Tandemworld Newsletter for April 2009

Availability Digest Asks, “Why Back Up?”

It’s a sad but true statement that the best lessons we learn are through the misfortune of others. So let’s learn from JournalSpace, the blogging platform that suddenly went out of business when it lost its database that wasn’t backed up. In its April issue, the Availability Digest examines JournalSpace’s demise as the result of a malicious act by its disgruntled IT manager. Having been caught stealing from the company, the manager was fired. On his way out the door, he overwrote the entire database with garbage. Even worse, all attempts at database recovery failed. Worst of all, the database was not backed up. Thousands of bloggers lost years of their work.

Also in this month’s Digest is our Recommended Reading selection, “High Availability Network Fundamentals,” a practical guide to predicting network availability. Our Geek Corner’s “Configuring to Meet a Performance SLA – Part 5” explains how to calculate a performance specification for a set of servers acting in tandem but with different distribution times and carrying different loads.

Digest Managing Editor Dr. Bill Highleyman will speak at the HP Technology Forum this June in Las Vegas. His preconference seminar will take place on the first day of the conference, June 15^th. The all-day presentation will include the defining of active/active systems, why they work, and how to implement them. To get a taste of what this seminar will include, you’re encouraged to participate in a Thursday, May 21^st, “sneak peek” webinar, sponsored by HP. Register for the free webinar at https://www1.gotomeeting.com/register/117299321.

The Digest is free online and focuses on continuous processing architectures, with particular attention given to active/active technology. www.availabilitydigest.com

Learn about Ban Bottlenecks® at the Shows!

Be sure to drop by our booth at the ACI Customer Exchange - Annual Meeting 2009 in Washington, D.C. May 31 – June 3, 2009.

Or, visit us at the HP Technology Forum & Expo, Las Vegas, June 15-18, 2009.

We’ll be presenting our payments-specific integrated service level and capacity management service. We’ll discuss how Ban Bottlenecks helps you become a hero by implementing an ITIL-style proactive service level and capacity management discipline. We show you how to process the future!

Jon Schmidt

www.banbottlenecks.com

NonStop - A Running Commentary – April, 09

The opinions expressed here are solely those of the author
and shouldn’t be viewed as reflecting any views
of his current or past employer

Readers of the articles I have published in TandemWorld, who are regular readers of the Real Time View Blog (http://itug-connection.blogspot.com/) as well as the new LinkedIn user group of the same name (http://www.linkedin.com/groups?gid=1878133&trk=hb_side_g), will not be surprised to know of my interest in the auto industry. They wouldn’t be surprised at how obscure a reference to a car can be woven into a commentary on the IT industry. And the differences between the auto and IT industries seem to be lessening daily.

In the TandemWorld article of February, ’09 (http://www.tandemworld.net/newsletter%20feb09.htm) I quoted from Road and Track magazine, and in the most-recent Real Time View blog posting (http://itug-connection.blogspot.com/2009/04/horse-isnt-dead-and-neither-is-viper.html) I quoted from Motor Trend. For me it’s very easy to draw parallels to the IT industry from what is happening today within the global auto industry, and the aspect I watch out for the most is the potential impact it may have on the choices we have. After all, with lots of different cars on the market, there are lots to choose from and each of us has our own criteria as to which car meets our needs. I have to admit that I tend to draw parallels between automobiles to just about anything else - that’s just the result of my mind never wondering from those wonderful art pieces called cars for too long. In London, there is a particular street I visit with the most amazing art galleries: Park Lane, Mayfair where you can view the most beautiful exotic and over-the-top luxurious cars! But I digress…

It’s almost impossible today to miss articles about the sale of Sun to Oracle. The saga started with Sun making it known that it was up for sale and shopping itself to a select number of potential buyers. Pretty much like throwing an advertisement in the local auto magazine – for sale, well performing, previously owned… computer company. The list of potential buyers was small – rumored to be only four or five companies – and IBM pulled the trigger first.

