Wednesday, December 26, 2012

Happy New Year

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

We don't know what the next year will bring us but I surely wish every one the best for 2013 : A Happy New Year and good health to you and your family.

Thanks to all my readers for stopping 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 feedback during 2012.

Monday, December 24, 2012

Looking back on 2012 - Year Review

Just try to sum up three important facts/evolutions that colored the mainframe world in 2012. With which ones do you come up in a split second ? I had some trouble you know. Was 2012 a rather quiet year for mainframe ? Well there's of course the announcement of the zEC12, oh yes, and there's the pre announcement of z/OS V2 and . . . and . . . Big Data perhaps ? Or something else ?
Well, let's find out.

My first post in January was about a belated New Years present by IBM with a price change for z/OS for VWLC and AWLC customers. In hindsight I think this was a more important price raise than the FWLC price raise later this year that caused a lot more of comment. I also broadened my perspective on storage. As I said it back then "I can no longer pretend to write about high-end storage attached to mainframe. There's a whole range of midrange storage devices that can also connect to this very mainframe and zBX". So I also started writing about XIV, StorWize ... and no longer just about DASD. One other thing I see throughout this year is that the customer Working Groups (especially GSE in Belgium) usually draw a lot of people and do one heck of a job. They also try to more and more involve young professionals into this. IBM also launches a try before you buy version of CICS TS 4.2.

In February we learn that the End of Support date of DB2 for z/OS 9.1 is set on June 27, 2014. For z/VM there's an EOS date for z/VM 6.1 : April 30, 2013. So, z/VM 5.4 outlasts z/VM 6.1 with half a year. We also had our first major storage announcement in February with Gen 3.1 of XIV. The most important feature of XIV Gen 3.1 is the introduction of SSD read cache.

In March Jeff Magdall, z/OS program leader from IBM, held the MVS Program Opening Keynote at Share and stated that the next z/OS releases would have a 2-year release scheme instead of the yearly releases up to now. And if you didn't get your copy of the source of true wisdom back then it's still there. I'm talking about the CA Technologies Reference Guide for DB2 10 for z/OS.

In April we got clarity about z/OS with a pre-announcement of z/OS 2. z/OS 2.1 will be released in the second half of 2013 and the current n-2 support for migration remains. Standard support is prolonged to 5 years. Price changes, now for Cobol V3 and Enterprise PL/I V3 follow the same train of thoughts as before. IBM has continuously added functionalities to these versions without significant price increases. Now it wants to level the prices with the more recent versions of the products. There were also some enhancements to z/VSE 5.1. Hardware withdrawal for the DS8700 is also announced this month. Mind you, it's only the base frame. This prevents you from buying new DS8700 boxes but you can still upgrade existing ones. And for those still on DB2 V8, by the end of the month the end of support date is reached.

May did not bring us a lot of new things. There was only an announcement on the DS8000 making clear if you're still on DS8100 or DS8300, it's time to move on as you can no longer order expansion frames nor practically anything else for them. Still there was an anniversary to celebrate in May : IBM Tape turned 60 and you can summarize this in a few words : "60 years of innovation – and counting".

More, no, a lot of storage announcements in June on XIV, SVC, StorWize, TPC and also DS8000. And, of course, it's especially the DS8000 statements of direction that caught our attention : Easy Tier will extend smart data placement intelligence to use server resources, communicating with servers to place hot data in server SSD cache to optimize performance. DS8000 will use fast, high-density SSD cache modules to enhance performance.and there will be an Easy Tier support for application APIs enabling servers to communicate with storage so storage can become aware of application workload needs and use that to optimize data placement. I can't wait to see this in action.

July and August seemed to be relatively quiet mainframe months but you know what they say : the sting is in the tail. But before that we should of course also mention that z/VM turned 40 in August and, more than ever, it stays alive and kicking. You know what they say : life begins at 40 ! And to complete this, support for z/VM 5.4 is prolonged. "z/VM V5.4 will continue to be supported until December 31, 2014, or until the z9 EC and z9 BC are withdrawn from support, whichever is later".
The most important announcement of August, and actually of 2012, was the announcement of the zEnterprise EC12 or zEC12. I'm nog going into too much detail, you can read everything over here, but I think it's reasonable to say that it's kind of a logical evolution of the z196 with the addition of two new features : Flash Express and IBM zAware. Not much news to offer on zBX, which only feeds some rumors that it might eventually be replaced by or merged with the new PureSystems of IBM. Something which is afterwards strongly contradicted by IBM. And surprise : there is no new pricing mechanism, allthough MLC software prices can go down up to 7%. That, we must say, is good pricing news of course.

