Resource Management Services for a Grid Analysis Environment

Ali, Arshad; Anjum, Ashiq; Azim, Tahir; Bunn, J.; Mehmood, Atif; McClatchey, Richard; Newman, H. B.; Rehman, W. ur; Steenberg, Conrad; Thomas, M.; Lingen, Frank van; Willers, Ian; Zafar, Muhammad Adeel

doi:10.1109/icppw.2005.76

Cited by 7 publications

(1 citation statement)

References 17 publications

(10 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ali et al [18] have described a job monitoring service which The contributions of this research are retrieves job status information from the schedulers. However, there are a few shortcomings in their service.…”

mentioning

confidence: 99%

MSIS: Meta-scheduler Infornation Service for Resource Performance Predictors

Bashir

Nagar

Tawfik

2007

2007 2nd International Conference on Communication Systems Software and Middleware

View full text Add to dashboard Cite

Contemporary Grid monitoring services lack the resources might lead to delayed, if not wrong, scheduling and functionality to monitor batch-schedulers' job queue information management decisions. For example a pre-emption process and the status of jobs submitted. Moreover, the monitored data could be initiated even when the situation on the ground has delivered by them is only virtually real-time. The former implies improved to normal. Consequently, it will induce an additional that these services are not capable of monitoring a batchoverhead instead of benefits. scheduler's inability to process a particular job. The latter implies that the strategies based solely upon the currency of Secondly, the inability of contemporary Grid monitoring monitored information can lead to delayed, if not wrong, services to monitor batched jobs makes meta-schedulers prone scheduling and management decisions. Batch-scheduler to mistakes like delegating jobs to a batch-scheduler where a monitoring is fundamentally imperative for future resource site is continuously failing a particular job. For e.g. according performance predictions that help meta-schedulers in deciding to condor manual [5] condor lacks the ability to reschedule the optimal schedules for deadline-constrained jobs. We have job elsewhere if a failing site still holds better credentials in developed a meta-scheduler information service (MSIS), which terms of least load and more free resources etc i.e. better load monitors the batch-schedulers for up-to-date information about index. It can result in a batch-scheduler scheduling the same their job queues, current status of the jobs and the underlying job to the same site which has a history of failing that job in the resources. MSIS can be used in its two totally independent earlier attempts.modes: as a retrofit API library to enhance capabilities of contemporary Grid monitoring services and as a web service. It, Above convinces us about the importance of monitoring job in combination with any Grid monitoring service, provides status in conjunction with the performance of underlying comprehensive information which can be used in estimating a resources simultaneously. It is imperative to note that resource's future state over a limited period of time.contemporary resource managers and batch-schedulers employ monitoring mechanisms that give them the ability to only I. INTRODUCTION update the availability of underlying resources which is , as already highlighted, only virtually real-time. In fact, what The dominion of meta-schedulers spans over multiple actually is required is the information through which we could batch-schedulers, which are geographically distributed. Their estimate the future availability of the underlying resources in resource management concerns include a combination of all terms of their performance capabilities at any future instant communicational, computational and data resources. and/ or a limited duration of time. However, the monitoring of Monitoring of dynamic status of resources is already an...

show abstract

mentioning

confidence: 99%

MSIS: Meta-scheduler Infornation Service for Resource Performance Predictors

Bashir

Nagar

Tawfik

2007

2007 2nd International Conference on Communication Systems Software and Middleware

View full text Add to dashboard Cite

show abstract

A standards‐based Grid resource brokering service supporting advance reservations, coallocation, and cross‐Grid interoperability

Elmroth

Tordsson

2009

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYThe problem of Grid-middleware interoperability is addressed by the design and analysis of a featurerich, standards-based framework for all-to-all cross-middleware job submission. The service implements a decentralized brokering policy, striving to optimize the performance for an individual user by minimizing the response time for each job submitted. The architecture is designed with focus on generality and flexibility and builds on extensive use, internally and externally, of (proposed) Web and Grid services standards such as WSRF, JSDL, GLUE and WS-Agreement. The external use provides the foundation for easy integration into specific middlewares, which is performed by the design of a small set of plugins for each middleware. Currently, plugins are provided for integration into Globus Toolkit 4, NorduGrid/ARC, and LCG2. The internal use of standard formats facilitates customization of the job submission service by replacement of custom components for performing specific well-defined tasks. Most importantly, this enables the easy replacement of resource selection algorithms by algorithms that addresses the specific needs for a particular Grid environment and job submission scenario. The algorithms in our implementation perform resource selection based on performance predictions, and provide support for advance reservations as well as coallocation of multiple resources for simultaneous job start. The performance of the system is analyzed with focus on overall job service throughput (up to over 250 jobs/minute) and individual job response time (down to under one second).

show abstract