Unfortunately, this gave IBM to quietly go through the books and after only a few days into its due diligence process, IBM reduced its offer by a few billion dollars. Sun blinked and said no, thanks. And then it waited. And waited some more. No news other than lawyers for the shareholders feeling miffed by the actions of Sun’s board getting organized and filing the first shareholder suits. A few days later, Sun returned to IBM and asked to have the discussions restarted. No-way, came the firm response from IBM, the deal is dead.

Quite unexpectedly to many, and following a weekend where McNealy placed a lot of phone calls, Oracle comes to the table with pretty much the same offer price as IBM’s rejected offer, but this time, both sides were happy. And short of being rubber-stamped by Sun’s board, the deal is done! Oracle is now in the hardware business – sure, it likes Java, the Java toolsets, and MySQL, but it’s also gained a whole bunch of hardware – servers and storage – at margins that are way below what Oracle shareholders are used to seeing in their annual reports.

Compare this with all that we are reading about General Motors and Chrysler. Of the two of them Chrysler is the shakiest, but if Fiat comes through Chrysler may survive, albeit as quite a different company. It was only a few years ago when Fiat had held it hand out for financial support from, of all places, GM – and when the clouds gathered over Fiat, GM gave to Fiat several billion dollars to exit gracefully from any future obligations. Today, Fiat has the money to buy a chunk of Chrysler – enough where the current head of Fiat could become a candidate for the CEO role at Chrysler – and, just to rub salt into the wound, has kicked off discussions with the German government to buy a chunk of Opel (as soon as it finishes with the Chrysler deal), probably spending money GM provided in the first place!.

Today the auto industry is dominated by a push for shared components, and commoditization. Each manufacturer rationalizes production on just a few platforms, and all their models leverage these base platforms. They have learnt to share with each other, and some platforms are being rebadged by competitors. Product lifecycles have lengthened over the past couple of decades to where we now see new models featuring new technology arriving every five or six years – sometimes longer.

Markets are being defined not so much by the mechanicals of the car, but by the “softer” subjective components. Whether it is the engine management system, the chassis / transmission / suspension behavior, or the integrated navigation, entertainment, and communication capabilities, it’s all about software and the programming decisions that establish the product differentiation.

With Oracle snapping up MySQL will the choices of database software shrink? Will Oracle’s move into hardware maintain the choices we already have today? Or is there more to come? With respect to Oracle gaining access to Java, Java toolsets, and MySQL, this is a big win for Oracle and Larry Ellison is excited – if for no other reason than allowing him to further stick it to his old nemesis Bill Gates and Microsoft. But the hardware? Already the stories are coming of how Oracle will sell the server business to HP, or another partner.

In a late braking story “Sun, Oracle address sale worries” the journalist Greg Avery in the Silicon Valley / San Jose Business Journal, that came out after the bell rang on Wall Street this Friday, wrote “several analysts who follow Oracle, including Forrester Research Group analyst James Staten in a Denver Business Journal interview, predicted the corporate database and software company would look to sell Sun’s hardware business after the merger because it would lower Oracle’s historic margins.”

Avery had come across proxy filings by Sun late Thursday reassuring employees, particularly my neighbors here in Boulder, Colorado where Oracle and Sun were “trying to allay concerns that Oracle could sell Sun’s hardware business after the planned merger of the two tech innovators is completed.” Later in the same filing, according to Avery, “Oracle plans to grow the Sun hardware business after the closing, protecting Sun customers’ investments and ensuring the long-term viability of Sun products,” before Sun adds “Oracle also intends to focus the server and storage businesses on our common enterprise customers, where we believe we bring competitive advantage, relationships, and a track record of helping to reduce costs and complexity.”

However, I think that this is mostly posturing and Sun will try to clean up the server business’s balance sheet and manage it to a point where it can be sold. As for buyers , according to Avery, “that wouldn’t necessarily rule out the kind of sale Staten (of Forrester Research Group) predicted that Oracle could pursue. He said Oracle could sell Sun’s hardware business to an Oracle partner company, such as Hewlett-Packard, that would sell Sun hardware in coordination with Oracle’s software and consulting, giving customers the benefit of integrated technology but without Oracle itself having to expand into hardware manufacturing and sales.”

So much for choice – if the hardware does end up with HP, are we seeing our IT options reduced? Coming at a time when global politicians are telling us there will be a future where only five or six global auto companies will remain, will IT end up with a choice between two computer hardware vendors? Perhaps three, if you count Cisco? Four if you add Dell? And is this necessarily a bad thing?

I happen to think that the appearance of completely new hardware technology is not likely to happen anytime soon. New companies, exploiting commodity hardware with innovative software, will definitely be with us for some time – just look at what’s happening in the EDW / BI space. Globally, we look to be heading towards having a small number of base platforms on which all systems are built.

Competition between vendors building on these base platforms will be fierce and the battleground will be littered with losers. Nothing here that would tell you whether we are covering the auto or the IT industries – they are looking a lot alike. And yes, as in the auto industry, where we can still learn so much, it will be all about software and the software will differentiate the products.

Richard Buckle

HP – ITUG: Board of Directors (2000 – 2006) Chairman (2004 & 2005)

IBM – SHARE: Board of Directors (2007 – 2008)

Real Time View blog: www.itug-connection.blogspot.com

NonStop Systems Execution Modes Performance Study

Ross Systems International (RSI) have been investigating the comparative execution rates of the various execution modes: Interpreted, Accelerated and Native, for a MIPS R Series processor running at 125 MHz. and therefore applicable for all Himalaya and S-Series Processors.

They performed the following steps:

a) Took some very CPU intensive cryptographic code, inserted timing code and the ability to switch it on and off by means of console commands.

b) Generated some test scripts for PIN translate commands to be executed on their SVRTEST component and compiled three versions of the HSEMM program containing the cryptographic code.

i) Full CISC interpreted mode, compiled with C

ii) Accelerated version of (i) using AXCEL

iii) Native Mode, compiled using nmc

c) Ran the standard test in timed mode using these three versions.

The table below shows the average results found:

Mode

Transaction time

Speed Up\CISC

CISC Interpreted

90ms

1

Accelerated

5ms

18 times

Native

2.3ms

39 times

These results were considerably more dramatic than the official figures:

Mode

Speed Up

Accelerated TNS mode

5 times faster than CISC interpreted

TNS/E mode

7 times faster than CISC interpreted

However, it must be remembered that the speed up achieved will depends on the processing mix in an application. For instance an application which is performing a sophisticated SQL query will spend most of its time in the operating system where it will be running native anyway. Whereas an application which is doing complicated multithreaded dispatching and cryptography will spend virtually all its time within the process and that's where the greatest speedup can be expected.

Another advantage of going native, especially in C and C++ is portability. The Native C and C++ compilers are much stricter than the interpretive compilers. This means that the code in inherently more portable, but you will have to work at firming up your code, typecasting and checking boundaries. Also with Integrity the maximum addressing mode is 64 bit, this may cause problems with boundaries and fields containing addresses, so you will need to think about the compilation addressing mode.

However, you can not go any faster than native, so if you need the extra cycles:

1. Look at your code

2. Use measure find out which are your most CPU intensive applications

3. Optionally Accelerate them to give quick relief
Use OCA on Integrity, otherwise use AXCEL for RISC

4. Recompile them native, with appropriate fixes.

The results may astonish you.

We are also performing a similar study on the Integrity and Blade range of NonStop Systems and will put the results when we have them on the news page of our web site: http://www.rsi-ns.com/

Alternatively you can contact Rupert Stanley of Ross Systems International at rupert@rsi-ns.com or call +44-1206-392923 for additional information.

TANDsoft Introduces New Enscribe-2-SQL Toolkit

TANDsoft, Inc., a long-time provider of database and application middleware solutions for HP NonStop customers, announces the availability of its new Enscribe-2-SQL Toolkit. The toolkit’s library offers application developers a flexible, affordable alternative to more expensive conversion products or manual conversion techniques.

