SHARE Attracts Mainframers to Providence

IBM CEO Ginni Rometty keynoted the 4-day SHARE conference, which opened in Providence, RI, on Monday. DancingDinosaur didn’t arrive until Tuesday—sorry Ginni, missed your speech–but found the event packed with sessions and a lively vendor expo.

10-core z14 processor

z14 10-core chip design

The session on making the business case for Linux on z and IBM Linux One proved interesting. Dancing Dinosaur has covered this topic multiple times. The main point: Linux on the z drives greater cost efficiency when compared with Linux on distributed servers.  Just add up the number of servers and admins you will need to host, say, a few thousand Linux instances.  Other points the speaker, an IBMer named Eduardo Olivera, were security, even without the z14’s pervasive encryption, which is coming to Linux on z, and high availability. It also brings non-disruptive scalability as well as delivering excellent Linux performance on the z. Linux on LinuxOne, he added, costs less since the core that runs Linux costs less than cores running z/OS.

Multiple instances of Linux benefits from a hypervisor as an added layer. On LinuxOne the primary hypervisor is KVM.

Another session was titled IBM MQ and the IBM Integration Bus for Docker Container and Cloud environments and led by Mark Taylor and David Coles. With the advent of the z14 and its new container pricing, containers will become a bigger topic going forward. DancingDinosaur covered container pricing a couple of weeks ago here with more coming in upcoming posts. Suffice for now that there is more in the z world than Docker containers.

Most of the session focused on MQ, which is IBM’s message hub. MQ essentially decouples applications and systems, which opens a range of possibilities. For starters, MQ v9 works with IBM’s Bluemix and has been simplified through the availability of pre-configured default and templates. According to Taylor there is almost no situation in which, with a little creativity, you cannot deploy MQ to advantage. IBM, he suggested, is preparing to offer MQ as a hosted service (MQaaS) through the IBM cloud. In that case, MQ will become considerably easier, especially for mainframe novices.

Rocket hosted an interesting session titled the API Revolution on Z—unlock your mainframe for accelerated digital transformation. In effect, Rocket laid out the business case for IBM z/OS Connect and went on to describe how it augments the IBM product with connections to more obscure products, like Adabas, and different ways data sets are organized, such as physically sequential, indexed sequential, and others.

Driving the need for APIs, RESTful and others, coincides with the fading away of data warehouses. Until recently, organizations created large data warehouses in which they parked data for analytics.  Today, if an organization wants to be competitive, it needs to access the data where it resides in, at the least, near real time and in its native form. Organizations don’t have the time to transform or normalize the data before analyzing it.

Instead, applications need to grab the latest data through an API and analyze it in whatever format it takes—SQL, JSON, NoSQL, SMF, basically anything, on the spot through Spark or something even faster. And Rocket wasn’t talking about data that was read-only; the data has to be fully transaction compliant.

Brocade hosted an interesting session titled Data Center Modernization for Mainframe Environments. DancingDinosaur has sat through many mainframe modernization presentations and they always focus on something different. Sometimes they focus on managing software deployment to minimize peak hourly charges. Other times the modernization strategy focuses on storage. Since this was Brocade’s session mainframe modernization focused on the network and the switches to optimize packet speed.

This is becoming increasingly important as flash storage gets coupled with really fast, highly optimized servers that easily saturate an older network design. Brocade’s latest Gen6 Directors promise to solve the problem, at least until the next generation of even faster flash and server processors arrive. Remember, the new z14 has a 10-core, 5.2 GHz processor that has been optimized at more levels than we can track.

For Brocade modernization comes down to chasing anything that impact latency. To modernize your mainframe environment just identify all the little things that add latency. It’s not just about buying a bigger pipe or faster switch; it might require cleaning dirt from fiber optic cables that add to latency in barely discernible increments.

But maybe the biggest delight for many at SHARE on Tuesday was the lobster and shellfish served during the vendor expo. For over an hour everyone feasted on lobster rolls and other shellfish. Unfortunately, DancingDinosaur is allergic to lobster.

DancingDinosaur is Alan Radding, a veteran information technology analyst, writer, and ghost-writer. Please follow DancingDinosaur on Twitter, @mainframeblog. See more of his IT writing at technologywriter.com and here.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: