Tuesday, March 15, 2016

Potential exposure for undetected loss of data for z/OS users writing data to zEDC compressed data sets using QSAM

Here's a new Red Alert. I'm taking over the content.

Potential exposure for undetected loss of data for z/OS users writing data to zEDC compressed data sets using QSAM on releases z/OS 1.13, 2.1 & 2.2.

Data loss may occur when writing to a zEDC compressed data set using QSAM when CLOSE is issued and there exists a partial QSAM buffer that has not yet been written, and all allocated space in the data set is filled. In this case, the unwritten partial buffer may not be written after CLOSE processing obtains the new extent for the data set.

The problem does not apply to BSAM or other access methods.

Please see APAR OA50061 for additional information.

Recommended Actions 
++APAR for OA50061 should be applied for all environments with zEDC compressed data sets. This requires an IPL to activate.

You can find the APAR information over here where you can find some more info :
"During creation of a new extent during CLOSE processing, any
user blocks that have not been compressed and written out to the
zedc compressed dataset yet will not be written out.

This does not apply to BSAM processing since all outstanding
WRITE requests must be CHECK'd for completion before CLOSE is
For the moment it also says : "++APAR will be available soon".

If you haven't signed up to the Red Alerts by now, you really should do it. Just go over here.

Thursday, February 25, 2016

New Cobol Compiler V6.1

At the same time of the announcement of the z13s, IBM also announced a new version of their Cobol compiler : 'IBM Enterprise COBOL for z/OS, V6.1 delivers new and enhanced COBOL statements for added function and flexibility (ZP16-0093)'.

Here's an overview from the announcement :
"IBM® Enterprise COBOL for z/OS®, V6.1 exploits the capabilities of IBM z/Architecture® and continues to add the following new features and enhancements:
  • The compiler internals are enhanced so that you can compile your large programs.
  • New and changed COBOL statements are included to provide you with additional functions.
  • New and changed compiler options are implemented to provide you with added flexibility. 
  • Support for generation of JSON texts
  • The ARCH compiler option allows you to exploit and tune your code to run on your choice of z/Architecture levels of the z/OS platform.
  • The OPTIMIZE compiler option allows you to select from multiple levels of increasing optimization for your code.
  • Support for the latest middleware includes IBM CICS®, IBM DB2®, and IBM IMS™."
Enterprise COBOL for z/OS V6.1 requires z/OS V2.1 (5650-ZOS), or later and is available on March 18, 2016. Program number is 5655-EC6. It runs on z9 systems and higher.

Like with previous versions, there is also a trial version announced (ZP16-0094).
"IBM® Enterprise COBOL Developer Trial for z/OS®, V6.1, an evaluation edition of Enterprise COBOL for z/OS, is a no-charge offering and does not initiate a Single Version Charging (SVC) period.
Enterprise COBOL Developer Trial lets you assess the value that could be gained from migrating to Enterprise COBOL for z/OS, V6.1, before making a formal decision to upgrade. This trial enables the evaluation of the latest IBM COBOL capabilities, in a nonproduction environment, without the prerequisite time and resource commitments that are required for a full-production migration project. The trial period is 90 days."
Next to this, there's also a Value Unit Edition (ZP16-0095).
"IBM® Enterprise COBOL Value Unit Edition, V6.1 delivers a compiler with a one-time charge price metric that is based on Value Units. The Value Unit price metric supports both full-capacity and sub-capacity environments."
In other words, this is the IPLA (or OTC) version of Cobol V6.1 with a program number for the product (5697-V61) and one for the Subscription&Support (5697-ECS).

Well, that's about it for the new Cobol Compiler. Check out the COBOL for z/OS documentation library for more information. I'm sure the 6.1 documentation will be soon available. If you need even more info you can also have a look at this Enterprise Cobol for z/OS Resources Page.

Tuesday, February 23, 2016

Preview of IBM z/VM V6.4

On February 16, 2016 along with the z13s, IBM also made a preview announcement of z/VM 6.4 : 'Delivering industry-proven advanced virtualization capabilities to support the increasing demands of your business (ZP16-0014)'

z/VM is planned to become available in the fourth quarter of 2016 and it will support the two LinuxONE versions and System z from z114 and z196 onwards. Mind you, z/VM 6.3 was still supporting the z10 systems, z/VM 6.4 will NOT !

Here's a short overview directly from the announcement :
"z/VM V6.4 enables extreme scalability, security, and efficiency, creating cost savings opportunities, and provides the foundation for cognitive computing on z Systems and LinuxONE.  z/VM V6.4 is planned to deliver:
  • Increased efficiency with HyperPAV paging that takes advantage of DS8000 features to increase the bandwidth for paging and allow for more efficient memory management of over-committed workloads.
  • Easier migration with enhanced upgrade-in-place infrastructure that provides an improved migration path from previous z/VM releases.
  • Improved operations with ease of use enhancements requested by clients, such as querying service applied to the running hypervisor and providing environment variables to allow programming automation based on systems characteristics and client settings.
  • Improved Small Computer System Interface (SCSI) support for guest attachment of disk and other peripherals, and hypervisor attachment of disk drives to z Systems and LinuxONE systems to:
    • Increase efficiency by allowing an IBM FlashSystem® to be directly attached to z/VM for system use without the need for an IBM System Storage® SAN Volume Controller (SVC).
    • Enable ease of use with enhanced management for SCSI devices to provide information needed about device configuration characteristics.
  • Increased scalability by exploiting Guest Enhanced DAT to allow guests to take advantage of large (1 MB) pages, decreasing the memory and overhead required to perform address translation.
  • Integration of new CMS Pipelines functionality which previously was not formally incorporated within the z/VM product, allowing a much more inclusive set of tools for application developers. "
Next to that, there's also support for simultaneous multithreading (SMT) and the vector extension facility (SIMD) instructions on the z13 systems. The last one helps "accelerating Business Analytics workloads on z13, z13s, and LinuxONE".

