2021. What an interesting year. With the world turned upside down by a pandemic that seemingly had its sights set on...
NonStop as a Service
By Justin Simonds, Master Technologist
If you have attended the NonStop Bootcamp in the past few years you have likely heard the term NonStop as a Service (NSaaS). This can mean many things to the person hearing the term. Will this be NonStop in a public cloud? Will this be a NonStop portal that can be accessed from a browser? Is this a desktop NonStop? The imagination can indeed run wild.
I’m sure most lovers of NonStop have a wish that more people could be exposed to NonStop. That is certainly a goal of NSaaS. Further that, if exposed, more software houses might port their software onto NonStop sparking a resurgence of interest in the platform. I think we can all agree that would be a great thing.
As NonStop has taken more and more advantage of our POSIX 1 & 2 compliance, known as OSS on the system, more and more open-source products have been ported to NonStop. Recently we have seen the use of GIT and NSGIT for those wanting to stay on the Guardian side of the house), Jenkins, Python and many other common tools and software running on Linux.
There are claims that new developers being given a BASH shell running in OSS are immediately productive without understanding NonStop. While I agree a programmer could write effective code for the system without necessarily understanding the NonStop Operating System one still needs systems managers, DBAs and administrators that do understand NonStop.
Back in the beginning days of NonStop, when Dr. Jim Gray wrote his famous paper “Why computers fail and what can be done about it” (June 1985 – just Google Jim Gray and why computers fail. It’s all over the place). One of his points back then was that ‘System administration, which includes operator actions, system configuration, and system maintenance was the main source of failures — 42%’ (page 8). With everything that has been added to the system since the 1985 version, things are generally more complex now and good NonStop people are worth their weight in gold if they can keep the system running.
Of course there are many more management tools and more automation but I contend there is still much more complexity. Our single system image, parallel architecture is a system designed for ease of management and generally fewer people are needed to manage the system but fewer required doesn’t mean the value is diminished. The cost of an outage always dwarfs the cost of preventing it.
So let’s say ‘what if’ a NonStop system could be made available with ‘some’ management? Perhaps for a reasonable price general system management, that is, an ability to remotely monitor a system for CPU busy, disk file usage, memory utilization, queues all the general day to day stuff. What if that were done for a customer allowing them to concentrate their staff on applications and modernization of the environment? Would that be an interesting service? Maybe NonStop basic management as a service?
This would free existing personnel to explore some of the new software and perhaps play around with DevOps on NonStop as basic system management was being performed in the background.
Although not NSaaS in full, it might be a step in the right direction.
This step could take the form of remote management as a service or perhaps as the HPE Greenlake model. In Greenlake, HPE will ‘build a private cloud’ in a customer’s data center and provide basic system management. A customer is charged for consumption, based upon agreed models for an agreed upon amount of time. Growth generally has been prebuilt into Greenlake systems so a customer can ramp up and also ramp down processing and being charged for what they use not by the amount of equipment.
NonStop is participating in the Greenlake model but based on the way customers use NonStop we are in phase one, a full commit model. Whereas you might acquire a Greenlake system from HPE and spin up Oracle and SQL Server databases for development or even spin up a large Kafka Spark environment to experiment with, generally NonStop systems are started up and don’t change much other than possible changes to the transaction rates.
If you’re a large bank you don’t want NonStops being spun up and down. You want NonStop to be stable and “always-on.” In later phases we will be offering more options but initially, under Greenlake a NonStop system will always be on. However there is the possibility I spoke of about providing some basic system management freeing up resources to see what else can be done on the NonStop or to modernize the environment.
For a more in-depth discussion on this or other areas of interest reach out to your NonStop sales account manager or your NonStop Solution Architect. They are always happy to talk NonStop!