In September we had another price 'change' for some CICS and IMS products. And in Belgium we had, yet again, a sucessful version of the Total Solutions Event at the IBM Forum. Don't forget to mark the days for next year's event : September 24 to 26, 2013.

Quite a lot of announcements in October. I start off with the new DS8870 which promises three times more performance than the previous version. It has a new Power 7 processor supporting a lot more cores and cache. But let's move on. Next one is the announcement of IDAA for z/OS V3.1. The most striking new things : you can put tables on the IDAA without keeping a copy in DB2. So it acts like a kind of archive, of very fast archive, that is. Secondly there's an incremental update feature giving you near real-time accessability instead of 'outdated' copies. Also CICS Transaction Server V5.1 is announced in October. And just like we saw with CICS TS 4.2 in January, it also gets a developer trial version. And there's more : the Early Support Program (ESP) for DB2 11 is announced. GA should be somewhere in 2013. No time to waste : Quality Partnership Program (QPP) for IMS 13 is also announced. This is also some kind of ESP. So also there, estimated GA in 2013. The last announcement is the Withdrawal from Marketing of the base frame of the DS8800. This means, similar to what we saw with the DS8700 in April, that you can no longer order a new DS8800 but you can still order all other components like expansion frames, disk capacity . . .

In November quite a lot of comments popped up about the FWLC price change IBM announced going as far as stating that customers would see a raise of their software bill of about 10%. I begged to differ there because definitely not all customers are affected and as far as I can see, only a small portion of their stack is taken into account making it more like a 0.5% price increase.
For the rest November and December were rather quiet months. I followed my first twitter mainframe debate (#mainframedebate). I'll let you know when there's another one, because it's actually quite interesting as lots of different people are there at the same moment. Just don't do like me by asking questions on customer experiences for products that haven't GA'd yet (FlashExpress that was).

Well, that's all folks. So, let me end by repeating my question : try to sum up three important facts/evolutions that colored the mainframe world in 2012. With which ones do you come up in a split second ? Well ? Or was it a quiet year after all ?

Wait a second, you'll say. Where are the 2013 predictions ? Well, it's not because I blog and, I admit, sometimes think about those things, that I have a crystal ball. I mentioned Big Data at the beginning but didn't really talk about it any further. Well, it's perhaps because I don't really like all those Big Words, but I'm definitely looking forward to evolutions in Data Warehousing and what role IDAA will have in this. Furthermore I'm curious on how the zBX story will continue, because let's be honest, we haven't seen much evolution here in 2012. And yes, I guess there will be a zBC12 in 2013.

To be continued . . .

Friday, December 21, 2012

Much ado about Nothing . . . euh . . . FWLC

I was going to write my year review, yes there's one coming up and I wanted to refer to the FWLC price raise that IBM announced a couple of months ago. But now I see this story popping up again in other year reviews and I do not totally agree with everything I read. I might be looking at this from a more European, or even Benelux, point of view, but still . . .

When I first read the articles and all the excitement about an announcement of a 10% price raise for FWLC softwares (American version 2012 (312-129)), I just thought : no big deal, we had ours last year (EMEA version 2011 (ZA11-1013)) that also bluntly announced that 10% raise. 'Oh well', I thought 'most of the mainframe watchers are focused on American announcements, they just missed this one here in Europe'. I also noticed they did not interpret the announcement all that correctly but what the heck. As I already said : much ado about nothing. I'll tell you why in a minute. But then in December there was another European announcement (ZA12-1087) once again announcing a 10% raise on all FWLC software. Hey, that's adding up to let's say 20%. So I started searching the announcements and found out that there had been the exact same announcement for the United States in January 2011 (311-002). Why no one mentioned it back then ? I don't know.

But why am I not al that worried about it ?