Have a look at the z/VM page and the z/VM 6.4 Resources page. I'm sure, in time, more information will certainly become available on those pages. And it's always interesting to follow the discussions on the z/VM discussion list.

Friday, February 19, 2016

Hardware withdrawal of zEC12, zBC12 and zBX Model 004

It’s become customary by now, the announcement of one system is accompanied by the End of Marketing announcement of the previous generation, in this case the zEC12 and the zBC12 : 'Hardware Withdrawal: IBM zEnterprise EC12, IBM zEnterprise BC12, and IBM z BladeCenter Extension (zBX) Model 004 - some replacements available (ZG16-0021)'

However, it’s a bit more complicated this time because what used to be a two step operation is now somewhat of a three step generation. Remember with the z10 EC and z10 BC or with the z196 and z114.
First there was the End of Marketing of the system itself and all hardware MES upgrades like disruptively adding memory, adding books or I/O cards.
And secondly, a year later, all LIC upgrades like e.g. adding a processor within the same book or drawer, non-disruptively adding memory was ended as well.
This is again the case with the zEC12 en zBC12, except that six months before the first step all delivery of and upgrades to a z12 are no longer possible “for the listed RoHS Jurisdictions”. This includes, I think, most of the European countries. So, this is how the withdrawals go

June 30, 2016 : RoHS jurisdictions
  • All Models of the zEC12 and upgrades towards them
  • All models of the zBC12 and upgrades towards them
December 31, 2016 : all countries
  • All Models of the zEC12 and upgrades towards them
  • All models of the zBC12 and upgrades towards them
  • Model conversions and hardware MES features applied to an existing zEC12 or zBC12 server
December 31, 2017
  • Field installed features and all associated conversions that are delivered solely through a modification to the machine's Licensed Internal Code (LIC)

You know I mentioned before that I thought the End of Marketing for previous generations was quite tight from a customer point of view. Here’s a brief overview.

Click on image for larger version

I also mentioned the zBX 004. Well, as of March 31, 2017 the last zBX Model 004 will have been marketed and no more upgrades from previous models will be possible either. I guess no further comment is needed with regards to the zBX ?

Thursday, February 18, 2016

Two Flash Alerts for DS8870

Well, life goes on, and this week I saw two flash alerts passing by on DS8870. You can find them here and here. They're not that long, so I'm just taking over both of their contents.

DS8870 Global Mirror suspend caused by a Track Format Descriptor mismatch

Global Mirror suspends caused by a microcode logic error introduced in R7.4 that results in a Track Format Descriptor mismatch. 

Global Mirror suspends caused by a microcode logic error introduced in R7.4 that results in a Track Format Descriptor mismatch. Microcode is improperly setting a flag in a PPRC control block. This problem is pervasive in Global Mirror environments. R7.4 code levels below and R7.5 levels below are exposed to this issue.
A mandatory ECA 712 is being released to the field, we recommend upgrade to code bundle R7.5

High Performance Flash Enclosure drive failures can cause loss of access

IBM released microcode with improved error handling for DS8870 High Performance Flash Enclosure (HPFE) flash drive errors. 

IBM has developed microcode enhancements for error handling in HPFE. DS8870 (R7.5 SP2.3 ) microcode levels contain the changes to improve the reliability and availability of High Performance Flash Enclosures. This enhancement streamlines error detection and isolation when a failing Flash Drive exhibits excessive errors.

This change is designed to be concurrently installable on DS8870 presently running the R7.x families of microcode.

A mandatory ECA 737 is being released to the field, we recommend upgrade to code bundle R7.5 SP2.3 as soon as possible.

ID: 312723 312369 313417

Tuesday, February 16, 2016

z13s - a new generation of z Systems built for cloud and mobile

General Introduction

A couple of weeks ago IBM announced its new version of the LinuxONE and especially of the smaller version, known as the Rockhopper. As a matter of fact, any one who follows the mainframe market a little bit could guess it was based on the new little brother of the z13. And it was of course. So today IBM announced the new z13s. Allthough LinuxONE was announced via Press Channels, it is now coupled with the z13s in the official announcement : "Expanding the IBM Systems' portfolio with additions to IBM z Systems and IBM LinuxONE (ZG16-0002)". I'm going to give you an introduction to the z13s, but I will also refer from time to time to the LinuxONE Rockhopper of course.

You know I largely concentrate on the technical aspects of the announcement rather than focusing on the strategic importance of the platform. Lots of other sources will give you plenty of information about this. Take e.g. a look at this IBM page covering z13s and its relation with Cloud, Security, Analytics and DevOps. And you can find quite some videos on Youtube as well.

I must immdiately say, I don't like the naming of this new system : z13s. This was my plural for two z13s, oops, I guess that will be two z13's and two z13s's. Any one a better suggestion ?

But let's get to the content of the announcement. I'll give you the usual survey of the new system starting with some technical specifications. Of course the Software Pricing is also always an interesting part. I'll conclude with some key dates and a couple of references to already available documentation. Here's an overview of the z13s.

Click on image for larger version

As usual, along with the announcement of the new 'Business Class' model there's also the GA2 for the 'Enterprise Class' model, the z13. No separate announcement as before though, it's all in the z13s announcement.

Technical specifications

The z13s has machine type 2965. Just like its predecessors it's also a one frame, air-cooled machine. Just like the z114 and the zBC12 it has top-exit I/O cabling and top exit power cabling. Unlike the z13, the processor is, with its 4.3 GHz, just a bit faster than its predecessor, the zBC12 (4.2 GHz). Of course it also has the new hardware features that were introduced with the z13 like Simultaneous Multi Threading (SMT) and Single Instruction Multiple Data (SIMD). Look for my post of the announcement of the z13 for more details about these. SMT is made available for IFLs and zIIPs, just like on the z13.

