Manag-E Nordic

HP supports SPACL

2010-12-17T14:35:00.003+01:00

In this previous blogpost we describe the effort which is currently ongoing towards standardising Service Catalogs through the SPACL Consortium (Service Portfolio and Catalog Language). One of the leaders in implementing SPACL is HP Software and Solutions, and their just released version 9.21 of the Service Request Catalog module to HP Service Manager is designed to support SPACL. For this reason, they describe this module as desgined to support "other" backends as well as HP Service Manager.

This is an important step towards standardising how IT services are defined, managed and measured. It will over time enable better provisioning of services through Cloud Computing, and make it easier for custoemrs to "shop around" for services with a given set of descriptive criteria and service levels.

We look forward to seeing where HP will take this over the coming years.

How to Pick an End-User Experience Tool

2010-10-11T18:27:00.003+02:00

Aberdeen Group has stated that they have found "solid evidence" that monitoring the end-user experience leads to a better overall application experience. This not only applies to the experience of users but also gives support staff a better experience. In this article in Network World, Beth Schulz summarizes ten recommendations for choosing an end-user experience tool.

The article cited above lacks a few aspects of such tools that are also important, and that should be included in any evaluation or selection of tools. One important question you should ask yourself is: What is my definition of end-user experience, and how do I quantify it? Because end-user experience may be seen from different perspectives, that may reflect different aspects of the truth.

Wikipedia defines user experience as "User eXperience is about how a person feels about using a system"; a definition that is based on ISO standard 9241-210. This is a much wider and more generic definition than that generally assumed when IT people talk about measuring the end-user experience. But it is still much closer to the reality of users' experiences, than the simple measures of response times and availability that IT operations assumes.

In other words, a concept of user experience should include not just those performance and availability aspects that normally are measured, but also more subjective factors. The look and feel of an application may influence how the user feels about it, simplicity may also impact the experience. A slow but easy to use application may be a lot better - particularly for casual users - than a fast but complex one.

Coming back to the question of criteria for selecting an "end-user experience tool", we need to define whose experience we want to catch. Is it the infrequent user who spends a long time in the application when he first gets here, the frequent but fast user who knows exactly what she wants and need to go straight to the core? And to what extent do we include "related services" (such as responsiveness of a help desk, or users' response to surveys on satisfaction) into the total experience that we (want to) measure?

Should you Customize your IT Service Management tool?

2010-10-07T07:21:00.003+02:00

For years, IT Service Management customers have been heavily into customizing the software they use for Incident Management, Problem Management and the whole ITIL process portfolio. The reason? Because these implementations have to a large extent been driven by a set of "ideal" process definitions, developed by external or in-house consultants, and the desire to 100% match the tool to the process. And also because these tools have traditionally been very "customization friendly", some of them almost being generic workflow development toolsets rather than solutions.

This heavy use of customization has had several consequences that over time have become burdens rather than advantages for the customers:

The ongoing maintenance costs have been high, since many customizations need to be redone or modified after version upgrades and other changes to the ITSM tool.
The tool as a whole becomes more error-prone, as the new/changed features that are customer-specific may add their own set of bugs.
Many integrations (e.g. with network or application monitoring) will also need to be customized, making these more expensive.
And frequently, we've seen that customers get stuck with a certain version level of the tool, because some customizations are not compatible with a new version.

In this article in Network World a user gives his views on ITSM customization, and their acid test for whether to perform a specific customization is: 'Is this something that we truly need to be more effective in how we run the business?' A very good question to ask yourself, and not only related to ITSM tools.

IT Security Acquisition Spree

2010-09-23T10:53:00.002+02:00

Over the last year, HP and IBM have both done several acquisitions in the realm of software for IT security, and it looks like this trend will continue. Some of HP's latest acquisitions are:

Fortify Software: Tools for security testing of source code, to ensure that applications are inherently secure.
ArcSight: Security Incident and Event Management, to discovery security related incidents and automatically handle most, passing the remainder on to humans.

These solutions fit well into HP's overall Business Technology Optimzation portfolio: Fortify will enhance Application Security Center, while ArcSight will add security related event management into the BSM (Business Service Management) stack, and optionally add some other (non-security) features such as log file analysis, something that HP does not have today.

Why are the big IT Management players doing this? We think that one of the main reasons is that the rise of Cloud Computing, and the entire realm of "outsourcing" is provoking a review of data security. Large players such as HP and IBM want to position themselves better for expected growth in this area. And it is always faster to acquire an existing solution than developing a new one. The challenge, though, will be to integrate the aquired software with the company's legacy solutions, some of which may have been acquired previously.

An interesting set of viewpoints on HP's ArcSight acquisition may be found here: http://www.pcw.co.uk/computing/analysis/2269984/arcsight-acquisition-give

Does Cloud Computing Change Service Monitoring?

2010-08-24T11:25:00.003+02:00

In these days when everybody is talking about Cloud Computing - and many vendors are changing their data sheets to say that the outsourcing services they provide are "cloud based" - many companies are asking themselves what the advent of Cloud Computing will mean for their investment in monitoring of the IT services that (somehow, sometime) will move to the Cloud.

The answer - as is often the case with any shifting paradigm - is: That depends. It depends on what type of monitoring you are doing today (and want to do tomorrow), and what type of Cloud Computing you will want to utilize.

Generally speaking, monitoring that is agent-less, and based on open/standard protocols will quite probably still work with few or no changes. E. g. monitoring web applications through "virtual users" is the same whichever technology or vendor actually hosts the application server.

Agent-based monitoring is subject to possible change to a greater degree, depending on the type of Cloud Computing being used:
For Infrastructure as a Service (i.e. the customer has access to the server, and provides the OS and application) or Platform as a Service (the customer has access to the OS and provides the application) agent deployment is more or less the same as with servers that are based in your own data center: It is up to the customer whether he wants to use an agent or not, but there may be restrictions on protocols and connections betweenm the agentinstrumented server and the management server run by the customer.
For Software as a Service (where the customer uses software provided by the vendor) agent-based monitoring is generally not possible, except in situations where the vendor will allow the customer some level of access to agents installed, maintained som operated by the vendor. But this is more likely in "traditional outsourcing", rather than in Cloud Computing.

Network Automation: 4 Essential Reasons

2010-07-14T12:59:00.005+02:00

IT of any shape or form in today's world stops if the network stops. So network uptime needs to be as close to 100% as humanly possible. In order to maximize that uptime, the best approach may be a non-human one: Using automation instaed of people. Jim Frey, Research Director for market and technology analyst company Enterprise Management Associates provides four very good reasons for automating network change in this article.

Are Cloud Services Ready for Prime Time?

2010-07-06T15:57:00.005+02:00

Cloud Computing is not just a buzzword, but a number of vendors have over the last few years started offering new "cloudy" services and not just rebranding their old ones. One venture into this area is (or was, as the case may be) EMC's cloud storage offering Atmos Online. This pioneering service is being shelved (see http://www.networkworld.com/news/2010/070110-emc-shuts-down-online-cloud.html), and no guarantees about continuity of availablily are given to customers.

This is just one example of what often happens with new technologies going through the Hype Cycle: The potential seems endless early on, but when the actual market doesn't take off as expected, the vendors drop out as the market plunges into "The Trough of Disillusionment" (see http://en.wikipedia.org/wiki/Hype_cycle).

So now the questions we ask are: Will more providers disappear? Will customers run away? Or will the market, after a period of turmoil, return and grow into significance?

From this writer's perspective, the ability to manage and monitor cloud services at all levels (availability as well as performance) in addition to simplified service provisioning and ease of comparing and migrating between different service offerings are key to the success of Cloud Computing. And the current lack of a standard service catalog is definitely a barrier to adoption for many potential users.

Does Cloud Computing lock you in?

2010-06-03T17:14:00.006+02:00

The great enthusiasm that many people show when talking about Cloud Computing, has created a perception that this is the new holy grail, that has only advantages and no problems or challenges. But as more and more people investigate what Cloud Computing means in real terms, the gaps in the current offerings become more apparent.

One problem is the lack of standards in this area. Standards processes are underway, but it will take a long time before they are properly established, and the marketplace has ensured that the majority of players support them. The lack of standards creates lock-ins, i.e. it becomes difficult to move to another cloud vendor if you want to - for technical, service or financial reasons. Red Hat's CEO James Whitehurst has expressed his views in this article
http://www.networkworld.com/news/2010/060210-red-hats-ceo-clouds-can.html

Looking into e.g. the "infrastructure as a service" (IaaS) offerings currently available, you quickly see that there is a plethora of vendor specific APIs and commands that you need to learn in order to utilise those offerings. Add to that the differences in what is actually provided, the service levels or other contractual details, and you quickly see that what in theory makes life simpler, actually adds complexity.

The "Management Tool" communitry - open source as well as licensed - is slowly getting its act together, to create the tools and middleware to make the underlying Cloud more transparent, and thereby allowing users more easily to move between clouds. But for now, when you go fo Cloud Computing, you must also select a specific vendor.

CloudSleuth - Performance Comparison of Clouds

2010-05-30T12:12:00.004+02:00

Compuware, with their subsidiary Gomez, has created a new service for anybody interested in Cloud Computing: A generic measurement network to track the performance and availability of applications running on a number of different Cloud Service Providers' infrastructure. Initially, they are using a small two-page "web catalog" type of application, with 40 small product info. pages on the first page and one large image on the second, but this is still a lot better than just relying on statements from the vendors themselves.

CloudSleuth, as the service is called, is in beta currently, and is not being tested from "everywhere" in the world, though from enough places that it gives you a reasonable "feel" for the performance of the different clouds. As Cloud Computing matures, so will most probably CloudSleuth, and future versions may support more complex applications than this simple website. What about a comparison of ITIL Service Management sites, pitching e.g. service-now.com against HP's Service Manager provided as SaaS?

Management of Virtualization is Coming of Age

2010-05-20T16:27:00.003+02:00

In many IT shops, application, technology or system specific tools are used extensively. This tends to work quite well as long as a technology is in its infancy - or even adolescence - but becomes a burden on the IT Operations staff when its usage becomes widespread. By that time, you will need to consolidate management of "silo'ed" systems, for better event correlation and root cause analysis, as well as to reduce the need for many separate specialized groups.

According to Network World's Denise Dubie, the time for integrating virtual system management into the rest of IT Management is now:
http://www.networkworld.com/newsletters/nsm/2010/051710nsm2.html

From a European perspective, this is long overdue. In the Nordic countries, where Manag-E is located, Virtualization moved into the Data Centers several years ago, and is today deeply entrenched there. Consolidating Virtualization Management into overall system management is something we've been eagerly awaiting for quite some time!

Cloud Computing and standards

2010-05-07T12:43:00.003+02:00

Everybody is talking about Cloud Computing these days, and a body such as the Cloud Security Alliance (www.cloudsecurityalliance.org) is actively building national organisations to spread its message ever more widely. Even in tiny Norway there is an effort under way to create a local chapter of CSA (www.joymount.no/csa/csa_norge.doc - in Norwegian!).

One of the proponents of standards for cloud computing in the US is the Nationalal Institute of Standards and Technology (NIST - see www.nist.gov). Peter Mell from NIST is the chair of a federal council which is tasked to - among other things - aggregate cloud computing standards. Many people are hoping their results will apply generally, and not just to US government agencys. See more on Jon Oltsik's blog www.networkworld.com/community/node/60847.

More on Social Collaboration in IT

2010-05-03T08:07:00.003+02:00

HP's new apporach to IT Management through socially collaborative media (see our blogpost http://manag-e.blogspot.com/2010/04/social-collaborative-it-management.html) is being commented on in many blogs. Dana Gardner har a nice screenshot in his blogpost on the topic (see http://www.it-director.com/enterprise/content.php?cid=12057). We will clearly see a lot more on this in the coming months.

Social Collaborative IT Management

2010-04-30T08:44:00.005+02:00

How can IT become more "human" as seen from the users' perspective? How can we manage the systems and applications, solving end-user requests, getting our job done, but without being mired down in formal processes and filling out forms - albeit on a screen rather than on paper?

HP Software - for quite a long time one of the most eager ITILists around - is clearly looking at radically new ideas for how IT should do its job, and it does not mean creating new standardised processes and procedures. If you look at this HP site http://www.48upper.com/, you'll see that they have an intent of creating a social media platform for IT Management, including (of course, you'll say!) knowledge sharing. Probably to be offered through a Software as a Service approach.

Matt Schvimmer, the HP manager featured on this site, came to HP through its acquisition of Peregrine Systems (a HelpDesk vendor) and he has worked within the Service Management area of HP Software since. But it is interesting to note that ten years ago Matt worked for Salesforce.com, so he has probably followed this market closely for years. Looking at this background, one can draw some interesting conclusions as to the direction HP Software is heading in.

Watch this space as this story develops!

Integrating Network Performance and Fault

2010-04-27T08:11:00.005+02:00

HP recently announced version 9.0 of HP Network Node Manager (http://www.infoworld.com/d/networking/hp-updates-network-mode-manager-901). NNMi (as it is now called) has been a network management market leader through the last two decades. Perhaps the most interesting part of the update to the product, is the close integration of network performance (from quality assurance based on IPSLA to traffic monitoring based on NetFlow and good old MIB data collection) into this traditionally topology and fault management focused product.

