Failure Recovery in Distributed Environments with Advance Reservation Management Systems

Abstract

Abstract. Resource reservations in advance are a mature concept for the allocation of various resources, particularly in grid environments. Common grid toolkits such as Globus support advance reservations and assign jobs to resources at admission time. While the allocation mechanisms for advance reservations are available in current grid management systems, in case of failures the advance reservation perspective demands for strategies that support more than recovery of jobs or applications that are active at the time the resource failure occurs. Instead, also already admitted, but not yet started applications are affected by the failure and hence, need to be dealt with in an appropriate manner. In this paper, we discuss the properties of advance reservations with respect to failure recovery and outline a number of strategies applicable in such cases in order to reduce the impact of resource failures and outages. It can be shown that it pays to remap also affected but not yet started jobs to alternative resources if available. Alike reserving in advance, this can be considered as remapping in advance. In particular, a remapping strategy that prefers requests that were allocated a long time ago, provides a high fairness for clients as it implements similar functionality as advance reservations, while achieving the same performance as the other strategies.

VRM: a failure-aware Grid resource management system

“…The general benefits of remapping in advance are shown in [4] where estimations of the failure duration were used. As the end of the failure is usually unknown, it is not easy to decide which active jobs on the resource have to be taken into account for remapping.…”

A distributed load-based failure recovery mechanism for advance reservation environments

“…In [7], the basic requirements and opportunities for failure recovery in planning based resource management systems have been examined. In particular, it was shown that remapping admitted but not yet active jobs is essential in order to reduce the number of terminated jobs.…”

“…However, estimations of the actual downtime are a questionable approach as these estimations are inherently unreliable and underestimations lead to a significantly higher termination ratio than possible with exact knowledge. In this paper, we extend the approach of remapping in advance from [7] and provide strategies that determine the remapping interval depending on the actual load on the available resources in the grid rather than on estimations which cannot be expected to be accurate.…”

“…The benefits of remapping in advance are shown in [7]. As the end of the failure will be usually unknown it is not easy to decide which active jobs on the resource have to be taken into account for remapping.…”

“…As shown in [7], predictions of the actual downtime are critical, as the downtime cannot be accurately anticipated. Underestimations in this case lead to a drastically increased number of terminated jobs.…”