Taking advantage of our proven OPTA2000 intercept capability, the Enscribe-2-SQL Toolkit permits easy conversion of Enscribe calls to SQL/MP or SQL/MX statements. No changes to the Enscribe applications are necessary. TANDsoft’s toolkit library includes an API that enables developers to write procedures specifying how each Enscribe operation maps to the desired SQL operation. As a result, the full conversion flexibility otherwise available only with manual conversion is retained.

Enscribe-2-SQL Toolkit - Features and Benefits:

Supports both SQL/MP and SQL/MX.
Available on all HP NonStop servers.
Offers unlimited conversion flexibility.
Permits file-by-file incremental migration.
Allows application-by-application incremental migration.
Supports complex set operations for maximum performance.
Supports both compiled and dynamic SQL.

The new Enscribe-2-SQL Toolkit was created for use by companies and ISVs who seek flexibility, economy, and performance for their Enscribe to SQL application upgrades. Visit us at http://www.tandsoft.com/Enscribe-2-SQL.pdf for further information, or call us at +1 (514) 695-2234.

TANDsoft’s HP NonStop solutions also include time-zone virtualization, time-sensitive application testing, and file synchronization for disaster recovery.

Insider Technologies – WebSphere MQ Monitoring

Join Insider Technologies at the EMEA BASE24 User Group (EBUG) Prague 5^th – 8^th May

Join Insider Technologies at the EMEA BASE24 User Group (EBUG) meeting in Prague, from the 5^th – 8^th May at the Corinthia Towers Hotel. The theme will be 'Transforming for the Future'. Please visit the following link for further details:

http://www.insidertech.co.uk/NewsEvents/Events/Events2009/tabid/173/Default.aspx

For visitors who wish to discuss and see Insider’s RTLX Reactor product in action (Real-Time transaction / payments monitoring for BASE24 and BASE24-eps including EMS, WebSphere MQ and XML monitoring), visitors can now register via the Insider Technologies website for booth demos; demo registration can also be performed for the upcoming ACE conference (Emperex).

If you do wish to register for a product demo at EBUG (and/or for the upcoming ACE conference (with Emperex) in Washington on 31/05/09), then please register via the following link:

http://www.insidertech.co.uk/NewsEvents/Events/EventCalendarandProductDemoRegistration/tabid/194/vw/0/Default.aspx

BITUG SIG “WebSphere MQ and Comms” 15/04/2009 (real-time monitoring of WebSphere MQ)

Dave Shields (Insider Technologies) attended the above Special Interest Group to provide a presentation on “Managing WebSphere MQ as a Service”.

The theme of this presentation was to discuss how Insider’s product set can monitor and manage WebSphere MQ as a Service,

not just on the HP NonStop platforms but across a range of disparate platforms, such as IBM, Unix, Linux and Windows.

Monitoring MQ events & objects on HP NonStop

The IBM WebSphere MQ EMS events have all the relevant information such as MQ Managers, Channels and Queues

embedded in the EMS event token, which results in the management of these events being very difficult to retrieve, process

and escalate.

A solution to this problem is provided by the use of Insider’s EMS MQ Gateway product, which is an off-the-shelf

solution (integrating seamlessly with Reflex for graphical monitoring) which finally gives you full EMS event tokens from all

those business-critical MQ WebSphere events (including version 5.3.1.4 on Open System Services (OSS)) that are suffering

from a lack of visibility on the HP NonStop platform.

Reflex also provides a built in Status Agent to give detailed status and configuration information for all Guardian and

OSS WebSphere MQ components, as well as providing exact graphical mapping for the MQ Gateway EMS events.

By using simple screens, the EMS MQ Gateway enables you to set up the rules by which WebSphere MQ events will be converted into fully tokenised EMS events. The process is simple: you identify the portions of the MQ event you want to convert and then identify the conversion rules (what you want to keep from the original event and what you would like to see in the new event) to create the new EMS message.

