Docker containers enable to package an application together with all its dependencies and easily run it in any environment. Thanks to their ease of use and portability, containers are gaining an increasing interest and promise to change the way how Cloud platforms are designed and managed. For their execution in the Cloud, we need to solve the container deployment problem, which deals with the identification of an elastic set of computing machines that can host and execute those containers, while considering the diversity of their requirements.In this paper, we provide a general formulation of the Elastic provisioning of Virtual machines for Container Deployment (for short, EVCD) as an Integer Linear Programming problem, which takes explicitly into account the heterogeneity of container requirements and virtual machine resources. Besides optimizing multiple QoS metrics, EVCD can reallocate containers at runtime, when a QoS improvement can be achieved. Using the proposed formulation as benchmark, we evaluate two well-known heuristics, i.e., greedy first-fit and round-robin, that are usually adopted for solving the container deployment problem.

Elastic Provisioning of Virtual Machines for Container Deployment

Today, a new technology is going to change the way platforms for the internet of services are designed and managed. This technology is called container (e.g. Docker and LXC). The internet of service industry is adopting the container technology both for internal usage and as commercial offering. The use of container as base technology for large-scale systems opens many challenges in the area of resource management at run-time, for example: autoscaling, optimal deployment and monitoring. Specifically, monitoring of container based systems is at the ground of any resource management solution, and it is the focus of this work. This paper explores the tools available to measure the performance of Docker from the perspective of the host operating system and of the virtualization environment, and it provides a characterization of the CPU and disk I/O overhead introduced by containers.

/pdf/measuring-docker-performance-what-a-mess-58hddf6x16.pdf

Measuring Docker Performance: What a Mess!!!

Containerization is a lightweight virtualization technology enabling the deployment and execution of distributed applications on cloud, edge/fog, and Internet-of-Things platforms. Container technol ...

The state‐of‐the‐art in container technologies: Application, orchestration and security

Today, The cloud industry is adopting the container technology both for internal usage and as commercial offering. The use of containers as base technology for large-scale systems opens many challenges in the area of resource management at run-time. This paper addresses the problem of selecting the more appropriate performance metrics to activate auto-scaling actions. Specifically, we investigate the use of relative and absolute metrics. Results demonstrate that, for CPU intense workload, the use of absolute metrics enables more accurate scaling decisions. We propose and evaluate the performance of a new autoscaling algorithm that could reduce the response time of a factor between 0.66 and 0.5 compared to the actual Kubernetes' horizontal auto-scaling algorithm.

Auto-Scaling of Containers: The Impact of Relative and Absolute Metrics

Object detection and recognition is commonly used in diverse computer vision based applications and many algorithms are proposed in literature. However, application of object detection algorithms in real-time systems demand minimal computation time and comprehensive performance analysis needs to be conducted before deployment. This paper aims to conduct performance analysis of deep learning based algorithm i.e. single shot detector (SSD) in IoT based embedded devices for smart home appliances control. We have developed a smart home automation system using object detection algorithm based on model view controller (MVC) architecture deployed on Cloud of Things (CoT) i.e. Amazon Web Service (AWS) cloud for users to remotely monitor their homes. Message queuing telemetry transport (MQTT) protocol is used for communication with connected IoT devices. For load-balancing, we have proposed the concept of distributed broker to support high number of publishers and subscribers. For experimental analysis, we have connected a camera with Raspberry Pi for object detection based on deep learning algorithm (SSD) using OpenCV library in our proposed system. Experiments are conducted to evaluate the performance of object detection algorithm under varying environmental conditions by changing the light intensity level, distance of object from camera, and frame size of video. Results show that communication delay is very low (i.e. 0.2 s) as compared to processing delay in Raspberry Pi. Furthermore, changing environmental conditions have very low/insignificant impact on the processing delay of object detection algorithm i.e. average delay of 1.7 s (stdev. 0.18) and 1.8 s (stdev. 0.24) in bright and dark lighting levels, respectively. However, accuracy is deteriorated under low lighting intensity level and increased frame sizes i.e. from 95–100 to 80–85%. Selected embedded device, camera model and object detection algorithm limits the performance of object detection in real-time systems and shall be carefully selected to fulfill the requirement.

Object detection mechanism based on deep learning algorithm using embedded IoT devices for smart home appliances control in CoT

Nowadays, the Closed-Circuit Television (CCTV) surveillance system is being utilized in order to keep peace and provide security to people. There are several defects in the video surveillance system, such as: picture is indistinct, anomalies cannot be identified automatically, a lot of storage spaces are needed to save the surveillance information, and prices remain relatively high. This paper describes the design and implementation of a low-cost system monitoring based on Raspberry Pi, a single board computer which follows Motion Detection algorithm written in Python as a default programming environment. In addition, the system uses the motion detection algorithm to significantly decrease storage usage and save investment costs. The algorithm for motion detection is being implemented on Raspberry Pi, which enables live streaming camera along with detection of motion. The live video camera can be viewed from any web browser, even from mobile in real-time.

Low cost real-time system monitoring using Raspberry Pi

We consider the problem of similarity search over the large datasets in the distributed environment. The proposed framework employs the Vp-Tree algorithm that integrated on top of the MapReduce framework to achieve good performance as well as meet the scalability and fault tolerance requirements for the system while data scale up. Since VP-Tree algorithm was implemented initially for partition and searching data in the local disk access, we proposed a new approach to using it in the parallel environment. The key point of the Vp-Tree algorithm is that it distributed the similar data points into groups, thereby reducing number of data need to scan during the searching stage. Consequently, the response time of the entire system has been improved. Otherwise, we used an open source computer vision library OpenCV for detect the similarity among images in the dataset. We evaluate the performance of our proposed framework using a synthetic data to show the positive of our approach. The experiment shows that our proposed framework achieves 57% improvement in response time in comparison with running searching job in tradition Hadoop framework. We also compared our application running time on Docker container against VM-based environment. The result points out that deploy our system over Docker container provide higher performance than VM-based environment in term of response time.

An Index Scheme for Similarity Search on Cloud Computing using MapReduce over Docker Container

Screen content has some different characteristics from camera-captured content, such as large motion and repeating patterns which lead to low encoding speed and higher bit-rate. To cope with these problems, homography-based motion detection is proposed to better explore the temporal correlation in screen content. After detecting motion, the motion parameters are forwarded to JPEG encoder for motion compensated predictions. Therefore, it can improve the coding efficiency. Experimental results show the proposed algorithm achieves efficiency in terms of both encoding time and encoding complexity.

Homography-Based Motion Detection in Screen Content

According to the application demands and characteristics of smart home control system, the home smart system is analyzed and studied. The kind of smart home control system based on Raspberry Pi is also designed. Our system is designed to assist and provide support in order to fulfill the needs of elderly and disabled in home. Also, the smart home concept in the system improves the standard living at home. This paper presents the overall design of Home System with low cost; we used a Raspberry Pi Kit that communicates with an Android application. Wifi module in Raspberry Pi received the signal and processed the input signal to control the home devices. The proposed system intended to control electrical appliances with relatively user-friendly interface and ease of installation.

Ton Thi Kim Loan

Papers

Low cost real-time system monitoring using Raspberry Pi

An Index Scheme for Similarity Search on Cloud Computing using MapReduce over Docker Container

Homography-Based Motion Detection in Screen Content

Android-based Low Cost Control System for Smart Home using Raspberry Pi