[Info-vax] Marketing ideas for VSI ?

[Bill Gunshannon]

On 12/15/18 1:37 PM, Dave Froble wrote:
> On 12/15/2018 10:33 AM, Kerry Main wrote:
> 
>> Actually network latency is primarily based on speed of light (hops 
>> also),
>> and this has not changed since the speed of light was originally 
>> calculated.
>>
>> Other factors to consider for long distance clusters and how far apart 
>> the
>> can be is the R/W ration of the application This is because active-active
>> clusters with load balancing across both sites requires  synch 
>> replication.
>> A heavy write app will mean you will likely need to have sites closer
>> together than the standard 100km rule of thumb that is usually stated 
>> in the
>> context of long distance clusters. This causes performance issues if a 
>> heavy
>> write bogs down the app while it is waiting for writes to occur at both
>> sites before the update is considered complete at either site.
> 
> For most problems, there can be solutions, if the problem is properly 
> analyzed, and one is open to all possible solutions.
> 
> CPUs have been hitting this wall for a while now, either in true speed, 
> or, in the amount of heat they can tolerate.  One of the solutions there 
> is more CPUs, and even switching between them, based upon heat.
> 
> Now, I'm not a network person.  I'm lucky if I can properly plug in an 
> RJ45 plug.  But, I'd ask, if you have more to push through a pipe than 
> the pipe can handle, why not more pipes?
> 
> Someone is going to solve this problem.  Might be nice if it was first 
> implemented on VMS clusters.
> 

Been around for ages.  Back in the dialup era it was common
to use multiple connections to get increased bandwidth.  I
seem to remember someone (USR?) actually having something
called "Shotgun Modems" to do it directly in the hardware.
Multiple Serial ports on one end, multiple phone lines on
the other PPP was capable of aggregating multiple streams.

bill