Wednesday 29 June 2011

Things that will make capacity management easier to do well…



KISS

‘Keep it simple, stupid’, the old business saying is more relevant than ever. 

Many capacity management techniques used in the ‘old days’, such as simulation modelling, are no longer relevant.  Configuration of systems can be achieved too quickly and resources are too amenable to rapid change.  Capacity management techniques need to evolve in keeping with this. 

Focusing on key metrics, not all metrics, allowing the varied audiences for capacity information to select what they need to see, all play a part in meeting the new demands on the capacity manager. 
At heart, there will still be a ton of data needed to satisfy all requirements, so automation of data capture and being able to bring all relevant data, technical and business, into one place will be a major benefit.

Business, business, business
As mentioned in a previous blog in this series, getting close to the business has become ever more important for the capacity manager due to technological change such as virtualization and Cloud. 
It’s easier to make mistakes quicker with virtual systems, and whilst Cloud promises elasticity of resource availability, that elasticity will come at a premium cost. 

The analyst community, ITIL and more all stress the need for capacity management to be closely aligned to the business.  Getting good processes in place to enable this to happen is the key to enabling your organization to enjoy the benefits of good capacity management, benefits that are typically worth about 10% of your IT budget per year. 
Getting essentially technical staff to be able to fulfil this need often needs outside training and support initially.

Let the end justify the means
Capacity management includes many techniques and approaches to achieve results.  One size doesn’t fit all requirements. 

For highly dynamic systems, or systems that are non-critical, automated reporting on a few key metrics, report generation only if thresholds are breached, trend based alerting ad reporting, might all be sufficient to meet the business need. 
A critical application with a heavy financial dependency to the business might require a more detailed approach, such as using analytical modelling to identify cost savings on merger and acquisition.  Build a team with a range of skills and match the approach to the business need.

There is more that can be said of course, but that will have to be the subject of future blogs. My series is now at an end and has hopefully given you some thoughts about good practice for capacity management today. 
In particular my blog series is intended to help those organizations wanting to establish or re-establish their capacity management function in response to technological changes such as virtualization and Cloud computing. 
I hope you've enjoyed the series and welcome any comments.
White papers, podcasts and more on capacity management can be accessed free by joining our Metron community


Andrew Smith
Chief Sales & Marketing Officer

Monday 27 June 2011

Things that will make capacity management harder to do…

Some challenges inevitably get thrown in the path of the Capacity Manager as he tries to do his job.  While challenges are just that - challenging and thus not easy to overcome, not everything thrown in the path of the capacity manager is a valid issue.

‘We’re too dynamic’

Both within IT infrastructure and outside, there seems to be delight in saying that things change too quickly to make the predictive aspects of capacity management feasible.  Perhaps this goes with the fire hose issues in my previous blog. 

If you feel you are the hunky hero striding in to put out the fire and carry the rest of the organization to safety, being told that the fires can be avoided or minimized might not be what you want to hear. 

Certainly with virtualization and other trends, we can provision and change IT resources faster than ever before. 

That also means we can waste resources and money faster than ever before as well of course, if we don’t put the effort in to understand what really will be needed and when.

‘But cheap hardware keeps getting even cheaper’

As above, it’s then easier to waste money faster. 

Gartner has been warning of ‘virtual server sprawl’ replacing server sprawl for several years now. 

If your host utilizations are less than 30%, you’ve got money you can save by using resources better.
Under-provisioning and using the Cloud is often touted as the way to avoid this virtual server sprawl. 

Of course, Cloud resources are like your bank overdraft – nice to know they’re there but expensive if you have to call on them. 

There will always be a premium price on resources provisioned from a third party at short notice (whether that third party is internal or external) and letting the capacity manager plan what is really needed within sensible limits of flexibility will always prove more cost effective.

‘We’ve consolidated suppliers and our key partners look after us’

This logic has moved through the outsourcing boom to Cloud providers, including having a single key hardware provider. 

It all still comes down to asking the fox to look after the chickens. 

From the provider’s perspective, there is less risk in selling ‘safe’ (over-provisioned) services in advance than in selling a cheaper service and maintaining a client relationship characterized by service quality problems if under-provisioned. 

Service penalties are not really an answer: I just got $50 back from a car rental firm for supplying a faulty GPS in a car.  Nice, but I still lost an hour of my life navigating out of and back to Logan airport in Boston to get the car changed. 

A key thing to remember, especially with Cloud, is where the responsibility for getting your sizing correct lies.  Unless you’re buying full Software as a Service, it is down to you to buy the right resources from your provider, not for them to know your business and define for you what you will need. 

It’s more critical than ever in a Cloud world that capacity managers get the opportunity and the support to define what resources are required, to avoid those nasty premiums that go with that wonderful elasticity of resource provision.

‘We don’t know what's going to happen’

Often when the capacity manager talks to users, they claim it is impossible to predict in their world. 

How much the business will grow, when new initiatives will take affect are all business drivers that they cannot control.  When they do take effect however, those managing the IT infrastructure are supposed to support them at a cost defined before the changes occur. 

This is where capacity management needs to be implemented with high level management support within the business. 
It also needs to be set up as a regular, iterative part of the business process. 

Demand becomes clearer over time, so regular interaction between capacity management and business leads to an ever-better refined picture of needs. 

This also allows capacity management to feed back early the broad implications of initial guesstimates’ by the business of what they might be planning. 
This can rule out gross errors at an early stage and help the business refine its own understanding of what they might need. 

As mentioned earlier in this blog series, this is a sensitive and skilled area of capacity management that might benefit from external consulting from experts such as Metron, to take the heat out of that IT-business relationship.

‘No one understands me’

The capacity manager can of course be part of his own problem.  If no one understands you, the problem is yours not theirs. 

Capacity managers need to be able to talk with their varied audience (IT infrastructure; application/service owners/business).  It's the capacity manager’s job to do this in terms that that particular audience can understand.  This is an area where external consulting support can help educate capacity managers and set up a viable capacity management process. 

While the capacity management role technically has become less esoteric than it was 20 years ago, the need to liaise with business has become stronger.  This requires just as much skill as understanding the inner workings of analytical models, but capacity managers need to evolve and understand that it's a different set of skills.

Enough about challenge, let’s have some thoughts about how these challenges can be overcome.  Not now though, that will be the topic of my next blog.

If you’d like more thoughts on capacity management  feel free to browse and download white papers on capacity management by joining our Metron community


Andrew Smith
Chief Sales & Marketing Officer




Friday 24 June 2011

Things you will not need for capacity management….

A pager

Does anyone still have one?  A Capacity Manager shouldn’t need a pager or whatever has replaced the pager in the modern age.  As I mentioned in a previous blog, there is a grey line somewhere that separates performance monitoring and system administration from capacity management.  The Performance Manager or Systems Adminstrator needs a pager or other means of alerting them to a crisis.  The Capacity Manager should be operating on the other side of the grey line, beyond day to day crises, looking ahead to see how those crises can be more effectively avoided in the future.
A dashboard

More contentious certainly, as many view a dashboard as a reporting mechanism for any reports.
A capacity manager might well want to set up a reporting dashboard for recipients of information about capacity issues.  It could equally be called a portal, framework or something else – in essence it is just the modern form of the periodic capacity report. 

What the capacity manager shouldn’t be concerned about is an on-line real or near real-time dashboard.  If the information is getting refreshed on the performance side of that grey line that divides current and future issues, it’s not a capacity management issue.  If someone wants to know what their CPU utilization is at a given moment in time, tell them the truth: it is either 0% or 100%.  Leave reporting that to the system administrators. 

If they want to know what it will be next month if the marketing department runs the planned advertising campaign, look at the capacity dashboard/portal/report.

A fire hose
Again, as per my previous blogs around this topic, capacity management is about knowing how many fire hoses to have in place, where to place them, what chemicals to have in them and more.  Putting the protective suit on and pointing the fire hose is for performance managers and system adminstrators. 

A good maxim for a capacity manager is:
If I’m reacting to incoming user complaints about poor performance, I have already failed.”

In economically challenged times, the fault can lie with management.  There is clearly knowledge overlap between a performance manager and capacity manager.  There can also be product overlap as well, as both need to rely on similar or even the same sources of data to do their job. 

Putting the responsibility for both tasks on one set of shoulders is too large a burden however.  Something has to give, and it will usually be the longer term proactive management of systems in favour of the short term crisis resolution.  The shame is that the longer term view offers much more effective management and reduction of costs than the former.

In the next of this series on Monday we'll look at things that can make the job of the capacity manager harder to do successfully.

Browse and download white papers on capacity management by joining our Metron community


Andrew Smith
Chief Sales & Marketing Officer


Wednesday 22 June 2011

Things you will need for Capacity Management

Generally capacity management is seen as a ‘good thing’. Generally it gets little attention, budget or assistance.  A contradiction, but why? 

The issue is usually short term issues and crises.  If the building is on fire, everyone knows why you need a fire hose and using it gets priority.  To handle that fire effectively though, someone needs to have thought out and answered how many fire hoses might be needed, where they need to be located, what chemicals they need to dispense for different sorts of fire and more.  Unfortunately in IT, we rarely get ‘fire free’ periods when these plans can be laid down.   That’s why we need capacity management separated from system administration, so the longer terms plans can be addressed without the distraction of the day to day problems.

Having a plan always helps.  Things that you will need to move towards an effective capacity management capability are: -

• A clear set of objectives

• Senior management commitment

• Process/flow definition

• A realistic plan

• The right people

Key within this is having a clear idea of the service level that must be provided.  This doesn’t have to be enshrined in any sort of formal Service Level Agreement, although a good SLA is of benefit to all.  However, without some idea of the target, it’s difficult to know whether or not you have hit, or what you will need to do to hit that target in the future. 
Essentially there is a balance between three items:

Change one and you change the others.  For example, if the business wants to grow, i.e. workload will increase, either you need to change the resources to cope or the service level will degrade.  Use this balance as a mechanism for talking through with the business what they want from IT.  Do they want to reduce costs (fewer resources)?  OK, but that means a degraded service level or less work can be processed.  Do they need faster response times (better service level)?  OK, but that means less work can be processed or more resources will need to be bought. 

We like to hide behind new terminology to go with new technology, but really the issues at heart are the same.  Today’s rapid provisioning of VMs is yesterday’s CPU upgrade or additional memory – resources.  Today’s hosting another 20 VMs is yesterday’s processing another million CICS transactions – workload.  The key is to understand the variables in the equation and use that equation as a means of sharing understanding with the business. 

On Friday we’ll look at things that can get in the way of good capacity management.

In the meantime feel free to browse and download white papers on capacity management by joining our Metron community


Andrew Smith
Chief Sales & Marketing Officer

Monday 20 June 2011

Capacity Management is not Monitoring

My blog series outlines some thoughts about establishing capacity management in your organization. 
A major concentration is the differentiation between capacity management and related activities such as monitoring and provisioning. 
As you’ll see, I strongly believe that capacity management is all about getting your eyes raised from any immediate day to day crises and concentrating on getter longer term capacity provisioning correct.  Especially in today’s hurly-burly roll in, roll out of virtual systems, short term decisions about provisioning are being mistaken for strategic capacity planning and management.

The overall flow of my successive blogs is:

         1.        Strategic Capacity Management and its objectives

2.       Things you will need for capacity management

               ….and things you don’t need

3.       Things that will make capacity management harder to do

4.       …..and things that will make it easier to achieve

 Strategic Capacity Management and its objectives

There have been many definitions of capacity management over the years, but despite changing technology, the core is still best expressed within initiatives such as ITIL, with wording along the following lines:

The continuing provision of consistent, acceptable service levels, at a known and controlled cost.

So what does this definition mean for the would-be capacity manager?

(a) ‘Acceptable service levels’

• Given a requirement to support a given workload at a given service level, derive the resources required

• Given a set of resources and a workload level, define the service levels that will be provided

• Given a set of resources and a service level requirement, define the workload levels that can be supported

All capacity management questions you get asked by management or the business will effectively be one of these three issues, however expressed.

(b) ‘Continuing provision’

This means that it is not sufficient merely to ensure that current performance is satisfactory.  Capacity management is not about ‘current’ provisioning, it’s about ‘continuing’ resource requirements.  Resolving a short term performance crisis or adding a new VM in response to a user request is system administration.  Knowing what overall resources and how they will need to be configured best to support all provisioning requests for the next six months is capacity management.  Somewhere there is a grey line where the ‘here and now’ becomes ‘the future need’.  Finding this grey line and learning how to manage it is where assistance from external  experts,  such as Metron,  will help.

(c) ‘At a known and controlled cost’

This means that a mechanism must be provided which will define the required resources
to support a given workload at a given service level.  This takes us to where capacity management interfaces with the business. 

Day to day system administration tasks are technical and don’t necessarily need alignment with business needs.  Longer term capacity issues around managing costs, service quality and expectations do.

On Wednesday I’ll take a look at what thoughts and processes you need to start putting in place to make capacity management effective.

In the meantime feel free to browse and download white papers on capacity management by joining our Metron community


Andrew Smith
Chief Sales & Marketing Officer



Friday 17 June 2011

Cloud Computing and Capacity Management

One significant change with the move to Cloud based systems is that capacity management becomes a much more strategic activity than in the past. 

Analyst groups such as Gartner are promoting this evolution and it is further supported by initiatives such as the ITIL® v3 refresh.  Rather than a purely resource level task, capacity management now needs to be an integral part of how the business chooses what is the best solution, for example between private and public Cloud.

What you buy across the cloud could vary from simple Infrastructure As A Service (IaaS) such as processing power and disk storage through to full Software As A Service (SaaS) offerings like salesforce.com, where the provider delivers you hardware, application and more importantly, is responsible for service quality.

For anything other than SaaS, the need to use capacity management techniques to plan requirements in advance will be ever more important.  Buying something in advance is cheaper than buying at the last minute. Emergency buying of Cloud resource (cloud bursts) might be easy to do but it is likely to be prohibitively expensive over time.

With service potentially coming from a variety of internal and external sources, guaranteeing service quality becomes both more difficult yet more necessary. 

What is without doubt is that the range of environments you will have to manage will become ever more complex.

What you can and should do in terms of capacity management will vary with the nature of your own implementation.

From a capacity perspective you won’t be able to measure everything you want in the Cloud so measure what you can, control what you can and don’t worry about the rest.  Tools and processes that support this open approach such as those provided by us are an essential.

Find out what it will be realistic for the capacity manager to provide to the business in this complex world of interacting services by coming along to our free webinar on June 30.


Andrew Smith
Chief Sales and Marketing Officer

Monday 13 June 2011

Capacity management of SAN storage


Capacity management of SAN storage is typically a lower priority than more tactical operations like allocating storage, hardware updates, and firefighting.
Determining who is responsible for capacity management can be a difficult question to answer.
There are overlapping areas of responsibility and internal politics that complicate the issue. Those who normally do traditional capacity reporting may not have sufficient knowledge of the SAN environment to effectively report on it.
Once capacity management processes are put in place, acting on the information is critical in gaining value from the effort.
Quantifying the value gained can be difficult to measure and publicize, but it needs to be done.
Storage data can be looked at from multiple perspectives such as the host, fabric/network, storage controller, and other devices.
Collecting data from all these sources for centralized reporting is a major challenge and should be prioritized. Two key areas to focus on are response at the host and storage capacity utilization.
I’ll be discussing ways to assist the storage administrator in the complex task of managing SAN attached storage at my webinar ‘Capacity Management for SAN Attached Storage’ this Thursday, June 16 – register now and join me.


Dale Feiste
Principal Consultant


Friday 10 June 2011

Growth and capacity management

As many economies move out of recession, their capacity concerns will quickly change to managing growth and Capacity Management will have a leading role to play in ensuring that this growth is supported.

Some markets are already starting to see a flurry of merger, acquisition and consolidation and effective capacity management takes the risk out of these changes.
One way it does this is through providing the means to control suppliers.  In a Cloud world where resource is controlled outside your own environment, this is ever more important.  Use Capacity Management tools and techniques to plan what you need and manage your supplier relationship with this knowledge: a sound view of what you need and when, rather than being driven by their influence.

In this way the business requirement will drive capacity, not technological factors.  While the CFO might loosen the purse strings as we enter a growth phase, they have seen the costs of over-provisioning on IT infrastructure and are unlikely to countenance the same excesses of risk avoidance again.  By matching what the business needs against what that service needs to cost, approval for genuine requirements will be easier to achieve.
Take a look at our free downloads section for white papers on capacity management http://www.metron-athene.com/_downloads/_published_papers/index.html


Andrew Smith
Chief Sales & Marketing Officer

Wednesday 8 June 2011

Managing SAN attached storage

Doing capacity management properly means getting in front of potential capacity related problems before they happen and efficiently sizing resources.

SAN storage is one area of the IT infrastructure that is typically not as transparent as others are. This lack of visibility means that if problems are not discovered proactively the risk for degraded end user performance, outages, and under-utilized resources increases.
There are many examples where high impact problems could be evolving in your SAN infrastructure hidden from view. Identifying these problems early on, before end users are impacted, is what capacity management is all about.
Determining who should be responsible for capacity management of storage, that is SAN attached, is not always clear. Storage administrators are typically busy doing other high priority tasks and capacity analysts do not have sufficient knowledge of the SAN storage environment.

I’ll be discussing ways to assist the storage administrator in the complex task of managing SAN attached storage at my webinar ‘Capacity Management for SAN Attached Storage’ June 16 – register now and join me.


Dale Feiste
Principal Consultant


Monday 6 June 2011

Why the CIO’s secretary can be your greatest asset in Capacity Management

More and more these days, the Capacity Planner is asked to present their information to high levels within the organization. 

The types of reports and data that you present to your fellow technical colleagues is different from what you should present to the CIO. 

 It’s not because the CIO is uninterested in the detailed information. It’s simply that their time in which to digest the information is limited and this is where their secretary can be a valuable asset to your presentation. 

Usually at the CIO level, assistants have been with that person or in that position for years.  They are in tune to the habits and workings of the CIO and what makes them tick.  The secretary knows what makes up a successful presentation for that executive and getting their buy-in can help you get your message understood. 

Finding out the questions the executive always asks when reports are submitted to him, how he likes data to be presented to him and what items or areas to focus upon will all help you fine-tune your message. 

Spending time with the secretary going over your presentation can give you an advantage in getting your information across – she’s already probably helped prepare more presentations for him than you will ever present to him. 

With this in mind, two key items highly likely to be asked are “How much is this going to cost?” or “What is the impact on the business?” 

This is where introducing cost metrics into your capacity reporting can be valuable.  To display cost information alongside the technical metrics adds value in your presentation and in addition to showing the CIO what the issue is, it shows him the impact of not acting, in terms he understands. 

Our recent webinar, ‘Capacity Management – Is it a business service?’ covered just this area. 
We will be re-running the session in due course, but if you would like a preview of what is covered, feel free to download the slides http://www.metron-athene.com/documents/webinar_presentation/metron_cm_business_service_2011-05-19.ppt

Charles Johnson
Consultant

Friday 3 June 2011

No one should forget the mainframe

Once seen as the lumbering beast in the background, now acknowledged by those in the know as the fastest, most secure, tools rich large server there is. Metron’s Acquire for zOS data capture facility is a non-intrusive, zero impact performance data capture tool, gathering relevant performance and capacity metrics from SMF/RMF records.

It’s also child’s play to implement for anyone with a zOS background.

Our Athene for z/OS offers a range of features to enable you to analyze z/OS performance across your entire estate, report and create trends of behavior and analytically model future scenarios to optimize configurations and expenditure.

Athene allows an enterprise to optimize on-going capacity, minimize over-spending on hardware, avoid the costs of performance crises, guarantee service levels, and alert on trends to know when a capacity issue will begin to emerge.

Our Athene for z/OS modelling capabilities help you to avoid potential performance problems and evaluate tuning and management strategies well in advance, not in crisis mode.

We’re offering one FREE z/OS Capacity Audit and Projection service per organization for a limited period.

To use the download to collect and return data for a FREE z/OS Capacity Audit and Projection visit our Downloads section www.metron-athene.com/_downloads ,

Andrew Smith
Chief Sales & Marketing Officer