Again, we have two models : the N10 and the N20. The N10 has one drawer (13 PUs) and the N20 can have one or two drawers for its 26 PUs. As you can guess the N10 has 10 configurable engines, the N20 has, indeed, 20 of them. The N10 has 2 SAPs and one IFP (Integrated Firmware Processor) resulting in 10 configurable engines The N20 has an IFP, 3 SAPs and 2 spare engines. But what's different from the zBC12 is that the z13s N20 can be a one drawer or a two drawer model. The second drawer is added when more than 2TB of memory is needed. As you can see it has 52 PUs, but only 20, as with the 1-drawer model, are available to the customer. The second drawer is only there for the extra memory from 2TB up to 4TB. Later on I'll tell you more about the memory.

If you go through the announcement, you will notice there are two other models as well : the L10 and the L20. Well, that's the indication for the Rockhopper and it's the same story with the drawers as well as for the configurable engines. Except, you can of course only have IFLs here.

As I already indicated, both models (N10, N20) can have up to 6 traditional CPs ranging from A01 to Z06 giving us 156 capacity settings. The smallest model, the A01 has once again a higher capacity than its predecessor. On the z114 it was 25 mips, on the zBC12 it was 50 mips and now it's become 80 mips. Same story as with the zBC12, allthough it now has 10 MSUs, it still has zELC pricing and the software price for the A01 remains the same as for the zBC12 A01 with its 6 MSUs.

Here's an overview of how the z13s compares to its predecessors.

You know by now what PCI stands for, of course. so a uniprocessor goes up from 1064 PCI on a zBC12 to 1430 PCI on a z13s.

Memory - the first 64GB are for free

I rememeber when I wrote about the announcement of the z13, one of the subtitles was Memory Memory Memory. And I can say, you're in for the same treat on the z13s. The maximum memory is 8 times larger than what we had on the zBC12 resulting in 4TB. Model N10 will be limited to around 1TB. But what's perhaps even more interesting is the minimum capacity you get. The first 64GB are free, you can't start any lower. And as bizarre as it may sound, tripling your memory will cost you less than e.g. adding just some capacity. Let me give you an example. Say you have 48GB on a zBC12. If you ask no additional memory, you'll always start at 64 on the z13s. If you want to double your capacity to 96GB, you're actually better off with 152GB than with 120GB. Here's an overview for the one drawer N20 configurations.

If you're a bit confused about the strange numbers, it's because the capacity needed for HSA and for RAIM is not taken into account in the Client GB. Upgrades within one line are non-dsiruptive, upgrades to a following line are disruptive.

More memory really makes a difference nowadays. On the one hand with 'traditional' workloads like e.g. DB2 but on the other hand definitely with Linux workloads. There's a specific redbook covering this matter : 'Benefits of Configuring More Memory in the IBM z/OS Software Stack'.

Under the covers

Here you see an 'under the covers' illustration of a z13s one drawer N20 model.

As you can see this resembles the zBC12. With new z13s systems you can only have PCIe I/O drawers but you can still carry forward one traditional I/O drawer but then only for FICON Express8 cards. As with the zBC12, empty slots in this I/O drawer cannot be filled after or during the upgrade to the z13s.
As with the z13, you can now have a rack mounted 1U HMC, which is installed in a customer supplied rack. Also similar is the placement of the Support Elements at the top of the frame.

Upgrades are possible from any z114 or zBC12. Upgrades from an N10 to an N20 and from a 1 drawer N20 to a two drawer N20 are disruptive. Only the N20 allows an upgrade to the larger z13 N30.


Here's an overview of the current connectivity features.

As you can see this is all pretty straightforward. There are some improvements on FICON and OSA-Express functions, but do have a look at the announcement for all the details.

New features and z13 GA2

Now we'll have a look at a couple of new things and I can immediately say that these become also available on the z13 (GA2) now. So, an overview of z13 GA2 might be in place here

I'll pick out some of them. I will come back to other features in future posts like LPAR Group Absolute Capping, SMC-D . . .

zACI or z Appliance Container Infrastructure

zACI is a new framework to support new types of virtual software appliances. The goal of this new framework is to simplify the way IBM will install software 'appliances'. Here's the entire definition
"z Appliance Container Infrastructure (zACI) is a new partition type which, along with an appliance installer, enables the secure deployment of software and firmware appliances. zACI will shorten the deployment and implementation of firmware solutions or software solutions delivered as virtual software appliances. The zACI framework enforces a common set of standards and behaviors, and a new zACI partition mode for a virtual appliance -- requiring a new zACI LPAR type."
If that doesn't make sense to you, it kind of reminds me of the concept of containers such as used with Docker. You can read a good introduction on Docker by one of my colleages over here.
His definition of Docker is pretty similar to what IBM seems to be targetting to accomplish with zACI. Well, at least, that's what I think about it for the moment.
"Docker containers wrap up a piece of software in a complete filesystem that contains everything it needs to run: code, runtime, system tools, system libraries – anything you can install on a server. This guarantees that it will always run the same, regardless of the environment it is running in."
I think that an example will make this immediately a lot clearer. One of those virtual software appliances is IBM zAware. IBM zAware is a well-defined virtual appliance that can be installed in a separate LPAR. The zAware partition mode is replaced by the zACI LPAR mode.

One of those appliances that is planned is z/VSE Network Appliance that will also be installed via zACI. It will avoid using a TCP/IP stack of z/VSE and it's a way to directly communicate from a z/VSE LPAR to an external network just using this appliance. or as the announcement states it :
"The z/VSE Network Appliance builds on the z/VSE Linux Fast Path (LFP) function and provides TCP/IP network access without requiring a TCP/IP stack in z/VSE. The appliance utilizes the new z Appliance Container Infrastructure (zACI) introduced on z13 and z13s servers. Compared to a TCP/IP stack in z/VSE, this can support higher TCP/IP traffic throughput while reducing the processing resource consumption in z/VSE."
The first examples of zACI are announced today and more will follow later on.

Dynamic Partition Manager (DPM) for Linux and KVM

This is a new administrative mode for creating partitions for customers that are running Linux and KVM only. It's a way of helping out people who are not familiar with the traditional ways of creating, managing and monitoring LPARs. In other words, this mainly targets new LinuxONE customers that have no mainframe background whatsoever. This means that instead of IMLing a CPC in traditional PR/SM mode customers will be using Dynamic Partition Manager (DPM) Mode. Therefore it only supports Linux partitions. z/OS, z/VM, z/VSE and the likes are not supported. And, it supports FCP storage only. Or as the announcement puts it :
"The new mode, DPM, provides simplified, consumable, and enhanced partition lifecycle and dynamic I/O management capabilities via the Hardware Management Console (HMC) to:
  • Create and provision an environment -- Creation of new partitions, assignment of processors and memory, and configuration of I/O adapters (network, FCP storage, crypto, and accelerators)
  • Manage the environment -- Modification of system resources without disrupting running workloads
  • Monitor and troubleshoot the environment -- Source identification of system failures, conditions, states, or events that may lead to workload degradation"

Software pricing

My story is pretty straightforward here. Similar to what happened with the zBC12 we don't see a new pricing mechanism but, again, a reduction to the AEWLC pricing. This means a reduction for standalone MLC pricing as indicated in this accompanying announcement : 'Technology Transition Offerings for the IBM z13s offer price-performance advantages (ZP16-0012)'. As the title of the announcement states, there are also transition offerings for migrations in a sysplex environment.

Mind you, however, when looking at those reductions : they talk about a reduction of 13% up to 30 MSUs, 10% up to 45 MSUs and 9% for higher MSUs. Keep in mind that this is the reduction as related to the original AEWLC pricing on the z114. As compared to the zBC12, you will have a reduction of approximately 5%.

This reduction does not apply to IPLA (OTC) softwares.

As I already mentioned the A01 has 4 extra MSUs but remains on zELC pricing. This means the price remains the same as on previous A01 models.

Operating Systems

  • z/OS V2.2 with PTFs
  • z/OS V2.1 with PTFs (Exploitation)
  • z/OS V1.13 with PTFs (Limited Exploitation) 
  • z/OS V1.12 with PTFs : be aware that TSS Service Extension will be required
More info can be found in the z/OS support for the z13 and z13s chapter of the announcement.

  • z/VM V6.4 (preview, available 4Q2016)
  • z/VM V6.3 with PTFs  - Exploitation support
  • z/VM V6.2 with PTFs – Compatibility plus Crypto Express5S support
  • z/VM V5.4 - Allthough still supported NOT compatible with z13s
  • z/VSE 6.1 with PTFs
  • z/VSE V5.2 with PTFs – Compatibility + Crypto Express5S
  • z/VSE V5.1 with PTFs – Compatibility (EOS 30/6/2016)
  • SLES 11 SP3+ and SLES 12 SP1+
  • Red Hat RHEL 6.6+ and RHEL 7.1+
  • Canonical Ubuntu 16.04 LTS – GA 2Q2016
  • KVM for IBM z V1.1.1 – GA 18/3/2016
  • KVM for IBM z V1.1.0

Physical planning

Nothing much to discuss here : the z13s has a similar footprint and weight compared to the zBC12. The differences
  • it's ASHRAE Class A3 (40°C and 85% relative humidity operating limits
  • Like the z13, it has door locks : both doors (front/rear) come with the lock installed
  • As already said, you can have a separately rack installed 1U HMC. The rack is supplied by the customer.


Apparantly, it's still a bit early to come with full documentation, but this is what is already available for the moment. I would recommand you have a look at the Publications chapter in the announcement.

Redbooks (drafts) :
Web :
  • IBM z13s Web Page
  • Datasheet
  • FAQ
  • z13s Virtual Tour (Link from z13s web page but at the moment of publication of this post not yet active)
  • z13s Designed to Outcompete video on Youtube gives a good summary of the announcement

    Some key dates for the z13s and z13 GA2

    February 16, 2016
    • Day of announcement
    • First day for GA orders
    • Resource Link support available
    March 10, 2016
    • Features and functions for the IBM z13s (Type number: 2965)
    • IBM z13s Models N10, N20
    • z114 and zBC12 upgrades to z13s Models N10 and N20
    • Features and functions for the IBM LinuxONE Emperor, and Rockhopper
    • IBM LinuxONE Emperor Models L30, L63, L96, LC9, and LE1 (5 models)
    • IBM LinuxONE Rockhopper Models L10 and L20 (2 models)
    June 30, 2016
    • MES features for Models N10, N20
    • z/VSE Network Appliance using the z Appliance Container Infrastructure (zACI) 
    • MES features for LinuxONE Emperor Models L30, L63, L96, LC9, and LE1
    • MES features for LinuxONE Rockhopper Models L10 and L20
    • MES upgrades for IBM LinuxONE Emperor Lxx models to Lxx models
    • MES upgrades for IBM LinuxONE Rockhopper Lxx models to Lxx models
    September 26, 2016
    • MES upgrades for IBM LinuxONE Emperor Lxx models to IBM z13 Nxx models
    • MES upgrades for IBM LinuxONE Rockhopper Lxx models to IBM z13s Nxx models
    As usual, I could only scratch the surface and touch upon the highlights of this new system.
    There's much more to discover about it.
    And along with this announcement, there were other announcements focusing on compilers, a preview of z/VM 6.4 and End Of Support Dates for zEC12 and zBC12.

    So I can only say : Stay Tuned !

    Monday, February 15, 2016

    141 cores on a z13 ? Wrong ! It's actually 808 !

    Sometimes you just have to hit the right slide in order to see things in perspective. We're always telling you about the PUs that are available in a system. E.g. the z13 has 141 PUs that can be used by the user. And actually we know that there are more. You have the SAPs that are taking care of all I/O between the PUs and the I/O cage, but there are more. That's why I've come to like this slide so much : it gives you some insight under the covers ànd it's great publicity for z Systems.

    Just click on the picture for a larger version

    I guess a picture says more than a thousand words. I let this one lingering about on my desk, it's a great conversation starter. Especially of course with non mainframe people. Just print it out on an A3 page, make some publicity for our platform and let me know what reactions you are getting !

    Friday, February 12, 2016

    GSE - Regional Conference Belgium-Luxemburg 2016

    Well, two GSE posts in a row. Talking about some publicity. Here's the invite for the regional conference. After years of interchanging between Spa and Oostende, there's a different location and they also shifted one day forward having the conference on a Thursday. The subject looks promising :
    To all GSE Installation and Prospective Members

    It is our pleasure to announce the 35th Regional Conference which will be organised on Wednesday 20 and Thursday 21 April 2016 in Hotel Gosset in Groot-Bijgaarden.

    The theme of this conference is REAL-TIME INSIGHT

    Customers, experts and suppliers will bring papers such as
    - Real-Time Insight in the Context of Cyber Security;
    - Real-Time Insight in the Retail Business;
    - Real-Time Insight in Banking;
    - Real-Time Insight in Health Care;
    - Real-Time Analytics on z;
    - Real-Time Insight using cognitive Computing.

    In a few weeks, you will receive a formal invitation with the final Conference Programme and a Registration form.

    Please insert a note in your agenda for the evening of Wednesday 20 April (Reception and Country Dinner) and/or the full day of Thursday 21 April 2016 (Conference).

    We will enjoy welcoming you at this unique event !

    Thursday, February 11, 2016

    Upcoming GSE events in Belgium

    Here are some GSE Working Group meetings for the upcoming months.

    • Thursday February 25 - GSE Architecture Working Group
    • Wednesday March 9 - z/OS Working Group
    • Monday March 14 - GSE Large User Group
    • Thursday March 17 - GSE DB2 Working Group
    More details below.

    Let's do them in sequential order and start with the GSE Architecture working Group on Thursday February 25, 2016. It takes place in the afternoon in the Sogeti Belgium office, with the theme “Intelligent Systems and Processes” and here's the agenda :
    12:30 – 13:00 hrs. : Welcome Coffee offered by Sogeti Belgium
    13:00 – 13:15 hrs. : GSE and the Theme of Today by Eric Ombelets, Chairman
    13:15 – 14:00 hrs. : Cognitive Business Use Cases by Eric Michiels, IBM Belgium
    14:00 – 14:45 hrs. : Smart Operational Decisions and Processes by Prof. Jan Vanthienen, K.U.Leuven
    14:45 – 15:15 hrs. : Coffee Break offered by Sogeti Belgium
    15:15 – 16:00 hrs. : Machine Intelligence by Erik Van Ommeren, Sogeti Labs
    16:00 – 17:30 hrs. : Networking Cocktail offered by Sogeti Belgium
    You can register over here.

    In March we have the GSE z/OS working Group on Wednesday March 9, 2015. It's an all day event taking place at our own RealDolmen headquarters in Huizingen and the agenda is still to be determined. Registration vill be via the GSEbelux website.

    Next in line is the GSE Large User Group meeting. This is on Monday afternoon March 14, 2016 at the  Colruyt Group Services offices in Huizingen (behind the RealDolmen offices). The subject is 'IT Security in practice' :
    "What are the actual threats and what are the expectations for the future ? For what must we be prepared ? Previously, perimeter security was sufficient. Today we have to accept that  someone can still break in into our system.  How can we efficiently detect /monitor  that ? How to react on such an attack ?  How must you be organized ? Which professional roles / responsibilities must be defined in your company to be capable to react in a sufficient way ?"
    The event starts at 1pm with the following speakers :Stefaan Van Daele (IBM)  - Senior Security Architect at the Global Security Competence Center and Simon Caus - Security Architect at KBC.

    Registration via a mail to katleen.frisch@colruytgroup.com

    And finally we have the GSE DB2 Working Group meeting. This is on Thursday March 17, 2016 at The KBC offices in Leuven. The general subject is DB2 System performance and Tuning. Here's the agenda :
    13:00 – 13:30 : Registration
    13:30 - 14:00 : ‘System Performance at BNPPBF’ - Eddy Tiels, BNPPBF
    14:00 - 15:00 : ‘The A to Z of DB2 system performance’ - Phil Granger, BMC Software
    15:00 – 15:30 : Coffee break
    15:30 - 16:30 : ‘Analyzing performance using Statistics & Accounting reports’ - Cristian Molaro, MConsulting
    16:30 – 17:20 : Panel discussion
    17:20 – 17:30 : RFE (Bart Steegmans, IBM)
    17:30 - 19:00 : Networking & Drinks sponsored by BMC Software
    Registration and more information on the subject over here.

    Wednesday, February 10, 2016

    Welcome to the Mobile Era, the Era of Big Data, the App Era, the Cloud - The New World of Innovation

    This is the title of an article on the Fast Company website. As a matter of fact it's an article on the new IBM system that will be announced on February 16, 2016. It looks like IBM has somewhat changed its strategy around announcements. A couple of weeks ago, with the announcement of the new LinuxONE, IBM kind of hijacked the wired.com site. Now it's choosing another channel for its next announcement introducing it via Youtube.

