On 12/31/2015 04:58 PM, Anton Aylward wrote:
On 12/31/2015 04:41 PM, James Knott wrote:
IIRC, a few years ago, IBM ran abut 50,000 instances of Linux in virtual machines on one of their "big iron" computers. Yes, I recall reading about that.
It strikes me that if you were implementing an ISP, that is the way to go, completely software managed. Those mainframes are ultra reliable; that many Intel boxes would be failing daily, one way or another, not least of all because of the proliferation of components: wiring, routers/switches, power supplies, RAM chips and so on. To say nothing of the #1 failure mode that has beset electronics (and before that, steam engines) for the last 100+ years -- solder joints. A completely software defined data centre on one of these boxes would be the way to go. The latest are compact, would fit under your desk. The wall-to-wall RAID array and HVAC+3-phase to ran it is another matter.
I used to work at IBM and saw some of those big systems. However, these days my work often takes me into data centres, where there are many racks, each filled with several servers. A couple of days ago, I was in one that used to be an IBM warehouse. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org