[Info-vax] Time to turn DECUServe into a mixed VMScluster?

[Arne Vajhøj]

On 6/1/2023 9:53 AM, Dave Froble wrote:
> On 6/1/2023 12:44 AM, terry-... at glaver.org wrote:
>> On Wednesday, May 31, 2023 at 6:33:17 PM UTC-4, Dave Froble wrote:
>>> If my feeble memory still works, I think that has already been answered.
>>>
>>> x86 cluster communications will be modified, and to include Alpha 
>>> and/or itanic,
>>> a patch for either will be required. No patches for VAX, since VSI 
>>> does not
>>> have a VAX release to which patches can be provided.
>>
>> As a reminder, the VSI FAQ still says "Clustering of VSI OpenVMS for 
>> x86-64 with VAX/VMS systems is still under investigation at this 
>> time." See: https://vmssoftware.com/about/v9-qa/
>>
> 
> Yes, but if VSI decides to "improve" the cluster communications, and 
> there are reasons to do so, then I see two possibilities for clustering 
> with a VAX.
> 
> 1) Issue a patch for VAX/VMS, which my understanding is they cannot do.
> 
> 2) Include both any new protocol along with the current protocol on 
> Alpha, itanic, and x86.  And that would sort of wipe out any security 
> measures that might be in the new protocol.

I am not sure that I understand the problem.

We are talking about a PR stunt here - running a VMS cluster with
4 architecture to get some good press.

Nobody cares whether that system uses encrypted clustering
traffic or not.

Some will care about encrypted clustering traffic in a production
environment.

But such an environment is not likely to want such a 4
architecture cluster or to include VAX (it seems reasonable
to assume that a VMS VAX system today is "frozen" aka don't
want to change anything).

If the new VMS cluster software negotiate encryption and only
enable if both ends support it then sites could patch and
get encryption for Alpha - x86-64 and Itanium - x86-64
if they want to (or auditors force them to).

Arne