At the same time, HP has also more closely integrated the Real User Monitor product - acquired as part of the Business Availability Center (BAC) when HP acquired Mercury Interactive - into NNMi. RUM monitors network traffic at the packet/protocol level, and connects these into transactions and sessions. RUM can alert on slow transaction response times, triggering network management to determine whether this is caused by a network problem or by a server or application problem.

Monitoring with agents or not?

2010-02-23T00:37:00.005+01:00

Over the past decade – or more – a discussion has raged regarding the relative merits of basing your monitoring of applications, servers or networks on agents installed on the monitored system, versus monitoring without such agents. Some of the most obvious differences (or advantages) are:

Agents can collect data even when the connection to the management system is down
Agents may perform actions independently of the management system
Agents see things from the inside, agent-less sees them from the outside
Agent-less requires less software installation
Agent-less is less intrusive

A lot of parameters may be discovered/monitored irrespective of whether agents are used or not. And for some tools, you may have a free choice. For other tools there are differences in licensing or pricing, favoring one approach over the other. There are frequently also differences in the resolution possible – at least in a simple manner. There may also be differences in security level achieved, e.g. if one approach sends encrypted data over the network while the other does not.

Several agent-less approaches are not really that: They just use a pre-installed agent from somebody else, e.g. the vendor of the hardware and/or software in question. This is true of most SNMP-based monitoring, as well as monitoring that uses remote access protocols such as WMI (Windows Management Interface).

Improving Decision-Making and Resource Planning

2010-02-01T06:55:00.006+01:00

In a previous article (http://manag-e.blogspot.com/2009/07/it-needs-to-do-their-own-financial.html) we discussed the need for IT to do its own Financial Management and reporting. An example of a business where they have seen this need, and are doing something about it, is John Hancock Financial Services (http://www.johnhancock.com/). According to the executive in charge of IT finance at the company, Donna Alexander, there are two key reasons for their focus on this area:

Enhance their decision-making capabilities
Perform better capacity and resource planning

For more information on this case, see this recent article in Network World: http://www.networkworld.com/newsletters/nsm/2010/020110nsm1.html

IT Management Initiatives for 2010

2010-01-26T13:16:00.002+01:00

Which are the most important IT Management initiatives in your organisation for 2010? respond in a comment, or take this poll on LinkedIn: http://polls.linkedin.com/p/75052/oskwg

Is there a Market if there is only one Vendor?

2010-01-26T07:29:00.003+01:00

Steve Hodgkinson argues in this article http://www.enterprisenetworksandservers.com/newsflash/art.php?776 for the view that Amazon's introduction last month of a "spot prize system" for their (proprietary!) Cloud Computing Virtual Machine offering (EC2) creates a market. "The service operates in a similar manner to the spot electricity market," writes Hodgkinson. This is not correct: The spot market for electricity, pioneered in Norway around 1990, would not work unless there actually is a commodity traded that is offered by multiple suppliers. Amazon's spot prizeing - though of great interest to many - does not a market make!

Cloud Computing - What Does it Take?

2010-01-25T23:09:00.003+01:00

I've recently read a number of articles and blogs on Cloud Computing, as well as marketing materials, and I find that there is ever more hype, and ideas, but not a lot of substantial "this is the way we do it" stuff. One idea presented by several is that Cloud Computing requires some form of brokerage functrion in order to be successful. See, for example, http://www.datacenterknowledge.com/archives/2010/01/22/cloud-computing-brokers-a-resource-guide/ or this press release http://www.gartner.com/it/page.jsp?id=1064712. The latter is also referenced in the first article.

For any type of online, automatic brokerage to actually work in a non-proprietary manner, many prerequisites must be fullfilled. This includes fully "machine readable" and standardised service definitions and service levels, automated and standarised security and deployment systems, as well as payment systems that can be trusted for transactions at an enterprise level. And with the promise (or hype) of Cloud Computing, all of this needs to function at a global level, where customers can take advantage of resources and spare capacity across the entire internet.

As I see it, there is a long way to go before this can get off the ground. On the service definition side, the SPACL consortium (see http://manag-e.blogspot.com/2010/01/can-service-catalogs-be-standardised.html for a few words on this) is heading in the right direction, but they have a long way to go. But there are no non-proprietary solutions that have come any further...

In the mean time, Cloud Computing will rely on more manual forms of contracting for services than the automated, brokered dream that a number of us share.

Managing Virtual Infrastructure

2010-01-18T13:01:00.003+01:00

I just read a good blog article on managing virtual infrastructure:
http://www.communities.hp.com/online/blogs/managementsoftware/archive/2009/11/18/the-best-way-to-manage-vmware-environments.aspx

I've seen a lot of organizations where the VM-experts are stuck taking "normal" first- and second-line support calls because the IT Operations department hasn't gotten around to integrating fault management for the virtual infrastructure with the rest. Perhaps a well integrated solution for fault management that covers virtual as well as physical infrastructure is a good investment?

Cloud Computing Defined

2010-01-15T10:17:00.003+01:00

There are few IT Management subjects these days where the hype factor is higher than for Cloud Computing. Hardly a day goes by without some form of Cloud Computing announcement from a vendor or consultant. But they are not always based on the same definition of what constitutes Cloud Computing. Perhaps we should challenge them more consistently, and ask: Whose CC definition do you use? Your own?

To me, the US NIST Cloud Computing definition seems to be a good starting point, and it would be a good idea if vendors and others could base their use of the term on this definition.

Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.

For more details, see http://csrc.nist.gov/groups/SNS/cloud-computing/cloud-def-v15.doc.

Can Service Catalogs be standardised?

2010-01-11T06:57:00.008+01:00

Service Portfolio, Service Catalog and Service Request are important terms in the area of IT Service Management. And even though ITIL has defined ”best practice” for the processes that handle such items, there is as yet no standard that defines how to specify the content of a service in a generic and interoperable way. This is still something that each individual user, vendor or consultant is free to do, according to his or her likes and dislikes. The same applies to specifying how to quantify the service, and how to define the quality attributes that will be measured in order to report on the Service Level supplied, supporting Service Level Agreements (SLAs).

One consequence of this lack of standardisation is that most CMDB products and implementations connot be made interoperable. This means that once you have defined your services in your CMDB, you are limited in the products you can use to monitor and otherwise manage those services: Any product must be closely linked to the CMDB of your choice!

But now a small group of vendors (BMC, CA, FrontRange, IBM, NewScale, and Planview) have come together to develop a standard for specifying services. The standard is called Service Portfolio And Catalogue Language, abbreviated to SPACL. Recently the SPACL working group released its first set of draft documents: A whitepaper, an object model and a technical summmary. You can download the document from http://www.spacl.info/forum/topics/spacl-version-10-documents.

I recommend that anyone working in ITSM, or who is genuinely interested in this very important subject area, should download and read these documents. You may pick up a lot of good ideas to implement in your own work, even though a final standard is still a long ways away.
Anyone may freely submit comments and feedback to the SPACL working committee, as they explicitly ask for this. Just go their website and log in! It is free.

Network Management a Priority for Cloud Computing!

2009-11-24T10:34:00.004+01:00

I recently read this article: http://www.enterprisemanagement.com/web/ema_ac1109.php. It is interesting to see how central Network Management is becoming again! And this happens when everybody has said that network management is "so '90s". An told us that this century belongs to service management.

The definition of Network Management from a Cloud Computing perspective may over time change our views on what is neetwork management as opposed to system management. Because the Cloud Computing paradigm does change our views - with more and more of what was hardware and cables in the "old world" networks, becomes virtual devices and virtual connections in the "cloud world" networks.

Network Management supporting Service Management

2009-11-03T08:56:00.003+01:00

In many IT organisations, Network Management is one of the functions that is best "hidden" from the purview of the ITIL-driven Service Management organisation. Often this is caused by the very technical nature of network management as a discipline. Making Network Management functionality available to a helpdesk is therefore often seen as a waste of time: "They wouldn't understand it if they saw it" is the attitude we often see and hear at custoemr sites.

That may be true of the more detailed aspects of Network Management (and most low-level technical stuff in other disciplines as well!), but that does not mean that no integration should take place.

Lately we have had success with integrating HP's Network Node Manager (NNMi) product with HP Service Manager through the Universal Configuration Management DataBase. The idea is simple: To make the status of network nodes available to the Service Desk in the same manner as the status of all other Configuration Items. When the first level SD-operator receives a call stating that "it doesn't work", the uCMDB can procide the status for all relevant CIs, including network nodes such as routers and switches. This causes a more correct assignment of the trouble ticket, as the operator immediately can see what CI is not behaving as it should.

Unauthorized Change: The Major Culprit

2009-08-25T09:30:00.001+02:00

According to Gartner and others, unauthorized configuration changes are one of the most common causes of errors in IT Operations. This is not to say operators intentionally misconfigure something, it is rather the complexity and level of detail that configuration of current network elements, servers, OSs and applications require, that cause mistakes to be made. According to Network World, as many as seven out of ten IT support calls are caused by self-inflicted problems (see http://www.networkworld.com/news/2008/020808-productivity-gap-between-mid-market-large.html).

So what do we do? Is there a solution except just demanding that everybody needs to do a better job?

The solution is (in theory at least!) very simple: Do not make manual configuration changes, but leave it to automation instead. Because automation will do the same job in exactly the same manner each time. And automation will - or can - be configured to make a backup before a change, every time. And it will do all the boring stuff, that takes precious time, but becomes extremely tedious for a human.

Network Management is one of the first areas where it is easy to make a Return on Investment through automation, since there are a number of changes - really trivial ones, actually, that take so much time to perform on a large number of devices that they just don't get done:

Updating SNMP community strings
Changing admin or root passwords
Updating ACL's
Backup of configurations - after every change!

And of course, updating software on network devices is just as important as patching server and PC operating systems. More and more threats are directed at network devices rather that servers and PCs.

IT Needs to do their own Financial Reporting!

2009-07-21T15:30:00.001+02:00

For years we have been talking about the need for running IT like a business, but very often that has ended up with yet another SLA-initiative, or an Asset Management initiative; in an ITIL-oriented shop all of this often becomes a question of CMDB-implementation. But is that how businesses are run? Is it the quality and quantity of services (or goods) delivered that is central to how businesses are run?

Our contention is that service level management, asset management and inventory is far from the actual running of IT as a business. Perhaps the most important aspect of running any business is to know what happens to your financial resources. How much do you spend, producing what, and what is your bottom line? And what's the profitability of any product line, department or whatever other units that you can break the total operation into?

With IT often being accounted as a cost center, which in these times means focus on cutting costs - anywhere, at any time - IT is definitely not being run as a business. And one of the reasons is the lack of financial reporting tailored to IT as such. Dana Gardner's interview with two HP'ers in this article and podcast (http://www.ecommercetimes.com/story/67549.html) gives some relevant views on the topic.

Is "IT Service Management" an Industry?

2009-07-15T11:06:00.004+02:00

Keith Aldis, the current itSMF CEO, has recently (see http://www.bcs.org/server.php?show=ConWebDoc.27396) launched the idea that those organisations involved in IT Service Management as defined by ITIL make up an industry, and as such should have a common Standard Industry Code (SIC) assigned to them. Aidan Lawes, the former CEO of itSMF, has used his blog (see http://www.itpreport.com/default.asp?Mode=Show&A=2024) to raise a few questions about the topic, but still to a large extent agrees with it.

Aldis and Lawes seem to agree that one can collect more or less all of itSMFs members into something called an ITSM Industry, and that it is just a question of defining and classifying this in order to assign a SIC to it. But in the author's opinion this is not so. For among the many members of itSMF and practioneers of IT Service Management - based on ITIL or not - there are a vast number of differnet industries and government organizations represented, and they belong to other industries entirely (banks, hospitals, manufacturing companies, services companies, IT sector companies, etc.). To classify some banks as belonging to the "ITSM Industry", because their internal (or outsourced?) IT services are based on ITIL principles, while others are classified simply as "financial institutions" seems like an abuse of the whole intent of Standard Industry Codes.

If there is something that can relevantly be termed the ITSM Industry, it would be those organisations where a majority of their activities (and/or revenues) are built upon the software and services that embody the principles of IT Service Management. This would mean that the above bank would not be a part of the ITSM Industry, but their suppliers of ITSM software, consulting and training probably would.

This is akin to the idea that it is the cattle farmer who is in agriculture, not the restaurant that processes his product and sells it on as something else. Or that a steel mill is in the metals industry, while a car manufacturer that uses the steel to build cars is in the automotive industry.

Cloud Nine: Is Cloud Computing the Silver Bullet of IT?

2009-07-13T15:53:00.004+02:00

This is the year of Cloud Computing - at least from a Hype perspective. "Everybody" is writing about it, talking about it, and doing just about anything except actually using it. Because there is not all that much of an actual functional Cloud Computing service that you can buy and use.

In a recent and well-written article, Network World pours a little cold blood into the veins of clouderati - i.e. those people whose whole existence seems built around the hype of Cloud Computing. The article is called Busting the nine myths of cloud computing, and you can find it here: http://www.networkworld.com/news/2009/062209-busting-the-nine-myths-of.html.

Cloud Computing today, posts more questions than answers. That doesn't make it wrong or irrelevant, it just proves that this is an area in its infancy, and a concept that is yet to be properly defined. For Cloud Computing to fulfill its hyped potential, it will need to come to terms with reality:

A definition of what it actually is must be accepted by most of the players - product vendors, service providers and customers;

It must be possible to quantify it, since any price must relate to some kind of quantity, and no free services - like lunches - can exist for long;

It must be possible to managa it, so that provedesr know what they are delivering and custoemrs know what they are buying.

The variety of delivery and priceing models for outsourcing, ASP and SaaS services, which are much better understood than cloud computing, shows that there is a long way to go to get something new standardized to an extent that makes it applicable on a wide scale.

FCAPS - still going strong!

2009-07-10T12:04:00.004+02:00

Times change, but some things always stay the same. This also applies to much of the IT Management space, even though there have been lots of new protocols, standards and ideas. One of the "unchanging" concepts is FCAPS (see definition on Wikipedia at http://en.wikipedia.org/wiki/FCAPS).

FCAPS was defined at around the same time that we got the 7-layer Open System Interconnection Reference Model (from presentation through application and down to the physical layer, see http://en.wikipedia.org/wiki/OSI_model). This model has also withstood the test of time, and is as relevant in today's SOA-world as it was 25 or more years ago.

FCAPS came out the need for defining interoperable management standards (originally only of networks), and quite simply says that there are five areas that you need to handle:

Fault
Configuration
Accounting
Performance
Security

In today's ITIL version 3 management world, it is important to remember this as well, because it is just as important today as it was in the 80's. And though we may use terms such as Event Management or Incident Management, we are almost always talking about Events or Incidents that can be related to one or more of the FCAPS areas.

From the author: It always irritates me when people associate Event Management only with Fault Management, or even equate the two. You can have Fault Events, Configuration Events, Accounting Events, Performance Events and Security Events. And not all Events are Incidents from an ITIL perspective, since Incidents relate to Service Disruptions, and many events from infrastructure objects (CIs in ITIL terminology) may be masked from affecting the services delivered through high availability solutions.

Anyone designing or implementing IT Management solutions today, be it for networks, servers, applications or services, will do well to remember FCAPS and determine whether the solution handles all five areas or only one or two. And if the solution does not cover all, can you clearly articulate and defend why it doesn't?

SLAs based on Transaction Response Time

2009-07-09T14:57:00.004+02:00

All the hype surrounding SLAs (Service Level Agreements) over the last few years, has centered on a definition of service level consisting of two parameters:

Response Time of an application (or a specific transaction within an application)
The availability of said application and/or transaction

These two parameters, then, should be measured and the results repported based on accumulation over a period of time - typically a month or a quarter. This seems simple, doesn't it? It is even an approach that is used in capacity standards such as the TPC benchmarks, so this must be good!

Then to the crux of the problem: How do you actually measure this, in real life, not in the lab? Basically, all the cometing tools out there, do this by one of two different apporaches (some reporting tools may actually combine results from both types of measurement for an "advanced" SLA report):

Generate synthetic transactions on a test computer somewhere in the network (called a probe), and measure the actual response time of this "dummy" transaction. Given that the probe is located "close" to one or more end-users in the network, this wil lgive you an idea of the response time for the dummy transaction from one site; or several, depending on how many probes you want to install, manage and collect data from.
Insert a passive listening device in a network segment or (or mirror port of a switch) that the relevant transactions pass through, and decompose the packets in real-time to measure the time between the packet(s) that contain the input transaction and the ones that are the response to it.

Both of these approaches, though valid to a point, are flawed, because they do not measure what we actually want: The actual end-user experience. A customer of ours with 600 end-user locations is daunted by the administrative overhead of managing all those generators of synthetic transactions (and the additional network load they represent!), while still not being happy with measuring the transactions while they travel thorugh the data center, which is where it makes sense to install the passive listening device.

There is actually only one approach that can give this customers - and many others - what they really want. This approach was developed in the early 90's, and has been a standard since HP and IBM contributed it to the world in the mid-90s. It is called Application Response Measurement (see http://www.opengroup.org/management/arm/) , and could give customers such as the one in this example what they need. If only the application vendors implemented it!

So, go tell your application vendor to download the API from the OpenGroup website and insert it into their application clients. There are several tools available for collecting and reporting on this data, and it will - finally - give users the SLAs they want and need!

Is Automation a Requirement for Virtualization?

2009-07-08T10:54:00.003+02:00

As more and more enterprises of all sizes move into virtualization, a number of changes happen in their IT Management operations and procedures. The rate of server deployments increases significantly, which again means that operating systems and applications are reinstalled more frequently, rather than being upgraded and/or patched "in place". And though it is much faster to redeploy a virtual machine than a physical one, a lot of manhours are needed to ensure it is done correctly.

More repetitive tasks, often means a higher probabilty of failures, due to less focus on the task at hand, or the use of less skilled people to do it. But repetitive tasks are - of course - easier to automate than those with a higher degree of customization. So our suggestion is to add automation to to your virtualization strategy, so that you can do the more frequent deployment tasks even faster!

For more on this topic, download this ebook sponsored by HP from our website:
http://www.managenordic.no/sites/m/managenordic.no/files/890091842.pdf

Software as a Service for Infrastructure Management

2009-07-07T08:01:00.003+02:00

For a number of years, services vendors have been offering Infrastructure Management as a service, typically as a low-ened solution instead of complete outsourcing. With the current industry focus on anything that may be called Software-as-a-Service, or hyped with the word "cloud" attached to it, we are seeing a lot of new offerings presented in the guise of SaaS or Cloud Computing. But is the market really ready for this?

In a recently published survey on SaaS and Infrastructure Management (download from http://whitepapers.theregister.co.uk/paper/view/885/itbe-dell-infrasturcuture-mgt-study-20081216.pdf) only 12% of respondents use SaaS solutions for infrastructure management, and almost all of them in combination with regularly licensed software. SaaS obviously has a long way to go in this market space.

The report also finds that the five most common major concerns with SaaS-based solutions for IT Infrastructure Management - from the perspective of those currently not using it - are:

Lack of integration with other in-house applications
Security risks
Performance issues / more downtime
Not enough control
Higher long-term costs vs. on-premises solutions

As SaaS and Cloud Computing mature, many of these issues will be addressed, but for the foreseeable future, there will probably be a significant market for licensed as well as SaaS solutions in Infrastructure Management, as well as in the wider IT Management area in general.

2009-07-06T10:59:00.004+02:00

IT users make mistakes, but the most serious mistakes are generally made by those who manage IT systems and networks. In this recent article http://www.networkworld.com/news/2009/070609-network-managers-mistakes.html Verizon lists ten serious ones that network managers far too often make.

These mistakes can in most cases be prevented through the use of either better processes or better test and management software - or both! Process and Security standards are not just hot air and paper, they are good guidelines for good IT management. And with the proper management tools configured effectively, errors can be avoided!

Video Marketing

2009-07-06T08:39:00.003+02:00

J9 Technologies has created some very simple but innovative videos to present what they do. A significant improvement on using a zillion PowerPiont slides that bore the audience to death! Read their blogpost about it at:
http://j9tech.blogspot.com/2009/07/video-speaks-thousand-words.html

NOrwegian Vivit User Group

2009-07-06T08:29:00.003+02:00

In January of 1996 Norwegian users of the product family then called HP OpenView created the first Users' Group for OpenView outside of the USA. This was then called NOVUG - Norwegian OpenView User Group. Since the start, it has been affiliated with OpenView Forum International (OVFI), the international users' group. When OVFI changed its name to Vivit, NOVUG followed suit and changed the name, but not the abbreviation. Today it is called NOrwegian Vivit User Group, and still has its website at www.novug.no

The upcoming major event for NOVUG is the annual summer/autumn event, which takes place in Oslo on August 27th. The format is a full-day conference (four parallel tracks) followed by a seafood dinner and beer on a sailing ship on the Oslofjord. See the website for more information.

Join Nordic IT Management Forum on LinkedIn

2009-07-06T08:26:00.002+02:00

There is an interesting group available on LinkedIn for those interested in the IT Management area in the Nordic region. This is located at: http://www.linkedin.com/groups?gid=1587337

Blog Inauguration

2009-07-06T07:56:00.000+02:00

Manag-E Nordic AS is a consulting and software company in Norway and Sweden. We supply IT management solutions based on HP's Business Technology Optimization (BTO) portfolio to customers in Norway and Sweden. Having received HP Software Business Partner of the Year awards almost every year since we started in 1999, we can legitimately claim to be "the leading HP BTO partner" in the Nordic countries.

Check back every now and then - we hope to supply you with our views frequently, though at random intervals ;)