[duke900ssd]

"It sounds like you don't know too much about SVC. Small number of connections? No, not at all. A full SVC will support non-cached 960,000 IOs per second. Hardly a small number of connections, huh?"

So this little PC has more I/O bandwidth between its four fibre adapters than half a dozen, or more, pSeries sytems have amongst their multiple adapters? Yeah, right.

"Each node is an IBM xSeries server - not nearly he same as a PC - running a specialized Linux kernel. Node are deployed in highly available node pairs called an IO group. In the even of the failure of one node, the other node in the IO group takes over all I/Os. If you loose a IO group, the IO is equally divided among all IO groups."

Said it your self, xSeries = x86 = PC, OMG! it runs linux too, so it's a linux PC.

"Each node has its own dedicated (and required) UPS."

Which is probably so small it only allows a controlled shutdown. Why is this required? I suspect because the poor little linux PC would dump the cache when the power went if it didn't have some warning and whould trash your data with no way to recover it. And, hey, we all love adding reliable things like a UPS or two as the only means to stand a chance of keeping our data safe.

"Keep in mind that the SVC functions as an in-band controller. You don't cable anything directly to the SVC. Traffic is routed through the SVC via SAN zoning and makes your switch zoning much simpler to manage."

Great if you can't zone your SAN properly, I suppose, otherwise it is just a major point of failure.

"SVC introduces no IO latency. To the contrary, the additional layer of caching that SVC introduces typically improves IO."

So your data I/O request goes from your server, to the SAN, then to the SVC, which thinks about it for a finite amount of time, then back to the SAN, then on to the storage, the I/O request is released back to the SAN, which passes it back to the SVC, which thinks about it for another finite amount of time, then it passes the data back to the SAN and on to the requesting system. How does this happen without any I/O latency? Magic?

"Please have a better idea what your speaking about before giving bad information."

Nuff said matey.