Resource usage of Windows computer laboratories

Studies focusing on Unix have shown that the vast majority of workstations and desktop computers remain idle for most of the time. In this paper we quantify the usage of main resources (CPU, main memory, disk space and network bandwidth) of Windows 2000 machines from classroom laboratories. For that purpose, 169 machines of 11 classroom laboratories were monitored over 77 consecutive days. Samples were collected from all machines every 15 minutes for a total of 583653 samples. Besides evaluating availability of machines (uptime and downtime) and usage habits of users, the paper assesses usage of main resources, focusing on the impact of interactive login sessions over resource consumptions. Also, resorting to Self Monitoring Analysis and Reporting Technology (SMART) parameters of hard disks, the study estimates the average uptime per hard drive power cycle for the whole life of monitored computers. Our results show that resources idleness in classroom computers is very high, with an average CPU idleness of 97.9%, unused memory averaging 42.1% and unused disk space of the order of gigabytes per machine. Moreover, this study confirms the 2:1 equivalence rule found out by similar works, with N non-dedicated resources delivering an average CPU computing power roughly similar to N/2 dedicated machines. These results confirm the potentiality of these systems for resource harvesting, especially for grid desktop computing schemes.

[1]  John R. Douceur.  Is remote host availability governed by a universal law? , 2003, PERV.

[2]  David E. Culler,et al.  A case for NOW (networks of workstation) , 1995, PODC '95.

[3]  Andrew A. Chien,et al.  Entropia: architecture and performance of an enterprise desktop grid system , 2003, J. Parallel Distributed Comput..

[4]  Luis F. G. Sarmenta,et al.  Bayanihan: Web-Based Volunteer Computing Using Java , 1998, WWCA.

[5]  Marvin Theimer,et al.  Feasibility of a serverless distributed file system deployed on an existing set of desktop PCs , 2000, SIGMETRICS '00.

[6]  Joel H. Saltz,et al.  The utility of exploiting idle workstations for parallel computation , 1997, SIGMETRICS '97.

[7]  Gilles Fedak,et al.  XtremWeb: a generic global computing system , 2001, Proceedings First IEEE/ACM International Symposium on Cluster Computing and the Grid.

[8]  Paulo Marques,et al.  Distributed data collection through remote probing in windows environments , 2005, 13th Euromicro Conference on Parallel, Distributed and Network-Based Processing.

[9]  David E. Culler,et al.  A case for NOW (networks of workstation) , 1995, PODC '95.

[10]  Nicolas Maillard,et al.  The Virtual Cluster: A Dynamic Environment for Exploitation of Idle Network Resources , 2002 .

[11]  J BoloskyWilliam,et al.  Feasibility of a serverless distributed file system deployed on an existing set of desktop PCs , 2000 .

[12]  Jeffrey K. Hollingsworth,et al.  Unobtrusiveness and efficiency in idle cycle stealing for PC grids , 2004, 18th International Parallel and Distributed Processing Symposium, 2004. Proceedings..

[13]  César A. F. De Rose,et al.  The virtual cluster: a dynamic network environment for exploitation of idle resources , 2002, 14th Symposium on Computer Architecture and High Performance Computing, 2002. Proceedings..

[14]  Joseph F. Murray,et al.  Improved disk-drive failure warnings , 2002, IEEE Trans. Reliab..

[15]  Andrew A. Chien,et al.  Henri Casanova , 2022 .

[16]  Andrea C. Arpaci-Dusseau,et al.  The interaction of parallel and sequential workloads on a network of workstations , 1995, SIGMETRICS '95/PERFORMANCE '95.

[17]  Renato J. O. Figueiredo,et al.  A case for grid computing on virtual machines , 2003, 23rd International Conference on Distributed Computing Systems, 2003. Proceedings..

[18]  Walfredo Cirne,et al.  Non-Dedicated Distributed Environment: A Solution for Safe and Continuous Exploitation of Idle Cycles , 2005, Scalable Comput. Pract. Exp..

[19]  Sanjeev Setia,et al.  Availability and utility of idle memory in workstation clusters , 1999, SIGMETRICS '99.

[20]  Peter A. Dinda,et al.  Measuring and understanding user comfort with resource borrowing , 2004, Proceedings. 13th IEEE International Symposium on High performance Distributed Computing, 2004..