing a major shift from monolithic applications, to graphs of hundreds or thousands of loosely-coupled microservices. Microservices fundamentally change a lot of assumptions current cloud systems are designed with, and present both opportunities and challenges when optimizing for quality of service (QoS) and cloud utilization.

Performance unpredictability is a major roadblock towards cloud adoption, and has performance, cost, and revenue ramifications. Predictable performance is even more critical as cloud services transition from monolithic designs to microservices. Detecting QoS violations after they occur in systems with microservices results in long recovery times, as hotspots propagate and amplify across dependent services. We present Seer, an online cloud performance debugging system that leverages deep learning and the massive amount of tracing data cloud systems collect to learn spatial and temporal patterns that translate to QoS violations. Seer combines lightweight distributed RPC-level tracing, with detailed low-level hardware monitoring to signal an upcoming QoS violation, and diagnose the source of unpredictable performance. Once an imminent QoS violation is detected, Seer notifies the cluster manager to take action to avoid performance degradation altogether. We evaluate Seer both in local clusters, and in large-scale deployments of end-to-end applications built with microservices with hundreds of users. We show that Seer correctly anticipates QoS violations 91% of the time, and avoids the QoS violation to begin with in 84% of cases. Finally, we show that Seer can identify application-level design bugs, and provide insights on how to better architect microservices to achieve predictable performance.

"Cloud-native" container platforms, such as Kubernetes, have become an integral part of production cloud environments. One of the principles in designing cloud-native applications is called Single Concern Principle, which suggests that each container should handle a single responsibility well. In this paper, we propose X-Containers as a new security paradigm for isolating single-concerned cloud-native containers. Each container is run with a Library OS (LibOS) that supports multi-processing for concurrency and compatibility. A minimal exokernel ensures strong isolation with small kernel attack surface. We show an implementation of the X-Containers architecture that leverages Xen paravirtualization (PV) to turn Linux kernel into a LibOS. Doing so results in a highly efficient LibOS platform that does not require hardware-assisted virtualization, improves inter-container isolation, and supports binary compatibility and multi-processing. By eliminating some security barriers such as seccomp and Meltdown patch, X-Containers have up to 27X higher raw system call throughput compared to Docker containers, while also achieving competitive or superior performance on various benchmarks compared to recent container platforms such as Google's gVisor and Intel's Clear Containers.

Cloud multi-tenancy is typically constrained to a single interactive service colocated with one or more batch, low-priority services, whose performance can be sacrificed when deemed necessary. Approximate computing applications offer the opportunity to enable tighter colocation among multiple applications whose performance is important. We present Pliant, a lightweight cloud runtime that leverages the ability of approximate computing applications to tolerate some loss in their output quality to boost the utilization of shared servers. During periods of high resource contention, Pliant employs incremental and interference-aware approximation to reduce contention in shared resources, and prevent QoS violations for co-scheduled interactive, latency-critical services. We evaluate Pliant across different interactive and approximate computing applications, and show that it preserves QoS for all co-scheduled workloads, while incurring a 2.1% loss in output quality, on average.

Current cloud services are moving away from monolithic designs and towards graphs of many loosely-coupled, single-concerned microservices. Microservices have several advantages, including speeding up development and deployment, allowing specialization of the software infrastructure, and helping with debugging and error isolation. At the same time they introduce several hardware and software challenges. Given that most of the performance and efficiency implications of microservices happen at scales larger than what is available outside production deployments, studying such effects requires designing the right simulation infrastructures. We present j);qSim, a scalable and validated queueing network simulator designed specifically for interactive microser-vices. μqSim provides detailed intra- and inter-microservice models that allow it to faithfully reproduce the behavior of complex, many-tier applications. μqSim is also modular, allowing reuse of individual models across microservices and end-to-end applications. We have validated μqSim both against simple and more complex microservices graphs, and have shown that it accurately captures performance in terms of throughput and tail latency. Finally, we use μqSim to model the tail at scale effects of request fanout, and the performance impact of power management in latency -sensitive microservices.

