Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : The New QLogic (ANCR) -- Ignore unavailable to you. Want to Upgrade?

To: Craig Stevenson who wrote (14866)	3/4/1998 1:46:00 PM
From: George Dawson	Read Replies (2) \| Respond to of 29386

Craig, I am reading a lot about clustering in the book I got last weekend: Message 3563817 *In Search of Clusters* by Gregory F. Pfister (Prentice Hall, Upper Saddle River, NJ 1998). Pfister explains the hardware, software, strategies, and economics of clusters. He also talks about some I/O rules for clusters and attached peripheral devices that make a lot of sense, but that I have not seen articulated in other areas. For example, Internode communication bandwidth of a cluster for I/O intensive workloads should equal the aggregate bandwidth from all other I/O sources that each node has. He goes on to illustrate that if you have N Mbytes/sec of aggregate bandwidth to all disks, tape, and I/O devices attached to a node that the internode communication for this node (if it is in cluster) should be N Mbytes/sec. The book has hundreds of similar pearls and abundant examples. I would recommend it to anyone with an interest in knowing more about what comes up on this thread on a frequent basis. It also has implications for FC as both an internode communication and attached storage application. In the acknowledgements, he refers to the CXT 1000 as a trademark of Ancor-but I have not come across it in the text yet. George D.