    In Europe Fast Company is not that well-known but here's a description of them from their facebook page.
    "Fast Company sets the agenda, charting the evolution of business through a unique focus on the most creative individuals sparking change in the marketplace. By uncovering best and "next" practices, the magazine helps a new breed of leader work smarter and more effectively."
    Do read the article. It gives a good overview of the recent innovations that change your way of doing business with an emphasis on mobile, security and data (protection). You know of course where we're heading at. So the article ends with an invitation to a webcast on February 16, 2016 : "Be the first to see IBM unveil its new system in an exclusive webcast, and learn what it takes to outcompete in the era of the cloud. Learn More".

    So, as the customer invitation says "be among the first to see IBM unveil a new system and technologies that offer todays businesses unimaginable potential to out compete with speed and intelligence" and register over here for the webcast.

    Monday, February 8, 2016

    Notice to TS7700 customers to update TS7700 Licensed Machine Code to current levels

    Pfew, my first post for 2016, I've been so busy that I didn't find the time for blogging. But don't think you get rid of me that easily. Not with my 10th blogging anniversary coming up in September.

    Well, let me start off with a Flash on TS7700 that you can find over here.


    IBM advises all TS7700 customers to update to TS7700 Licensed Machine Code release 2.1 or higher. As of July 31, 2016, IBM will no longer provide code updates on TS7700 machine code levels up to and including release 2.0 (8.20.0.x).


    Although IBM will no longer develop and distribute code updates to TS7700 Licensed Machine Code levels R2.0 and earlier as of July 31, 2016, IBM will continue to provide customers who have a valid IBM hardware maintenance agreement (“HWMA”) for this product with replacement parts, application of available microcode upgrades, and technical assistance regarding routine, short duration installation, usage, and code-related questions (in accordance with the terms of the HWMA). Request Price Quotes (RPQ) will no longer be reviewed for TS7700 machine code levels prior to release 2.1.

    Upgrade and replacement paths may be available to enable the continued usage of existing hardware with R2.1 or later supported code levels. Contact your account representative for assistance.

    Following is a list of downlevel machine code and affected Machine Types and Models.
    Machine Code levels and affected Machine Types/Models
    As of July 31, 2016, IBM will no longer provide code updates for the following TS7700 machine code
    where x is any available fix level for the release
    Machine Types/Models that are capable of running one or more of the machine code levels referenced above:
    Machine Type Model Description
    3957 VEA IBM TS7720 Server
    3957 VEB IBM TS7720 Server
    3957 V06 IBM TS7740 Server
    3957 V07 IBM TS7740 Server
    3956 CC6 IBM TS7740 Cache Controller
    3956 CX6 IBM TS7740 Cache Drawer
    3956 CC7 IBM TS7740 Cache Controller
    3956 CX7 IBM TS7740 Cache Module
    3956 CC8 IBM TS7740 Cache Controller
    3956 CS7 IBM TS7720 SATA Cache Controller
    3956 XS7 IBM TS7720 SATA Cache Module
    3956 CS8 IBM TS7720 Cache Controller

    Thursday, December 31, 2015

    Happy New Year

    Well here's my last post for this year. If you missed my year review, just scroll down a bit.

    I don't know what 2016 will bring us but I surely wish every one the best for 2016 : A Happy New Year and good health to you and your family. Or as one of my friends on Facebook put it : "Enjoy 2016 as if there is no 2017!". And, as a Prince fan I had to add of course "... and party like it's 1999".

    Thanks to all my readers for coming by so regularly, to all the people who linked to this blog, to those who kindly retweeted me and to every one who gave me great support and feedback in 2015.

    Looking back on 2015 - Year review

    Personally, my year could've been a bit better. I had a hernia operation back in April and I was out for a couple of months from half September till November due to an unlucky fall, but apart from that, I still had a fruitful blogging year. And within a couple of months I can even celebrate my tenth blogging anniversary. But first, let us have a review of 2015.

    January started off fiercely with the announcement of the z13. No more EC or BC model and the naming also changed from System z to z Systems. I'm still getting used to that. One of the main messages was : memory, memory, more memory. Customers are easily getting at least half a TB of memory. It's the only way to compete with all the in-memory big data analytics stuff that's out there. But of course it remains, as Ross Mauri (General Manager, z Systems, IBM) points out in relation to the LinuxOne, "an I/O monster". Yeah, we like that ! All compilers immediately bring a new version utilizing the new functionalities of the z13. We like that too. Fast responsiveness.

    In February we start setting up our XIV campaign. Not all storage news turns out to be positive as Europe sees a price increase of 15% on all storage components. This seems to be due to the dollar-euro rate and, I must say, we see similar moves with other vendors as well.

    The next months we remain focused on our XIV campaign. The RealTime Compression is announced and we also see the software defined version, Spectrum Accelerate, arriving (ZP15-0090). You cannot not have a software defined version in this modern world. And it opens perspectives for smaller customers as well.

    In April IBM announces the 'z Systems Collocated Application Pricing for z/OS or in short zCAP (ZP15-0244). It's a new pricing mechanism for new z/OS workloads. In short it allows workloads that you would put in a separate environment for pricing reasons to be priced as if they are running in a dedicated environment while technically integrated with other workloads. I sometimes wonder why IBM is making this so complicated. Can't you just say e.g. if you grow 5% next year, we give you a reduction of 10% on your growth. But I guess this would be way too easy.

