Friedman R., and Birman, K.P., Using Group Communication Technology to Implement a Reliable and Scalable Distributed IN Coprocessor. TINA 96 Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
An Overview of the Galaxy Management Framework for Scalable.. - Vogels, Dumitriu (2000) (Correct)
....in [16] Scalability in cluster systems has been addresses in cluster research, but mostly from an application specific viewpoint: The Inktomi [10] technology is a highly scalable cluster targeted towards Internet search engine and document retrieval. A cluster based SS7 call processing center [11] with good scalability from a real time perspective, was build in 1998 using Cornell communication technology. Another cluster system that claims excellent scalability is Porcupine, which is very specifically targeted towards Internet mail processing systems [20] In terms of specific distributed ....
Friedman R., and Birman, K.P., Using Group Communication Technology to Implement a Reliable and Scalable Distributed IN Coprocessor. TINA 96
Design and Modelling of a Parallel Data Server for Telecom.. - Ronström (1997) (1 citation) (Correct)
....secondary indexes, foreign keys and schema changes. Also ideas from [CHAMB92] and [KARL96] have been further developed to achieve a copy fragment algorithm where only one tuple at a time is locked and shipped. CHAMB92] locks the whole fragment and ships it and [KARL96] ships one bucket at a time. [FRIED96] reports work on a prototype for Intelligent Network applications. It is optimised for simple read transactions. The communication is efficient using an IBM SP2 communication switch between 12 RS 6000 workstations. On top of the communication they use Horus, a group communication technology. Horus ....
....Our estimated results shows that 6,000 simple read transactions per second per processor node should be achievable with an Ultra 2 Server. This is achieved without packing messages and without an architecture which is highly optimised for simple reads. Discovering a failed node took six seconds in [FRIED96], 100 ms in [ClustRa95] In this work we assume that we will be able to discover failed nodes in 20 30 ms. ClustRa95] achieve a take over time of 0.3 seconds which is excellent. Our scheme performs the take over in 1 ms after the failed 233 node has been discovered. This take over means ....
[Article contains additional citation context not shown here]
R. Friedman, K. Birman, Using Group Communication Technology to Implement a Reliable and Scalable Distributed IN Coprocessor, submitted to "TINA" conference.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC