Asynchronous Migration of Process Replicas in a Cluster

Abstract

Application processes have to be efficiently and reliably performed on servers in a cluster. A process is replicated to increase the reliability and availability. However, the more number of replicas of a process are performed, the more reliable and available the system is but the more amount of electric energy is consumed. In this paper, we take a process migration (MG) approach to energy-efficiently and reliably performing an application process on servers in a cluster. A process performed on a current server is migrated to another server if the server is expected to consume a smaller amount of electric energy to perform the process than the current server while the deadline constraint of the process is satisfied. In order to reliably perform a process, multiple replicas of the process are performed on different servers. We consider synchronous and asynchronous ways to migrate process replicas to servers. In the synchronous migration, every replica moves to another server on a same computation state. In the asynchronous migration, each replica by itself makes a decision on when the replica is migrated to another server.

Energy-Aware Migration and Replication of Processes in a Cluster

“…In this paper, we consider another model, i.e. replica migration model [10], [11] where each replica of a process is performed on a server and may migrate to another server.…”

“…2 The replica instance p k ijt migrates to another server s u and a succeeding replica instance p k+1 iju is performed on the server s u . In papers [9], [10], [11], the model to estimate the expected energy consumption of each server s t to perform every current process and expected termination time of each current process on the basis of the SPC and SC models. Let EE t and EE u be the expected electric energy [Ws] consumed by the server s t and s u , respectively.…”

Simple estimation and energy‐aware migration models of virtual machines in a server cluster

“…In the work of Kataoka et al, an algorithm to estimate the termination time of each process is proposed based on the computation model. However, it is not easy to estimate the termination time of each current process p i on every virtual machine v m h of a server s t by using the computation laxity l c t i ( τ ) and process computation rate N P R t i ( n ) of each process p i where n (=| C P t ( τ )|) processes are concurrently performed . In order to make the estimation simple and practical, we assume the total number V C i of virtual computation steps of each process p i to be a constant, ie, V C = 1.…”

“…After an application process is issued to a host server selected by a selection algorithm, the host server might consume more electric energy than estimated to perform application processes. In the migration approach, an application process on a host server migrates to an energy‐efficient guest server which is expected to consume smaller electric energy than the host server after the application process is started on the host server. However, it is not easy to migrate types of application processes among heterogeneous servers with different architectures and operating systems.…”

An Energy-Aware Migration of Virtual Machines

“…A process migration approach is also discussed where a process on a host server migrates to another server if the host server is expected to consume more electric energy e.g. due to overloads [5], [6], [7], [8], [10]. In information systems, processes are required to be not only energyefficiently but also reliably performed.…”