Design & analysis of fault tolerant digital systems
Design & analysis of fault tolerant digital systems
In search of clusters (2nd ed.)
In search of clusters (2nd ed.)
PARMON: a portable and scalable monitoring system for clusters
Software—Practice & Experience
Network Management Standards: SNMP, CMIP, TMN, MIBs and Object Libraries
Network Management Standards: SNMP, CMIP, TMN, MIBs and Object Libraries
IPDPS '01 Proceedings of the 10th Heterogeneous Computing Workshop â"" HCW 2001 (Workshop 1) - Volume 2
A gossip-style failure detection service
Middleware '98 Proceedings of the IFIP International Conference on Distributed Systems Platforms and Open Distributed Processing
| Hi-index | 0.00 |
Clusters of workstations are increasingly used forapplications requiring high levels of both performanceand reliability.Certain fundamental services are highlydesirable to achieve these twin goals of network-basedcluster system analysis and management.Among theseservices is the ability to detect network and node failuresand the capability to efficiently determine computer andnetwork load levels.Furthermore, the ability to allow forthe distribution of administrative directives is alsointegral to the goal of cluster management.This paperpresents a scalable approach to providing these vitalsupport capabilities for distributed computing integratedinto a cluster management system.Previous approachesto cluster management have suffered from problems ofscalability and the inability to properly supportheterogeneous systems in a non-proprietary fashion.Thiscluster management system employs gossip techniques toaddress the problem of scalability in network-basedsystem management.The results of two case studies showthat the cluster management system is scalable and haslittle adverse impact on the performance of sequentialand parallel applications running on the managed system.