    In May IBM announces RealTime Compression on the XIV and RealDolmen has its XIV event. We also publish a brochure on XIV : "Powerful Storage with IBM XIV'. You can still download it over here. There was also a preview of z/VSE 6.1 since, you can hardly believe it, z/VSE turn 50 in 2015. And before we have the announcement of the DS8880 later this year, we still have some enhancements to the DS8870 with e.g. the 16Gb FC cards.

    In June we see the End of Marketing for z/OS 2.1 move from September to January 2016. Reason for me to also explain the difference between End of Marketing and End of Service. It's surprising how sometimes the most simple terms still create confusion. Perhaps I should start a series on the basics of software pricing. Actually, why not ? June was also the month I found out that the z13 was actually manufactured in my own home town. How about that ? I passed that sign hundreds of times and now all of a sudden it gets a new meaning.

    Two interesting events in July. z/OS 2.2 is officially announced. But for those who are more into hardware, the one on the change to the start of the warranty period on IBM material is even more interesting. Before warranty started the first of the month following the installation of the CE. Now it either starts the business day following the completion of the installation or it starts 45 days following the shipment from IBM.

    More software pricing news in August. IBM announces Country Multiplex Pricing (CMP) (ZP15-0318). CMP is a new sub-capacity offering that allows clients to use their z Systems capacity within a given country without the constraints of sysplex aggregation rules and many of the limitations of previous reporting methodologies. I'm not aware of any customers using this for the moment in Belgium, but then again we live in a rather small country.

    What's more important, IBM also announced LinuxONE in August : a Linux only mainframe introducing new applications, even new operating systems (like Ubuntu) and even more important a new hypervisor : KVM. As I already told you in my article about this, it's a kind of two worlds meeting each other for the first time. Will mainframers be willing to embrace the open world and on the other hand will, let me call them disrespectfully, open world Linux people embrace the mainframe just because it's become a bit more open ? I really don't know, the future will tell us. Hey, that's already my conclusion for this year.

    In September we have the new Mainframe Contest in the Benelux (and elsewhere of course) which will eventually lead to another worldwide contest among the best contestants as we saw with the announcement of the z13.
    On the tape front we eventually see the announcement that ends direct attachment from mainframe to tape via the tape controllers. In the meantime the last generation, the C07, has become End of Marketing. It became too difficult to support newer generations of drives like the TS1150 or supporting more than 4TB in z/OS. The alternatives are TS7700 and ProtecTIER Mainframe Edition (ME).

    In October there are announcements for IMS V14 (ZP15-0539) and for IBM CICS Transaction Server for z/OS 5.3 (ZP15-0495). But what especially caught my attention was of course the new generation of the DS8000 : the DS8880. More power, more performance, a new 19 inch footprint, a change to the HMCs and to the power requirements (only single phase) along with a completely new license pricing mechanism makes this quite an interesting announcement.

    z/VSE 6.1 and CICS Transaction Server for z/VSE V2.1 were announced in November. For the first time in 16 years (that means since last century in 1999) we have a new version of CICS for z/VSE.

    The last announcement I want to mention was already in November but became available in December : the new IBM XIV Gen3 model 314. The focus is on the improvement of RealTime Compression. Therefore CPU and SSD cache capacity are doubled. Playing advocate of the devil one might wonder how earlier models coped with this as IBM always told us RealTime Compression ran on those machines with no performance loss either ?

    Well, that's it for 2015. Any predictions for 2016 ? Not really, or rather an open question from an open mind. Last year I mentioned that OpenStack was here to stay and I guess that is pretty much the case, but today I'm, let's say, a bit confused. We see a lot of change towards cloud or hybrid cloud not only in the market but we, at RealDolmen, as a service provider are also heavily investing in it. I'm no longer doing infrastructure, I'm part of the hybrid cloud team. And I'm glad about that, it's the way to go, but where does that leave the mainframe in a world of dockers, containers, software defined environments . . . ? I guess we'll just have to find out in 2016. Perhaps it will just disappear . . . disappear into the cloud . . . and . . . be used more then ever.

    Wednesday, December 23, 2015

    Intoducing IBM LinuxONE

    Just watch the short video below, it looks (and even sounds) like a Walt Disney movie but it's reality.

    Since I was preparing a new issue of our z Systems eZine, I guess LinuxONE was an obvious choice to have an article about. So let me also share its content over here.

    The hardware
    In short, LinuxONE is a mainframe machine with only Linux. And when we look at the hardware there are two models : the Emperor and the Rockhopper. And yes, they are named after a couple of penguin species.

    The largest model is based on the z13, that was announced last January and the Rockhopper is based on the z12 BC. Since e.g. the LinuxONE Emperor inherits the characteristics of the z13, you have a machine with a possible 141 Linux engines and 10TB of memory. Or as Ross Mauri (General Manager, z Systems, IBM) puts it : "8.000 virtual servers on an Emperor, hundreds of thousands of containers in one system (...) 5Ghz microprocessors, 320 channels for I/O and each channel has two dedicated power processors, it's an I/O monster, up to 10TB of memory, a huge set of cache (...)". As we all know, being an 'I/O monster' is one of the biggest assets of these machines. If you have the time, you can find Ross Mauri's pitch in a 30-minute youtube video you find over here.

    The message that accompanies LinuxONE is based on three 'slogans'.

