One-Shot architecture search, which aims to explore all possible operations jointly based on a single model, has been an active direction of Neural Architecture Search (NAS). As a well-known one-shot solution, Differentiable Architecture Search (DARTS) performs continuous relaxation on the architecture's importance and results in a bi-level optimization problem. However, as many recent studies have shown, DARTS cannot always work robustly for new tasks, which is mainly due to the approximate solution of the bi-level optimization. In this paper, one-shot neural architecture search is addressed by adopting a directed probabilistic graphical model to represent the joint probability distribution over data and model. Then, neural architectures are searched for and optimized by Gibbs sampling. We rethink the bi-level optimization problem as the task of Gibbs sampling from the posterior distribution, which expresses the preferences for different models given the observed dataset. We evaluate our proposed NAS method -- GibbsNAS on the search space used in DARTS/ENAS and the search space of NAS-Bench-201. Experimental results on multiple search space show the efficacy and stability of our approach.

Differentiable architecture search (DARTS) has been a promising one-shot architecture search approach for its mathematical formulation and competitive results. However, besides its caused high memory utilization and a large computation requirement, many research works have shown that DARTS also often suffers notable over-fitting and thus does not work robustly for some new tasks. In this paper, we propose a one-shot neural architecture search method referred to as MergeNAS by merging different types of operations e.g. convolutions into one operation. This merge-based approach not only reduces the search cost (about half a GPU day), but also alleviates over-fitting by reducing the redundant parameters. Extensive experiments on different search space and various datasets have been conducted to verify our approach, showing that MergeNAS can converge to a stable architecture and achieve better performance with fewer parameters and search cost. For test accuracy and its stability, MergeNAS outperforms all NAS baseline methods implemented on NAS-Bench-201, including DARTS, ENAS, RS, BOHB, GDAS and hand-crafted ResNet.

Under the consolidation scenario in Clouds, the network dimension should be considered as important as the computing power of machines. Traditional consolidation procedure is usually made according to the experience, which mainly focused on the hardware capability of the target system, like CPU, Memory and etc. Along with the consolidation of the computing power, the network communication among machines is also consolidated. The consolidation procedure needs to cover this change and avoid network problems after moving the applications into target virtualized system. This paper presents a novel approach to provide optimization taking network traffic into account during consolidation.

The proper operation of Service-Oriented Architecture (SOA) depends on underlying system services of operating systems, so efficient and effective migration of Service Hosting Environment is critical to cope with intrinsic-changed nature of SOA. However, due to the large amount of configuration items, complicated mapping and complex dependency relationship among system services, migrating into a new Service Hosting Environment satisfying the operation requirement of SOA becomes an error-prone and time-consuming task. The SCM project in IBM develops a novel approach to migrate Service Hosting Environment shaped in Unix-like systems. Firstly, this approach builds a set of configuration models to describe various system services. Then based on models, it presents knowledge based mapping to translate system service configurations between Service Hosting Environments. Finally, it designs a dependency hierarchy deducting algorithm to compute the dependency relationship among system services for migration traceability and error determination. A SCM prototype has performed well on largely reducing time, labor and errors in real migration cases.

Efficient problem detection methods play an important role in system management. In this paper, a formal method is described for problem detection in large scale and distributed enterprise IT environment. Events from distributed system components are collected, filtered and correlated. Leveraging these correlated events, the behavior of a distributed system is presented as a problem detection state machine (PDSM). PDSM is built up automatically from system logs without any specification of the target system. This approach combines logs from multi-sources and does not require any human involved or experimental instructions. It is generally applicable to a large class of distributed systems. Experimental results show that the implementation of PDSM performs problem detection efficiently in typical distributed enterprise systems.

The configuration of a service-based system has a significant impact on the nonfunctional requirements of the system as a whole. However, finding the best configuration is very challenging and sometimes impossible for administrators because so many factors have to be considered. More importantly, a service based system has to be frequently reconfigured to adapt to rapid and continuous changes in user requirements and runtime environments. In this paper we propose an autonomic computing approach to the problem of reconfiguration, that is, enabling the service based system to configure itself by means of a loop of monitoring, analyzing, planning and executing actions. We begin by formalizing the definition of configuration and reconfiguration. Then, we describe how we implemented the autonomic computing mechanisms for reconfiguring service-based systems to satisfy service level agreements with minimal resource consumption. The approach is demonstrated on a resilient service provisioning environment. Finally, the preliminary experiments are evaluated to determine the effectiveness of proposed approach.