[Info-vax] Distributed Applications, Hashgraph, Automation

[Stephen Hoffman]

On 2018-02-17 02:06:37 +0000, Kerry Main said:

>> 
>> -----Original Message-----
>> From: Info-vax [mailto:info-vax-bounces at rbnsn.com] On Behalf Of
>> Stephen Hoffman via Info-vax
>> Sent: February 16, 2018 5:22 PM
>> To: info-vax at rbnsn.com
>> Cc: Stephen Hoffman <seaohveh at hoffmanlabs.invalid>
>> Subject: Re: [Info-vax] Distributed Applications, Hashgraph, Automation
>> ...
>> In terms of features and capabilities provided, RDMA is a 
>> next-generation cluster interconnect and not a next-generation cluster.
> 
> Which is what I stated. " this type of cluster communications capability"

Adding a new-generation memory channel (RDMA) interconnect isn't going 
to change the market perception of OpenVMS clustering in any 
appreciable nor meaningful fashion.

What you were discussing was the past, and with some incremental 
changes to the present, with the potential addition of RDMA.  Which 
hopefully also includes 40 GbE and some related updates.   Not about 
hauling the whole environment forward. Which was what started this 
thread, and what I was referencing.  Where OpenVMS is now has clearly 
not convinced a whole lot of folks to purchase OpenVMS and particularly 
to adopt clustering as implemented.

There's more than a little work deprecating and replacing the worst of 
the parts of OpenVMS while preserving the best and most of the rest.  
Rethinking cluster and app configuration and control for instance.  
Integrating IP, LDAP, SMB and other ubiquitous services.  Updating the 
DLM.  Scheduling.  Etc.   Clusters as implemented still have a couple 
of really good features, too.  Logical names — cluster or otherwise — 
as configuration tools are among the worst of ideas found on OpenVMS, 
and I could see replacing the whole of the V4-era design with an 
LDAP-based design even for device I/O redirection, and with an app 
configuration tools and API based on YAML or otherwise.

Integrating a distributed ledger as an operating system component — 
which is what started off this thread — I'm not so sure about.  
Certainly distributed ledgers are very useful for a specific apps and 
environments, and there's certainly ample fodder for marketing, but 
there's not a whole lot of (no pun intended) consensus around which 
distributed ledger schemes and how that's going to work, and there's 
certainly a concern that issues arising with cryptocurrencies such as 
fraud or theft could undermine the perception of distributed ledgers as 
marketing fodder.




-- 
Pure Personal Opinion | HoffmanLabs LLC