[Info-vax] Distributed Applications, Hashgraph, Automation
Kerry Main
kemain.nospam at gmail.com
Fri Feb 16 21:06:37 EST 2018
> -----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
>
> On 2018-02-16 03:15:48 +0000, Kerry Main said:
>
>
> > mission critical proven DLM,
>
> Handy, definitely. In competitive configurations, other DLMs exist.
>
Just not as proven in mission critical environments. Experience and
reputation does matter for components as critical as the DLM.
Yes, z/OS has a well respected mission critical DLM as well.
> > cluster logicals,
>
> Or as is commonly used on various other platforms, LDAP.
> https://directory.apache.org/apacheds/ or VSI Enterprise Directory, or
> otherwise.
>
> > cluster batch,
>
> Batch is not competitive as scheduling offerings goes. It's a pain in
> the rump, in practice. Third-party scheduling offerings for OpenVMS,
> and other configurations on other platforms have vastly better
> offerings. Hadoop YARN or Mesos, etc.
>
Most companies look to schedulers as add-on LP's - not as core OS
offerings.
The reason is that they want the same scheduler to run on all their
production platforms.
Heck, no one uses the native Windows Server batch service.
> > common file system (new one with significant new features cooking as
> well).
>
> The new file system is comparatively old, unfortunately. We're
> clearly headed toward in-memory processing and byte-addressable
> non-volatile storage too, and not toward main processing using HDD or
> SSD storage, save as archival and recovery and overflow.
>
Main memory is heading towards TB. Multiple disks (SSD and HDD) are
heading towards PB.
Apples and Oranges. There will continue to be a place for both.
> > However, the industry really only looks at shared disk or shared
> nothing.
> >
> > Btw, the modern day equivalent to memory channel and ultra low
> latency
> > data sharing is either Infiniband or RoCEv2 (RDMA over converged
> > ethernet)
> >
> > Not sure where it is at right now, but RoCEv2 is on the research
slide
> > of the OpenVMS roadmap.
> >
> > Imho, this type of cluster communications capability is critical to
> > next generation cluster scalability of shared disk clusters. It is
how
> > VSI can address the biggest counter argument to shared disk clusters
-
> > "shared disk clusters have scalability issues due to the requirement
of
> > a distributed lock manager"
>
> 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"
Regards,
Kerry Main
Kerry dot main at starkgaming dot com
More information about the Info-vax
mailing list