Monitoring MQ across ALL platforms

Sentra MQ will monitor the WebSphere MQ components of your system as well as the platforms and applications that they

reside on and connect to. You are provided with an overview of the health of your entire WebSphere network, being visible

via a single web-based console view.

Sentra for WebSphere MQ monitoring has been designed to display your WebSphere MQ information as clearly as possible and provides several means of alerting your Business and Support personnel to any issues before they become critical and in many cases, automatically fix the error using built in scripting.

Sentra for WebSphere MQ has been designed to be modular, customisable and incorporates an extensible architecture which automatically monitors your Windows NT/2000/2003, UNIX or LINUX and HP NonStop WebSphere MQ environment, 24x7.

Sentra for WebSphere MQ auto-detects your disparate MQ environments, utilising standard MQ API and forwards the data off-line to a Sentra server, where inbuilt MQ rules drives threshold monitoring and alerting, all in real-time to produce alerts, graphs, charts, dashboards and reports.

For a copy of the “Monitoring WebSphere MQ as a Services” presentation, please visit the Insider Technologies website, where a copy can be obtained.

http://internal.insidertech.co.uk/NewsEvents/Events/Events2009/tabid/173/Default.aspx

A copy can also be obtained via the BITUG website: http://www.bitug.com/downloads.php?cat_id=35

In addition to using standard MQ API for extracting the WebSphere MQ data, Insider have developed an XML agent which enables this data, as well as data from files or from TCP/IP socket-based messages, to be parsed and monitored for trending, alerting and reporting; again, all in real-time.

For further details on Insider’s XML monitoring, please visit the following link:

http://internal.insidertech.co.uk/Products/Sentra/SentraforXMLMonitoring/tabid/189/Default.asp

Insider Technologies – Sentra installed at Standard Bank South Africa (SBSA)

Insider Technologies has recently installed its Sentra service monitoring application at Standard Bank of South Africa (SBSA).

The Sentra application will be providing the Business Application Monitoring (BAM) layer for Logica's All Payment Solution (LAPS)

The LAPS BAM allows users to condense complex processes into a single comprehensive operational dashboard.

The BAM allows business users to define and extract crucial data which is then used to visualise the end-to-end payment process and alert users of exception conditions such as processing bottlenecks.

Key features include:

Monitoring facilities from both the business and technical viewpoint

Early detection and resolution of processing bottlenecks

Configuration and generation of alerts for user-defined exception conditions

Automated alert escalation and delivery via multiple channels (e.g. SMS, e-mail)

Ability to monitor multiple data sources (e.g. WebSphere MQ Queues, JMX Queues) via dedicated agents

Support for drill down to specific problem area to support resolution

Easy extension to provide enterprise-wide monitoring

Integration with other system management packages such as Tivoli and HP OpenView

Insider Technologies – Company Introduction

If the above interests you, then please visit our website for further details, or download our Company Brochure for all you need to know about Insider Technologies Limited.

http://www.insidertech.co.uk/Portals/0/ITL_Company_Information.pdf

Insider Technologies – Newsletter

You can also subscribe to our free newsletter where you will be kept informed of Insider’s latest news. If you wish to subscribe, or you just wish to view our wide range of our products, then please visit the following link:

http://www.insidertech.co.uk/Home/tabid/36/ctl/Register/Default.aspx?returnurl=%2fdefault.aspx

Insider Technologies Limited

Spinnaker Court, Chandlers Point,

37 Broadway, Salford Quays,

Manchester

M50 2YR

United Kingdom

Tel +44 (0)161 876 6606

Fax +44(0) 161 868 6666

Website: www.insidertech.co.uk

Email: support@insidertech.co.uk

Insider Technologies is Quality Certified to ISO9001:2000 and TickIT.

All Trademarks acknowledged.

Looking for a Solution or an answer to a particular problem ASK Tandemworld. We will contact the vendors for you and find a suitable solution that meets for current and future needs.