[Info-vax] Message from HP.

[alanfeldman48 at gmail.com]

On Tuesday, December 10, 2013 5:35:08 AM UTC-5, JF Mezei wrote:
> On 13-12-10 03:17, David Froble wrote:
> 
[...]
> 
> So why is NSK saved while VMS and HP-UX tied to a pipe on E deck of the
> 
> sinking Itanic ?
> 
[...]
> 
> > Look at entities such as Google assembling their own hardware.
> 
> 
> 
> Google doesn't care when a service goes down. People don't notice that a
> 
> search yielded only 100 results instead of 1000 because one or more
> 
> servers were down.

That's because Google can't count. I recall getting results like 1-7 out of 
100. Or 1-100 out of 7. I've heard stories like this from others. Well, at 
least one, anyway. :-)

And their new newsreader is in many ways worse than the old one. Just like iOS 
7 (!)

[...]
> 
> 
> But there are ways to assemble a system where you can hot swap
> 
> components that have failed, there are ways where there is some hardware
> 
> supervisor that detects hardware faiults and failsover at hardware level
> 
> etc etc.
> 
> 
> 
> Go back to last century with the VAXft. It was both a hardware and
> 
> software solution.
> 
> 
> 
> So new NSK servers may have come commodity components, but they will
> 
> also have some specialised components to provide the failt
> 
> detection/failover/tolerance that customers want. (and components choses
> 
> to enable hot-swapping)

How do they compare to Stratus, like their XA2000? We had one at work up until 
a few years ago. It had redundant everything except tape drive and backplane. 
It contained three sections: CPU, Mem, and I/O. The boards were all "mirrored". 
So you could do hot swaps of any of them! (Unless both boards of a pair die 
together, of course.)

[...]

AEF