In Search of the Ideal Storage Configuration for Docker Containers

Tarasov, Vasily; Rupprecht, Lukas; Skourtis, Dimitris; Warke, Amit; Hildebrand, Dean; Mohamed, Mohamed; Mandagere, Nagapramod; Li, Wenji; Rangaswami, Raju; Zhao, Ming

doi:10.1109/fas-w.2017.148

Cited by 25 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several Docker storage drivers were comprehensively evaluated in Refs. [12,13] and a number of key findings were obtained. For example, BtrFS performs better under Docker commands but has poor performance under real workloads, DeviceMapper has poor performance under Docker commands but performs well under actual workloads.…”

Section: Related Workmentioning

confidence: 99%

Optimizing the copy-on-write mechanism of docker by dynamic prefetching

Jiang

Liu

Shi

et al. 2021

Tinshhua Sci. Technol.

View full text Add to dashboard Cite

Docker, as a mainstream container solution, adopts the Copy-on-Write (CoW) mechanism in its storage drivers. This mechanism satisfies the need of different containers to share the same image. However, when a single container performs operations such as modification of an image file, a duplicate is created in the upper readwrite layer, which contributes to the runtime overhead. When the accessed image file is fairly large, this additional overhead becomes non-negligible. Here we present the concept of Dynamic Prefetching Strategy Optimization (DPSO), which optimizes the CoW mechanism for a Docker container on the basis of the dynamic prefetching strategy. At the beginning of the container life cycle, DPSO pre-copies up the image files that are most likely to be copied up later to eliminate the overhead caused by performing this operation during application runtime. The experimental results show that DPSO has an average prefetch accuracy of greater than 78% in complex scenarios and could effectively eliminate the overhead caused by the CoW mechanism.

show abstract

Section: Related Workmentioning

confidence: 99%

Optimizing the copy-on-write mechanism of docker by dynamic prefetching

Jiang

Liu

Shi

et al. 2021

Tinshhua Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…For the Docker images we make use of the Docker native capabilities of building images based on Dockerfiles. The VM images are generated by Packer 13 . Packer processes a Packerfile, which is similar to a Dockerfile, but uses a multitude of different virtualization providers to generate and store the image.…”

Section: Portability and Reproducibilitymentioning

confidence: 99%

“…An analysis and evaluation of the Docker storage drivers with respect to filesystem performance is presented by [13]. The results demonstrate that the choice of the storage driver can influence the filesystem performance significantly where the Btrfs storage driver achieves the best performance but less stability as the other storage drivers.…”

Section: Performance Overhead and Resource Isolationmentioning

confidence: 99%

The Impact of the Storage Tier: A Baseline Performance Analysis of Containerized DBMS

Seybold¹,

Hauser²,

Eisenhart³

et al. 2018

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Containers emerged as cloud resource offerings. While the advantages of containers, such as easing the application deployment, orchestration and adaptation, work well for stateless applications, the feasibility of containerization of stateful applications, such as database management system (DBMS), still remains unclear due to potential performance overhead. The myriad of container operation models and storage backends even raises the complexity of operating a containerized DBMS. Here, we present an extensible evaluation methodology to identify performance overhead of a containerized DBMS by combining three operational models and two storage backends. For each combination a memory-bound and disk-bound workload is applied. The results show a clear performance overhead for containerized DBMS on top of virtual machines (VMs) compared to physical resources. Further, a containerized DBMS on top of VMs with different storage backends results in a tolerable performance overhead. Building upon these baseline results, we derive a set of open evaluation challenges for containerized DBMSs.

show abstract

“…Third, the kernel often introduces software overheads (e.g., locking) whose mitigation requires complex restructuring of the application or system code [47,71]. Fourth, the implementation of new filesystems to meet the container storage needs follows the traditional kernel software development that is time-consuming [8,40,67]. Finally, the shared storage I/O path makes the colocated tenants more vulnerable to attacks or bugs that discourage the wider container adoption [22,56,65].…”

Section: Introductionmentioning

confidence: 99%

A user-level toolkit for storage I/O isolation on multitenant hosts

Kappes

Anastasiadis

2020

Proceedings of the 11th ACM Symposium on Cloud Computing

View full text Add to dashboard Cite

Software containers limit the performance of data-intensive applications due to storage I/O contention in the system kernel of the host. Although kernel partitioning has been claimed as a promising approach, it is impractical due to implementation complexity. As a pragmatic alternative, we suggest that the tenants of a host should run their own userlevel filesystems. We introduce the Polytropon toolkit as a collection of user-level components configurable to build several types of filesystems. The toolkit provides an application library to invoke the standard I/O calls, a user-level path to isolate the tenant I/O traffic to private host resources, and user-level filesystem services distinct per tenant. Furthermore, we introduce the RCQB concurrent queue that relaxes operation ordering for improved communication throughput, and we provide the SMO pipelined memory copy that is faster than standard methods. We use Polytropon to build the client of a distributed filesystem optionally combined with a union filesystem. Polytropon serves 32 tenants with RocksDB over network storage with 7.2-14x lower latency than kernel systems, and reduces up to 2.9x the timespan to execute source-code processing in the containers of 32 tenants. RCQB achieves higher performance than the state-ofthe-art queues by 4-52x, while SMO achieves 29-66% higher data transfer throughput than existing methods.

show abstract

In Search of the Ideal Storage Configuration for Docker Containers

Cited by 25 publications

References 9 publications

Optimizing the copy-on-write mechanism of docker by dynamic prefetching

Optimizing the copy-on-write mechanism of docker by dynamic prefetching

The Impact of the Storage Tier: A Baseline Performance Analysis of Containerized DBMS

A user-level toolkit for storage I/O isolation on multitenant hosts

Contact Info

Product

Resources

About