Heterogeneity has grown in popularity both at the core and server level as a way to improve both performance and energy efficiency. However, despite these benefits, scheduling applications in heterogeneous machines remains challenging. Additionally, when these heterogeneous resources accommodate multiple applications to increase utilization, resources are prone to contention, destructive interference, and unpredictable performance. Existing solutions examine heterogeneity either across or within a server, leading to missed performance and efficiency opportunities.

Cloud providers routinely schedule multiple applications per physical host to increase efficiency. The resulting interference on shared resources often leads to performance degradation and, more importantly, security vulnerabilities. Interference can leak important information ranging from a service's placement to confidential data, like private keys. We present Bolt, a practical system that accurately detects the type and characteristics of applications sharing a cloud platform based on the interference an adversary sees on shared resources. Bolt leverages online data mining techniques that only require 2-5 seconds for detection. In a multi-user study on EC2, Bolt correctly identifies the characteristics of 385 out of 436 diverse workloads. Extracting this information enables a wide spectrum of previously-impractical cloud attacks, including denial of service attacks (DoS) that increase tail latency by 140x, as well as resource freeing (RFA) and co-residency attacks. Finally, we show that while advanced isolation mechanisms, such as cache partitioning lower detection accuracy, they are insufficient to eliminate these vulnerabilities altogether. To do so, one must either disallow core sharing, or only allow it between threads of the same application, leading to significant inefficiencies and performance penalties.

Cloud computing promises flexibility and high performance for users and cost efficiency for operators. To achieve this, cloud providers offer instances of different sizes, both as long-term reservations and short-term, on-demand allocations. Unfortunately, determining the best provisioning strategy is a complex, multi-dimensional problem that depends on the load fluctuation and duration of incoming jobs, and the performance unpredictability and cost of resources. We first compare the two main provisioning strategies (reserved and on-demand resources) on Google Compute Engine (GCE) using three representative workload scenarios with batch and latency-critical applications. We show that either approach is suboptimal for performance or cost. We then present HCloud, a hybrid provisioning system that uses both reserved and on-demand resources. HCloud determines which jobs should be mapped to reserved versus on-demand resources based on overall load, and resource unpredictability. It also determines the optimal instance size an application needs to satisfy its Quality of Service (QoS) constraints. We demonstrate that hybrid configurations improve performance by 2.1x compared to fully on-demand provisioning, and reduce cost by 46% compared to fully reserved systems. We also show that hybrid strategies are robust to variation in system and job parameters, such as cost and system load.

FPGAs are a popular target for application-specific accelerators because they lead to a good balance between flexibility and energy efficiency. However, FPGA lookup tables introduce significant area and power overheads, making it difficult to use FPGA devices in environments with tight cost and power constraints. This is the case for datacenter servers, where a modestly-sized FPGA cannot accommodate the large number of diverse accelerators that datacenter applications need. This paper introduces DRAF, an architecture for bit-level reconfigurable logic that uses DRAM subarrays to implement dense lookup tables. DRAF overlaps DRAM operations like bitline precharge and charge restoration with routing within the reconfigurable routing fabric to minimize the impact of DRAM latency. It also supports multiple configuration contexts that can be used to quickly switch between different accelerators with minimal latency. Overall, DRAF trades off some of the performance of FPGAs for significant gains in area and power. DRAF improves area density by 10x over FPGAs and power consumption by more than 3x, enabling DRAF to satisfy demanding applications within strict power and cost constraints. While accelerators mapped to DRAF are 2-3x slower than those in FPGAs, they still deliver a 13x speedup and an 11x reduction in power consumption over a Xeon core for a wide range of datacenter tasks, including analytics and interactive services like speech recognition.