First of all FWLC is a pricing mechanism which has a flat monthly rate per server independent of its capacity. But, perhaps more important, it's only applicable to let's say Enterprise Class machines and sysplex pricing : VWLC and AWLC for most customers nowadays.
Customers on smaller systems having EWLC and AEWLC pricing do not even have FWLC pricing. So when I read that customers that are still on older machines running this old software are the first victims of this, I tend to disagree. These are usually small customers who do no longer update their systems ànd who are usually on small systems, hence, nót having FWLC pricing. So who stays : large customers and sysplex customers. But there again it will only be the larger customers as smaller customers who run sysplexes on BC machines are often cheaper off paying each machine separately with EWLC or AEWLC pricing.
Since this mainly effects larger customers I cannot imagine that lots of those customers, who still invest largely in mainframe technology will have many of these older softwares running.

Secondly, what about the 10% price raise ? Well, I checked it and it's actually correct. Customers will see a 10% price raise on their software stack . . . at least on their FWLC softwares. So, when I read in some year review that software prices for mainframe customers will raise with 10%, that is definitely not the case.  I've checked a number of software stacks and the maximum percentage of FWLC software I found for a customer was about 4% of the total bill. So . . . in reality we're speaking of a raise on the total bill of maximum 0.4% instead of 10%. Hey, now I can even live with the 20% I just mentioned.

So, you know I usually can be critical about price 'changes' by IBM, but not really about this one. Unless the American market is entirely different. But then again why did no one mention it last year ?

Wednesday, December 19, 2012

IBM zAware information

IBM zAware was introduced together with the zEC12. You can find my introduction to it over here. Let me fresh up your mind with this little picture. More info in my previous post.

Click on image for larger version

But that was just an introduction leaving people with lots of questions. So I'm glad I can refer to some additional information through a couple of blog posts by the real subject matter specialists.

Have a look at the IBM Mainframe Insights blog. There was a series of  4 articles on IBM zAware which is definitely worth your attention. The first one 'The Journey of IBM zAware' tells us how the idea for IBM zAware originated and how it was developed. The content of the second one 'zAware Installation and Start up' is pretty obvious as it tells you which steps to take to setup and activate the IBM zAware partition. The third one aswers the 'Top 10 Most Frequently Asked Questions About IBM zAware'. And the last one is actually an 'IBM zAware Demo'. It's a 15 minute demo on Youtube demonstrating the GUI. Oh well, I might as well include it here too.

Don't forget : there's also a Redbook on zAware : 'Extending z/OS System Management Functions with IBM zAware' which has the following content :
  • Chapter 1. Introduction
  • Chapter 2. z/OS System Management Functions
  • Chapter 3. Planning for IBM zAware
  • Chapter 4. IBM zAware Installation
  • Chapter 5. Maintaining and Managing IBM zAware
  • Chapter 6. Integration with other products
  • Appendix A. Syslog Message Analysis Program
  • Appendix B. Activating TCP/IP AT-TLS
  • Appendix C. Using automation to monitor IBM zAware connections
  • Appendix D. Problem determination sample
An just to be complete : there's also an 'IBM System z Advanced Workload Analysis Reporter (IBM zAware) Guide (SC27-2623-00)'.

Combining all this information, you'll be pretty well on your way to become a real IBM zAware specialist.

Tuesday, December 18, 2012

Distributed Energy - The Journal of Energy Efficiency & Reliability

People who know me a bit will probably raise their eyebrows at this title. If there is something I know absolutely nothing about, it's about electricity. When at a Solution Assurance (or should I say TDA) it comes to 220V or 380V, DC or AC, 16A or 32A, I'm at a total loss desperately looking around for any help I can get. It must be some kind of switch (no pun intended) in my head that's immediately turned off when it comes to electricity.

But hey, this is about Energy. And I have a slight suspicion that's not exactly the same. So, I first stumbled across this Distributed Energy Journal in 2009 when it published an article about 'The Green Machine', with the subtitle 'IBM unveils the “most technologically advanced, energy-efficient” data center ever constructed'. I think it's still an interesting read.

So I kept following the periodical and from time to time it published some articles on the building of Data Centers and Energy Efficiency when doing so. Because I know that practically all companies are interested in this matter (money, money in times of crisis) and I doubt a lot of IT people know this journal I guess it might be an interesting introduction for some of you.

