The impact

9 Genomic pipelines consist of several pieces of third party software and, because their experimental 10 nature, frequent changes and updates are commonly necessary thus raising serious distribution and 11 reproducibility issues. Docker containers technology offers an ideal solution, as it allows the 12 packaging of pipelines in an isolated and self-contained manner. This makes it easy to distribute and 13 execute pipelines in a portable manner across a wide range of computing platforms. Thus the 14 question that arises is to what extent the use of Docker containers might affect the performance of 15 these pipelines. Here we address this question and conclude that Docker containers have only a 16 minor impact on the performance of common genomic pipelines, which is negligible when the 17 executed jobs are long in terms of computational time. 18 Introduction 19 Genomic pipelines usually rely on a combination of several pieces of third party research software. 20 These applications tend to be academic prototypes that are often difficult to install, configure and 21 deploy. Furthermore their experimental nature can result in frequent updates, thus raising serious 22 reproducibility issues. In the past virtual machines were proposed as an answer to this issue. They 23 are indeed very convenient but come along with a few major issues that include high latency and 24 significant overhead. 25 Docker containers technology has been designed to address these issues. It has recently received an 26 increasing level of attention throughout the scientific community because it allows applications to run 27 in an isolated, self-contained package that can be efficiently distributed and executed in a portable 28 manner across a wide range of computing platforms. 29 The first most obvious advantage of this approach is to replace the tedious installation of numerous 30 pieces of software, with complex dependencies, by simply downloading a single pre-built ready-to-run 31 image containing all the software and the required configuration. 32 The second strength of Docker is to run each process in an isolated container that is created starting 33 from an immutable image. This prevents conflicts with any other installed program in the hosting 34 computing environment, and guarantees that each process runs in a predictable system configuration 35 that cannot change over time due to misconfigured software, system updates or programming errors. 36 Containers only require a few milliseconds to start and many instances can run in the same hosting 37 environment. This is possible because it runs as an isolated process in userspace on the host 38 operating system, sharing the kernel with other containers. 39 A study from IBM Research showed that Docker technology introduces a negligible overhead for CPU 40 and memory performance, and applications running in a container perform equally or better when 41 compared to KVM virtualization in all tests (1). 42 PeerJ PrePrints | https://dx.doi.org/10.7287/peerj.preprints.1171v2 | CC-BY 4.0 Open Access | rec: 12 Jun 2015, publ: 12 Jun 2015 P re P rin ts