    Let me clarify them a bit for you

    Linux Your way
    "IBM LinuxONE allows you to choose the Linux distribution, hypervisor, applications and databases, management tools, and services" (*). This is not only a valid statement for people already running Linux on mainframe but perhaps especially for people who are running Linux on other platforms for the moment.
    At the operating system level, this means we need more Linux distributions, so IBM announced the support for Ubuntu as well. But even more important to pull Linux people into the mainframe world is that they can use a familiar hypervisor. That's why next to PR/SM and z/VM, LinuxONE also supports KVM as a hypervisor. This means that people used to work on x386 platform can now far more easily make the step towards mainframe without having that steep learning curve that was there before.
    But what also intrigues me : when I look at the announcement of KVM on z Systems (ZP15-0359) it indicates that it does not only run on the LinuxONE but also on z13, zEC12 and zBC12. So that would be on IFLs next to e.g. z/OS. And it also support ECKD volumes on the DS8000.
    Then talking about programming languages, development and databases, just let me give you an extensive list, to give you an idea of what's cooking underneath all this.
    Programming Languages: Python, Ruby, Ruby on Rails, R Language, GO, Scala, Clojure, PHP, Java, oCaml, Erlang
    Development & Build Runtime Environments:  Node.js, Apache HTTP Web Server, Apache Tomcat, OpenJDK, LLVM, GCCGO, Zend Framework, Erlang native compiler  
    Cloud Management: IBM Wave, IBM Cloud Manager, IBM Urban Code, Openstack, Docker, Chef, Puppet, VMware vRealize Automation 
    Database: Oracle, DB2 LUW, Cassandra, CouchDB, MySQL, MariaDB, MongoDB, PostgreSQL
    Analytics Tools with IBM Big Insights (Hadoop), DB2 BLU and Spark
    This really opens up a whole new world for us, who are used to z/OS, DB2 and CICS. But it's also becoming a large part of the future of the mainframe. Even if we have to discover an entire new world. So, fasten your seatbelts and let's go. !

    Linux without Limits
    "This is the second pillar of LinuxONE. Lightning fast response times and virtually unlimited scale gives your applications the premium Linux experience they deserve." I think I already touched upon this, when I quoted Ross Mauri earlier with his description of the possibilities of the Emperor. Scale out at its best. But there's also the value of Scale up : you can support "the largest business applications within a single system, without having to spread transactions and data over multiple servers".

    Linux without Risk 
    "Ensuring that your data and services are fully protected and available when and where your clients need them with the industry’s most secure and resilient Linux system."
    I don't have to elaborate too much on this one either : we're talking about availability, business continuity and definitely also security : "The LinuxONE systems provide isolation at multiple levels - applications, containers, virtual servers and partitions, allow for full encryption of the data (...). They allow for end-to-end security, identity and access control to protect your clients and your business reputation".

    Pricing : Elastic pricing
    This elastic pricing is also described as 'Cloud on Premise'. You have a kind of monthly licensing for what you are using during a specific month. So, you don't buy the hardware, you order the hardware that you need, you order the software that you need, it's in your data center, you use it and you're billed on your monthly usage. Since pricing is always a very sensitive matter, I'll quote Ross Mauri again : "This is like cloud like pricing but on prem. It's for hardware and software, you order what you need, you pay for what you use. Your usage can go up, it can go down, it's metered, you pay by the month". But we all know IBM pricing is never that simple. So contact your IBM or BP representative if you want to get a clear picture on all the details.

    IBM zAware and Open Mainframe Project from the Linux Foundation
    One extra item I want to mention in this context is the Open Mainframe Project of the Linux Foundation. The Linux Foundation is a "nonprofit organization dedicated to accelerating the growth of Linux and collaborative development.  In collaboration with the Linux Foundation, IBM will support the Open Mainframe Project, a collaboration of nearly a dozen organizations across academia, government and corporate sectors to advance development and adoption of Linux on the mainframe".
    Now what about IBM zAware ? Along with the z13, we saw already that similar to z/OS, support for the analysis of Linux images on mainframe was announced. Now IBM zAware is becoming available on the LinuxONE machines for native and guest Linux images. But what's more, the largest contribution of source code IBM is making to the open source community is actually IBM zAware. Open source developers will be able to add their own code to it.

    LinuxONE Community Cloud for Developers and students for free
    And finally : do you want to have a go at LinuxONE at no charge. Then you should head over to the LinuxONE Community Cloud : you can obtain a LinuxONE virtual server for testing and piloting emerging applications for evaluation purposes. As a developer or student you can get a 90 day trial that includes up to 2 virtual CPUs, 2 GB memory and 40 GB of storage!

    Some reference material
    • IBM announcement of LinuxONE
    • Information about some Open Source software for IBM LinuxONE (like e.g. MongoDB, Docker, MariaDB, Node.js and Spark)
    • Manual : KVM for IBM z Systems : Planning and installation guide.
    • Manual : KVM for IBM z Systems : Administration Guide.
    • FAQ about access, registration and use of LinuxONE Community Cloud
    • Redbook : Getting Started with KVM for IBM z Systems 

    What I think will be the most difficult step to take is that every one who wants to take the step towards LinuxONE will have to leave their comfort zone. As mainframers, we're not familiar with KVM, MongoDB, OpenStack or containers like Docker, but it's a possible, real future for our (dare I say) beloved mainframe. For people familiar with all these things, it takes a leap of faith to leave the familiar x386 platform or whatever platform they're using for the moment and taking that step towards the mainframe. All in all, this could be a game changer for the mainframe.

    After I wrote this conclusion I started browsing through the Youtube videos again and you know, just have a look at the two videos below. They just show what I tried to express in words.

    (*) quote from the liner notes of a LinuxONE customer presentation. Any other quotes are from the same presentation unless mentioned otherwise

    RealDolmen z Systems e-zine 23


     The 23rd issue of our RealDolmen z Systems Newsletter was sent out yesterday. You can download it over here. Just like the last times, there's just an English version. No more Dutch or French versions. Some of it has already been posted over here too.

    The content ? There are three major topics. One about the new evolutions on the DS888x, I published earlier over here, one about LinuxONE and finally there's some extra publicity for the IBM Benelux Study Tour with a more elaborate agenda. I also focus on some announcements, as usual there are some hints and tips and of course the usual table of EOS dates of the operating systems is also present again.

    Enjoy the reading !