[Info-vax] The real problem that needs solving to grow VMS

[kemain.nospam at gmail.com]

> -----Original Message-----
> From: Info-vax <info-vax-bounces at rbnsn.com> On Behalf Of Arne Vajhøj
> via Info-vax
> Sent: Friday, November 11, 2022 10:13 PM
> To: info-vax at rbnsn.com
> Cc: Arne Vajhøj <arne at vajhoej.dk>
> Subject: Re: [Info-vax] The real problem that needs solving to grow VMS
> 
> On 11/11/2022 7:21 PM, kemain.nospam at gmail.com wrote:
> >> From: Info-vax <info-vax-bounces at rbnsn.com> On Behalf Of Simon
> >> Clubley via Info-vax
> >> Sent: Friday, November 11, 2022 2:19 PM On 2022-11-11,
> >> <kemain.nospam at gmail.com> <kemain.nospam at gmail.com> wrote:
> >>> How about an application environment whereby the nature of the
> >>> application environment required a well established shared disk
> >>> strategy rather than a shared nothing strategy?
> >>>
> >>> Reference: Pro's and Con's of shared disk vs. shared nothing
> >>> <http://www.benstopford.com/2009/11/24/understanding-the-shared-
> noth
> >>> ing-architecture/>
> >>
> >> Some reading for you Kerry:
> >>
> >> https://en.wikipedia.org/wiki/GFS2
> >
> > If you had read one of my recent posts regarding this topic, you would
> > have seen where I stated "shared disk (OpenVMS, Linux GFS2 and z/OS)"
> 
> Yep.
> 
> > But I stated well established .. this fully supported GFS2 technology
> > was only integrated into RHEL since V5.3. (2010 timeframe).
> 
> That (2010 timeframe) is pretty new in the VMS world (8.4 was released in
> 2010 I believe).
> 

Integrated OpenVMS Clustering was introduced back in OpenVMS V4.0 - back in
the 1983-84 timeframe. Hence, OpenVMS clustering has been battle tested and
refined in mission critical environments for approx. 40 years now.

> But that (release was January 20th 2009) is actually pretty well
established in
> the Linux world. RHEL is at 9.0 / 8.7 today. And 5.x is long out of
support.
> 
> >                                              Also, GFS2 is not
> > currently widely used in the Linux world.
> 
> True.
> 
> But that is because the IT world today use databases with their own sync
> mechanisms not file system with an OS provided DLM.
> 
> The demand is not there.
> 

Again - both shared nothing and shared disk cluster (including multi-site)
strategies have pro's and con's, but the data sync'ing across distributed
shared nothing systems while at the same time addressing typical CAP
(consistency, availability, partitioned data - pick two) is not an easy
thing to support in real life.

As a reference:
<http://highscalability.com/blog/2015/10/12/making-the-case-for-building-sca
lable-stateful-services-in-t.html>
Extract - "A stateless architecture is easy to scale horizontally and only
requires simple round-robin load balancing. What’s not to love? Perhaps the
increased latency from the roundtrips to the database. Or maybe the
complexity of the caching layer required to hide database latency problems.
Or even the troublesome consistency issues."

See rest of article.

Regards,

Kerry Main
Kerry dot main at starkgaming dot com





-- 
This email has been checked for viruses by AVG antivirus software.
www.avg.com