High-throughput computing (HTC) workloads seek to complete as many jobs as possible over a long period of time. Such workloads require efficient execution of many parallel jobs and can occupy a large number of resources for a longtime. As a result, full utilization is the normal state of an HTC facility. The widespread use of container orchestrators eases the deployment of HTC frameworks across different platforms,which also provides an opportunity to scale up HTC workloads with almost infinite resources on the public cloud. However, the autoscaling mechanisms of container orchestrators are primarily designed to support latency-sensitive microservices, and result in unexpected behavior when presented with HTC workloads. In this paper, we design a feedback autoscaler, High Throughput Autoscaler (HTA), that leverages the unique characteristics ofthe HTC workload to autoscales the resource pools used by HTC workloads on container orchestrators. HTA takes into account a reference input, the real-time status of the jobs’ queue, as well as two feedback inputs, resource consumption of jobs, and the resource initialization time of the container orchestrator. We implement HTA using the Makeflow workload manager, WorkQueue job scheduler, and the Kubernetes cluster manager. We evaluate its performance on both CPU-bound and IO-bound workloads. The evaluationmore »
Mind the Gap: Broken Promises of CPU Reservations in Containerized Multi-tenant Clouds
Containerization is becoming increasingly popular, but unfortunately, containers often fail to deliver the anticipated performance with the allocated resources. In this paper, we first demonstrate the performance variance and degradation are significant (by up to 5x) in a multi-tenant environment where containers are co-located. We then investigate the root cause of such performance degradation. Contrary to the common belief that such degradation is caused by resource contention and interference, we find that there is a gap between the amount of CPU a container reserves and actually gets. The root cause lies in the design choices of today's Linux scheduling mechanism, which we call Forced Runqueue Sharing and Phantom CPU Time. In fact, there are fundamental conflicts between the need to reserve CPU resources and Completely Fair Scheduler's work-conserving nature, and this contradiction prevents a container from fully utilizing its requested CPU resources. As a proof-of-concept, we implement a new resource configuration mechanism atop the widely used Kubernetes and Linux to demonstrate its potential benefits and shed light on future scheduler redesign. Our proof-of-concept, compared to the existing scheduler, improves the performance of both batch and interactive containerized apps by up to 5.6x and 13.7x.
- Award ID(s):
- 2045680
- Publication Date:
- NSF-PAR ID:
- 10312140
- Journal Name:
- ACM Symposium on Cloud Computing
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Data-intensive applications often suffer from significant memory pressure, resulting in excessive garbage collection (GC) and out-of-memory (OOM) errors, harming system performance and reliability. In this paper, we demonstrate how lightweight virtualization via OS containers opens up opportunities to address memory pressure and realize memory elasticity: 1) tasks running in a container can be set to a large heap size to avoid OutOfMemory (OOM) errors, and 2) tasks that are under memory pressure and incur significant swapping activities can be temporarily "suspended" by depriving resources from the hosting containers, and be "resumed" when resources are available. We propose and develop Pufferfish, an elastic memory manager, that leverages containers to flexibly allocate memory for tasks. Memory elasticity achieved by Pufferfish can be exploited by a cluster scheduler to improve cluster utilization and task parallelism. We implement Pufferfish on the cluster scheduler Apache Yarn. Experiments with Spark and MapReduce on real-world traces show Pufferfish is able to avoid OOM errors, improve cluster memory utilization by 2.7x and the median job runtime by 5.5x compared to a memory over-provisioning solution.
-
Data-intensive applications often suffer from significant memory pressure, resulting in excessive garbage collection (GC) and out-of-memory (OOM) errors, harming system performance and reliability. In this paper, we demonstrate how lightweight virtualization via OS containers opens up opportunities to address memory pressure and realize memory elasticity: 1) tasks running in a container can be set to a large heap size to avoid OutOfMemory (OOM) errors, and 2) tasks that are under memory pressure and incur significant swapping activities can be temporarily "suspended" by depriving resources from the hosting containers, and be "resumed" when resources are available. We propose and develop Pufferfish, an elastic memory manager, that leverages containers to flexibly allocate memory for tasks. Memory elasticity achieved by Pufferfish can be exploited by a cluster scheduler to improve cluster utilization and task parallelism. We implement Pufferfish on the cluster scheduler Apache Yarn. Experiments with Spark and MapReduce on real-world traces show Pufferfish is able to avoid OOM errors, improve cluster memory utilization by 2.7x and the median job runtime by 5.5x compared to a memory over-provisioning solution.
-
Container networking, which provides connectivity among containers on multiple hosts, is crucial to building and scaling container-based microservices. While overlay networks are widely adopted in production systems, they cause significant performance degradation in both throughput and latency compared to physical networks. This paper seeks to understand the bottlenecks of in-kernel networking when running container overlay networks. Through profiling and code analysis, we find that a prolonged data path, due to packet transformation in overlay networks, is the culprit of performance loss. Furthermore, existing scaling techniques in the Linux network stack are ineffective for parallelizing the prolonged data path of a single network flow. We propose FALCON, a fast and balanced container networking approach to scale the packet processing pipeline in overlay networks. FALCON pipelines software interrupts associated with different network devices of a single flow on multiple cores, thereby preventing execution serialization of excessive software interrupts from overloading a single core. FALCON further supports multiple network flows by effectively multiplexing and balancing software interrupts of different flows among available cores. We have developed a prototype of FALCON in Linux. Our evaluation with both micro-benchmarks and real-world applications demonstrates the effectiveness of FALCON, with significantly improved performance (by 300% for webmore »
-
Workflows are a widely used abstraction for describing large scientific applications and running them on distributed systems. However, most workflow systems have been silent on the question of what execution environment each task in the workflow is expected to run in. Consequently, a workflow may run successfully in the environment it was created, but fail on other platforms due to the differences in execution environment. Container-based schedulers have recently arisen as a potential solution to this problem, adopting containers to distribute computing resources and deliver well-defined execution environments to applications. In this paper, we con- sider how to connect workflow system to container schedulers with minimal performance loss and higher system efficiency. As an example of current technology, we use Makeflow and Mesos. We present five design challenges, and address them by using four configurations that connecting workflow system to container scheduler from different level of the infrastructure. In order to take full advantage of the resource sharing schema of Mesos, we enable the resource monitor of Makeflow to dynamically update the task resource requirement. We explore the performance of a large bioinformatics workflow, and observe that using Makeflow, Work Queue and the Resource monitor together not only increase themore »
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Publisher's Version of Record
- https://doi.org/10.1145/3472883.3486997