Acceleration in the form of customized datapaths offer large performance and energy improvements over general purpose processors. Reconfigurable fabrics such as FPGAs are gaining popularity for use in implementing application-specific accelerators, thereby increasing the importance of having good high-level FPGA design tools. However, current tools for targeting FPGAs offer inadequate support for high-level programming, resource estimation, and rapid and automatic design space exploration. We describe a design framework that addresses these challenges. We introduce a new representation of hardware using parameterized templates that captures locality and parallelism information at multiple levels of nesting. This representation is designed to be automatically generated from high-level languages based on parallel patterns. We describe a hybrid area estimation technique which uses template-level models and design-level artificial neural networks to account for effects from hardware place-and-route tools, including routing overheads, register and block RAM duplication, and LUT packing. Our runtime estimation accounts for off-chip memory accesses. We use our estimation capabilities to rapidly explore a large space of designs across tile sizes, parallelization factors, and optional coarse-grained pipelining, all at multiple loop levels. We show that estimates average 4.8% error for logic resources, 6.1% error for runtimes, and are 279 to 6533 times faster than a commercial high-level synthesis tool. We compare the best-performing designs to optimized CPU code running on a server-grade 6 core processor and show speedups of up to 16.7×.

Scheduling diverse applications in large, shared clusters is particularly challenging. Recent research on cluster scheduling focuses either on scheduling speed, using sampling to quickly assign resources to tasks, or on scheduling quality, using centralized algorithms that search for the resources that improve both task performance and cluster utilization.

Cloud computing promises flexibility and high performance for users and high cost-efficiency for operators. Nevertheless, most cloud facilities operate at very low utilization, hurting both cost effectiveness and future scalability.

Large-scale datacenters (DCs) host tens of thousands of diverse applications each day. However, interference between colocated workloads and the difficulty to match applications to one of the many hardware platforms available can degrade performance, violating the quality of service (QoS) guarantees that many cloud workloads require. While previous work has identified the impact of heterogeneity and interference, existing solutions are computationally intensive, cannot be applied online and do not scale beyond few applications.

Large-scale datacenters (DCs) host tens of thousands of diverse applications each day. However, interference between colocated workloads and the difficulty of matching applications to one of the many hardware platforms available can degrade performance, violating the quality of service (QoS) guarantees that many cloud workloads require. While previous work has identified the impact of heterogeneity and interference, existing solutions are computationally intensive, cannot be applied online, and do not scale beyond a few applications. We present Paragon, an online and scalable DC scheduler that is heterogeneity- and interference-aware. Paragon is derived from robust analytical methods, and instead of profiling each application in detail, it leverages information the system already has about applications it has previously seen. It uses collaborative filtering techniques to quickly and accurately classify an unknown incoming workload with respect to heterogeneity and interference in multiple shared resources. It does so by identifying similarities to previously scheduled applications. The classification allows Paragon to greedily schedule applications in a manner that minimizes interference and maximizes server utilization. After the initial application placement, Paragon monitors application behavior and adjusts the scheduling decisions at runtime to avoid performance degradations. Additionally, we design ARQ, a multiclass admission control protocol that constrains application waiting time. ARQ queues applications in separate classes based on the type of resources they need and avoids long queueing delays for easy-to-satisfy workloads in highly-loaded scenarios. Paragon scales to tens of thousands of servers and applications with marginal scheduling overheads in terms of time or state. We evaluate Paragon with a wide range of workload scenarios, on both small and large-scale systems, including 1,000 servers on EC2. For a 2,500-workload scenario, Paragon enforces performance guarantees for 91% of applications, while significantly improving utilization. In comparison, heterogeneity-oblivious, interference-oblivious, and least-loaded schedulers only provide similar guarantees for 14%, 11%, and 3% of workloads. The differences are more striking in oversubscribed scenarios where resource efficiency is more critical.

With the advent of social networking and cloud data-stores, user data is increasingly being stored in large capacity and high performance storage systems, which account for a significant portion of the total cost of ownership of a datacenter (DC) [3]. One of the main challenges when trying to evaluate storage system options is the difficulty in replaying the entire application in all possible system configurations. Furthermore, code and datasets of DC applications are rarely available to storage system designers. This makes the development of a representative model that captures key aspects of the workload's storage profile, even more appealing. Once such a model is available, the next step is to create a tool that convincingly reproduces the application's storage behavior via a synthetic I/O access pattern.