The latest November-December issue has an article called 'Supersize IT' writing on how "huge data centers' surging demands are met with innovative power solutions". it gives lots of examples of new data centers and how they are trying to handle energy as efficiently as possible. Yes, I know, I'm starting to sound a bit vague here, but you know why. So all you data center specialists, just head over there and read all the good stuff. And yes, the question whether DC is a better alternative is also discussed.

There, I should go into politics. I can fill a whole page delivering a valuable message without having a clue what I'm talking about. Sort of.

Thursday, December 6, 2012

Why every IT architect should understand something about CICS

Sometimes you stumble across a topic or some documentation that's just too good not to post about (Is that a correct English sentence ?). Even if it has already been brought to your attention through several other resources. So, I read about the new 'Architect's Guide to IBM CICS on System z' in the new CICS Portfolio eNewsFlash which you can find over here. And there were also some links to some other sources about it. So, instead of writing my own post about it, I'm going to do a bit of stealing (yes, even my post title) and I'll give you the references.

Bottom line is : every IT architect should know about the value of CICS. That's why Martin Keen starts his post on the IBM Software blog like this :
How many times did you use CICS Transaction Server this year? This week? Today? Unless you're already familiar with IBM's 43 year-old transaction server, you might be scratching your head and thinking “I've never used it!”.

Think again.

Have you had lunch yet? If so, did you pay with a debit or credit card? Then you've used CICS. Did you pay for lunch with cash instead? CICS entered your life then too – when you went to the ATM to withdraw the money.

And you're not alone. CICS Transaction Server handles a dizzying number of transactions every day. More than 30 billion transactions a day in fact(...)
Go on over there and read about the value of CICS and why every IT architect should know about it.

The same introduction to the Redbook is found on Youtube.

Still not convinced ? Then take a look over here to find out what inspired the writers of the Redbook.

But of course, do go and read the 'Architect's Guide to IBM CICS on System z' over here. And don't think : 'Oh, it's definitely beginner's stuff' because Martin Keen concludes :This is an essential book for architects, and a lot of other people besides. Even people who consider themselves CICS experts will learn something from reading it. I highly recommend it".

Still not enough ? Here's another 10-pager on the 'CICS Transaction Server Application Architecture'.

So, as I always say, do check them out !

Tuesday, December 4, 2012

#mainframedebate, twitter, mainframe blog and Enterprise Virtualization with z/VM

Some three years ago I had my little say about Twitter over here and back then (ages ago) I referred to an article on 'The four stages of the average Twitter user' and I think the article is still relevant. The four stages are : Confusion and Indignation, The first "Aha!" Moment, Remembering to Tweet and Thinking in 140 characters. Three years ago I was definitely a stage one person. But things - and people - evolve. I've discovered that lots of information and links to interesting blogs comes a lot faster via Twitter. So, I'm tweeting a reference to my blog posts myself, I'm checking the tweets of people I follow at least twice a day, I'm re-tweeting interesting stuff and this all mostly via Tweetdeck. Am I hooked to it, not really : I'm only using it for professional purposes and I haven't gone mobile ... yet.

And so my post starts with the hashtag #mainframedebate. For those not familiar with the term, a hashtag is a kind of search argument you can add to a tweet. If people search on it, they get all tweets around this topic. And, if all people start using that hashtag at the same moment, you can start a debate. Well, yesterday I received an invite from Steven Dickens (@StevenDickens3) which went like this :
planning to join us? 5th Dec | 4pm GMT | 11am EST
So tomorrow at 4pm GMT you can join this debate on twitter posing as many questions as you like about the mainframe. You need a twitter account to participate but you can follow the discussion over here as well.
If you want some more information, you can have a look at the Mainframe and Hybrid Computing blog from Steven who also organized the first debate and wrote about it in a couple of posts last September.

It's free association today, so this topic also brings me to Stevens blog. I didn't know it before but Steven writes quite some interesting stuff on mainframe and cloud computing. His latest post is about Cloud Computing and the Cult of x86. He refers to a white paper 'Comparing Virtualization Alternatives - What's best for your business?' which you can find via this link. "The purpose of this analysis was to examine the real-world  impact on businesses that deploy IBM’s z/VM virtualization product, compared to those using UNIX or x86 products". If you want to make a stand for z/VM as virtualization platform this is a very useful document with lots of arguments and graphs illustrating every aspect of the comparison.

And . . . don't forget to join the #mainframedebate tomorrow !