DCAV: A software system to evaluate next‐generation cloud data center availability through a friendly graphical interface

Abstract

To assess the availability of different data center configurations, understand the main root causes of data center failures and represent its low-level details, such as subsystem's behavior and their interconnections, we have proposed, in previous works, a set of stochastic models to represent different data center architectures (considering three subsystems: power, cooling, and IT) based on the TIA-942 standard. In this paper, we propose the Data Center Availability (DCAV), a web-based software system to allow data center operators to evaluate the availability of their data center infrastructure through a friendly interface, without need of understanding the technical details of the stochastic models.DCAV offers an easy step-by-step interface to create and configure a data center model. The main goal of the DCAV system is to abstract low-level details and modeling complexities, becoming the data center availability analysis a simple and less time-consuming task.Next-generation cloud data center infrastructures are being built on a hyperscale, also known as composable data centers, focused on promoting flexibility, automation, optimization, and scalability. For that, the next-generation data center refactors traditional ones into pools of disaggregated resource units (processors, memory, storage, and networking, plus power and cooling sources). According to Li et al 5 this design approach "offers the potential advantage of enabling continuous peak workload performance while minimizing resource fragmentation for fast evolving heterogeneous workloads."Several industry initiatives have emerged to turn this paradigm a reality. The Intel Rack Scale Design (Intel RSD) is an architecture for disaggregated composable data center infrastructure based on Redfish a Distributed Management Task Force (DMTF) industry standard. Many partners have developed solutions based on Intel RSD and Redfish; for instance, Ericsson has proposed the Hyperscale Datacenter System 6 that allows data center operators to configure and control a composable data center by using virtual performance-optimized resources (named virtual Performance-Optimized Data Center (vPOD)). To compose vPOD systems, only the required hardware resources are allocated and may be reallocated easily on demand reducing management complexities, idle resources, energy consumption, and maintenance time, which reflects in higher availability levels.Despite this, data center managers or planners must know in detail the data center architecture low-level details; how the power, cooling and IT subsystems work; and how they are interconnected to plan the data center with minimum downtime. Furthermore, the management of challenges, such as guaranteeing service level agreements with established availability levels, requires data center's planners to assess the availability of a data center, resulting in a complex task as it requires gathering information of a large-scale infrastructure.

Availability modeling in redundant OpenStack private clouds

“…In this section, a dynamic method for modeling cloud infrastructure 11,[43][44][45] is presented. The presented method applies an algorithm to use a hierarchical modeling approach proposed in Section 4.1.…”

Availability analysis of design configurations to compose virtual performance‐optimized data center systems in next‐generation cloud data centers

“…Figure shows an example of a Tier IV of a power subsystem. For more detailed information about data center tiers, see Reference . Please note that, in this article, we consider the tiers from I to III, once we have estimated the reliability values of these tiers in our previous work …”

“…We specified the four vPOD configurations proposed in this research using the Redfish Emulator . Then, we used a software system named DCAV System to automatically collect the vPOD configuration from the Redfish Emulator, generate and solve the RBD and SPN models (using steady‐state analysis and stationary analysis techniques, respectively) proposed in this article. Finally, based on the RBD and SPN models, the DCAV System calculates the availability results for each vPOD configuration.…”