HyperPlex --?

[ln_unix]

Hi All,

Am new to partitions..

and want to know about the HyperPlex?

Just I know that it comes under Hard Partitioning and is another form of nPar.

Waiting for your most easy understandable replies.

Thank You Very Much.

Regards,
LN

[Alzhy]

HyperFlex I thought was a proprietary Clustering / Interconnect solution. What it does is a high speed interconnet of several high end servers - SuperDomes(?) so you have a single system with a large number of CPUs acting as one. Mostly used with SuperCOmputing implementations.

It is one of the older clustering technolgies that I doubt people are still seriously looking at as there are other lower cost (Mainly Linux) coolutions out there plus InfiniBand.

[ln_unix]

Thanks Alzhy for your reply...

But I heard that HyperPlex is another form of hard partitioning...

Isn't it that ????

LN

[Alzhy]

Well... in a way yes. Since it GLUES several SuperDomes for instane as one via so called HyperFabric interconnects.

But let me ask you -- what are you trying to accomplish and why the interest with HyperFlex? DO you have a need to have moer than 128 CPU cores on a SuperDome?


I think it is "old" and dated and there are other cluster interconnects and schemes available today -- most of which are Linux based though.

[ln_unix]

NO Alzhy, I am just trying to capture the term in a broader way in my knowledgebase.

So, If you know much about the same ,Please tell me.Want to know more about the same.

Thank You Very Much for Your immediate replies :)

LN

[Viktor Balogh]

Hi,

here are some docs about HP's HyperFabric solution:

docs.hp.com/en/B6257-90032/B6257-90032.pdf

Regards,
Viktor

[Viktor Balogh]

https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=Hyper02

Overview

HyperFabric is a networking technology that runs on various HP 9000 and Itanium servers and workstations, allowing very high speed and low latency in communication between systems. It supports TCP/IP and Hyper Messaging Protocol(HMP). HyperFabric clusters use switched fabric technology to provide high speed system interconnect. HMP provides a zero-copy reliable datagram protocol which enables middleware products to use a low latency high bandwidth clustering solution.

what's new

HyperFabric version B.11.23.02 includes the transparent local failover feature on Hyper Messaging Protocol (HMP). If a HyperFabric resource (adapter, cable, switch or switch port) fails in a fabric, HMP now has the ability to transparently fail over traffic using other available resources. This feature is available only for customers who run Oracle RAC 9i using HMP on HyperFabric2 (HF2) hardware.

features and benefits

* Adapter is capable of link speed upto 2.0 Gbps each way
* Support for IP and HMP
* Very low latency
* High availability
* Integrated with EMS and ServiceGuard

A HyperFabric (HF) configuration contains two or more HP systems connected using HF adapters, cables and optionally HF switches. Each HP system acts as a node in the configuration. HF switches can be meshed, and configurations with up to four levels of meshed switches are supported.