Towards characterizing cloud backend workloads

Abstract

The advent of cloud computing promises highly available, efficient, and flexible computing services for applications such as web search, email, voice over IP, and web search alerts. Our experience at Google is that realizing the promises of cloud computing requires an extremely scalable backend consisting of many large compute clusters that are shared by application tasks with diverse service level requirements for throughput, latency, and jitter. These considerations impact (a) capacity planning to determine which machine resources must grow and by how much and (b) task scheduling to achieve high machine utilization and to meet service level objectives.Both capacity planning and task scheduling require a good understanding of task resource consumption (e.g., CPU and memory usage). This in turn demands simple and accurate approaches to workload classification-determining how to form groups of tasks (workloads) with similar resource demands. One approach to workload classification is to make each task its own workload. However, this approach scales poorly since tens of thousands of tasks execute daily on Google compute clusters. Another approach to workload classification is to view all tasks as belonging to a single workload. Unfortunately, applying such a coarse-grain workload classification to the diversity of tasks running on Google compute clusters results in large variances in predicted resource consumptions.This paper describes an approach to workload classification and its application to the Google Cloud Backend, arguably the largest cloud backend on the planet. Our methodology for workload classification consists of: (1) identifying the workload dimensions; (2) constructing task classes using an off-the-shelf algorithm such as k-means; (3) determining the break points for qualitative coordinates within the workload dimensions; and (4) merging adjacent task classes to reduce the number of workloads. We use the foregoing, especially the notion of qualitative coordinates, to glean several insights about the Google Cloud Backend: (a) the duration of task executions is bimodal in that tasks either have a short duration or a long duration; (b) most tasks have short durations; and (c) most resources are consumed by a few tasks with long duration that have large demands for CPU and memory.

Reliability and high availability in cloud computing environments: a reference roadmap

“…According to our studies, cloud tasks can be classified into two long-running and short-running tasks in terms of tasks' duration. Therefore based on Table 3, it can be said that task durations are bimodal, either somewhat less than 30 min or larger than 18 h [60]. Such behavior results from the characteristics and types of application tasks running on cloud infrastructures.…”

“…Providing this guarantee needs a scalable infrastructure including many computing clusters which are shared by various tasks with different requirements and quality of service in terms of availability, reliability, latency and throughput [60]. Therefore, to provide required service level (e.g.…”

“…A task classification has been done in [60] based on the tasks' resource consumption running in the Google Cloud Backend. The results of this study are presented in Table 3.…”

“…The first are interactive or user-facing tasks which run continuously so as to respond quickly to a user request. The second type of long-running tasks is compute-intensive, such as processing web logs [60]. Interactive tasks consume a large amount of CPU and memory during periods of a high user request rate.…”

Workload Modeling for Computer Systems Performance Evaluation

“…Initial analyses of cloud workloads have been published by Mishra et al [490], Reiss et al [562], and Di et al [177]. Di et al characterized the workload on a Google data center [177], and compared it with the workload on computational grids.…”