Saving the Ferrous Tiger

By : |March 14, 2016 0
Reviving mainframes with Linux can in all possibility sound as just another bubble-wrap. But what if it’s a needle that is injecting a lubricant that will keep the iron from rusting? A big ‘what-if’ indeed

Pratima H

SAN FRANCISCO, USA: Some may call it bricolage, some may label it as the ultimate IKEA-isation of computing wood, some may tag it as recycling and some may stamp this a crusade for conserving an old species.

But whichever way you look at it, some huddles look too fascinating and timely to be ignored.

___________________________________________________________________________________________________________

After all, it’s about going inside a really old lair and bringing the jungle king out alive.

From 701 to the z13, this animal has seen it all – the thick forestlands of transactional processing; those unfamiliar forest fires kindled by mobility, cloud and analytics; and a computing gentrification of its own kind right inside its neighbourhood.

First came PCs, then Mobile, Cloud, Digital Forces, SaaS and so on. And soon enough the unshakable giant powering big computing backyards appeared to be conking off in its fight with new kids who were moving fast from being mischievous urchins into sharp adolescents.

One could brandish gilded swords in this battle being fought against push-buttons or one could give it all up and take to the rocking chair.

But wait, one could also join these young adversaries in the big, well-sprawled-out park.

Like a community project or collaborations where problems can be fixed and the so-called jalopy be well-oiled from time to time for new roads. Pitstops and huddles soon surfaced.

Like the Open Mainframe Project. In this huddle you can spot names like IBM, CA Technologies etc and Canonical, Ubuntu at the same time.

Such interventions sound like the perfect recipe to re-invent mainframes from the sunset phase (most people had written them off). They mix development muscle of open source protein with lower costs and complexity that is usually associated with mainframes while allowing enterprises to continue using this iron for heavy, intensive workloads the way they did just a few years back.

Linux on system z, LinuxONE (The mainframe for open source OS), Ubuntu support etc are just a few glimpses of how Linux wheels are being planted behind this heritage chariot today.

The Open Mainframe Project echoes the progress as it attempts to identify ways to leverage new software and tools in the Linux environment that are ideal for taking advantage of the mainframe’s speed, security, scalability and availability and intends to coordinate mainframe improvements to upstream projects to increase the quality of these code submissions and ease upstream collaboration.

Recently the Open Mainframe Project showed fresh momentum around new membership investments, an internship program and technical focus areas like JIT for OpenJDK, Docker support, Blockchain and Linux monitoring tools along with new members like DataKinetics, East Carolina University, Hitatchi Data Systems and Sine Nomine Associate. The Linux Foundation, the nonprofit organization enabling mass innovation through open source, also announced an expansion of its Open Mainframe Project to accelerate further collaboration among industry and academic partners and create areas of focus for advancing Linux on the mainframe for the benefit of the community and industry.

Linux-development bottled nicely with the power and scale of mainframes and canned tightly with an open-source- community spin. Now that is indeed a case of two different worlds coming together.

Is it driving mainframe usage? Is it bringing beauty to the beast in a new form? Is it that proverbial twist in the plot of ‘mainframe is still relevant’ saga?

We try getting such doubts sorted in this quick Q&A with John Mertic, Senior Program Manager, The Linux Foundation. Mertic serves as the senior project manager at The Linux Foundation, running the day-to-day operations for projects such as the Open Mainframe Project and ODPi. He pours in all his hands-on incision, insight and intimacy with Open Source as he explains how and where the movement is progressing for mainframes.

And yes, we do talk about white-boxes, Dockers, Blockchain and age through it all.

What are the progress highlights of the project so far? What role have academic contributors played here? How do you respond to perceptions that corporate giants only stamp a lip-service for open community projects?

The announcement we did at IBM Interconnect hit some key highlights of the project, namely:

– Continued member growth across categories with four recent investments from companies becoming members of the organization
– Formation of the technical steering committee with direction on focus areas
– Launch of an internship program for Summer 2016

All of these milestones are important for advancing the project. I’d like to highlight the internship program, in particular, though. It will both help build lucrative careers for students, as well as contribute to the open mainframe talent pool. This is critical for advancing the state of the art of mainframe technologies for the future.

The Open Mainframe Project is an open source project and anyone can contribute and influence the direction of the work. It’s also worth noting that the board chair of the project is Len Santalucia, representing Vicom Infinity which is a Silver member. The sense of respect all parties have for one another is what makes open source projects like this successful.

Would the idea of re-inventing mainframes sound relevant and plausible today? Specially with areas like some degree of obsolescence, challenges on skill-sets, enterprise-scale Linux, production-ready environments, and other factors? How well does mainframe-on-cloud align as a trend?

It’s funny, because when I tell people I work with the mainframe community to help it advance and grow, the perception you bring up is certainly one I encounter. But when you dig into the pure economics of it ( a small z13 system can often replace a large cluster of x86 servers ) and recognize the trends of Big Data and high frequency transactions that many applications demand, it becomes obvious to the unique value the mainframe has.

Certainly there is a lot of work to do, and the areas you outlined are many of those. Overall outside of the technical work this group will do, there is a huge topic of education and promoting success stories of mainframe usage. The use-cases for highly available, fault tolerant computing are going to grow, and having a common OS like Linux at the core will help reduce the barriers to considering this platform.

Is the white-boxing wave likely to stay and disrupt the industry enough? Do commodity hardware, open compute project trends propel mainframes well?

Throughout the history of this industry, there is always the hunt for the magic bullet to solving technical challenges, whether that is at the software, hardware or infrastructure levels. Having more choices enables end users to pick the right tool for the problem at hand. The reality is that Linux is a unifying OS that removes the barriers to adoption and deployments.

Any observations on new pieces like Dockers, hypervisors etc? How significant is Blockchain Hyperledger to mainframe thrust?

Docker and containerization has led the drive to micro-services architecture and in many cases improved how large scale applications are deployed and managed. This idea of splitting larger applications into smaller services fits perfect with the mainframe architecture, adding with it fault tolerance and an easier ability to make workloads more elastic. Hyperledger is another ideal workload for the mainframe, with the bonus that this architecture is a known quantity by the financial world.

How much do the likes of IBM matter for the project, specially with intersections or equations between Softlayer and mainframes or LinuxONE and Z portfolio?

IBM’s role in the project is an important one but no more important than every individual and company contribution. The neutral Open Mainframe Project allows anyone to innovate and build for the future.

Any challenges being experienced so far?

Getting the word out; with a project this young, it’s still not on the radar for many mainframe users and vendors.

Participating at IBM Interconnect in February and SHARE San Antonio in March were two big steps to getting that awareness going. Open source community projects like this naturally take time to form and mature, and this one is no different. The good news is that this community is already well connected and aligned on the common goal of reducing objections to Linux on the mainframe, which lets us as a project focus more on how we accomplish these goals rather than debating on why the community exists.

No Comments so fars

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.