Blog Posts

Recall in our previous posting, we introduced the notion of an Enterprise Cloud, comprised of a combination of private, public, and hybrid cloud services.  Many enterprises will want to exclusively offer private services (meaning they are operating, by definition, a “private cloud”) for a variety of reasons (a topic of a future posting in this series).  However, more realistically, and more commonly  moving forward, IT will need to manage all three types of services.

Thus, the question is that of how IT wants to approach the design and management of the mixture of cloud services in its environment – whether the service types are blended together in a single cloud offering, or separated.  There is no single right answer to this, as it’s entirely dependent on the organization’s priorities, infrastructure, needs, and context.  Nonetheless, it’s clear to me that there’s a need and a demand for an overall management  platform that can handle all three variety of services in a unified fashion.

For example,  imagine that you’re a developer who needs a server OS temporarily provisioned for some development work. You make the request, and a few minutes later you have a new VM, which you access across the network via some sort of remote login. Does it *really* matter where that VM physically resides?  Much of the time, it doesn’t…it can just as easily and productively be in a remote cloud as in the server room down the hall.  That is, the policies and factors that determine where this particular VM is provisioned should be embedded in and executed by the management platform, which can make the best placement decision at the time of request.  There would need to be a compelling reason to instead have the requester make this decision, by explicitly choosing which cloud platform to use.

A related question, applicable to organizations operating only private cloud services, is whether to manage one cloud environment, or have distinct cloud infrastructures to serve different purposes. The answer is that there is no single answer…people will need to balance their technical and business needs with the financial, technical, and organizational constraints, and make an informed decision that’s right for them.  Nevertheless, I think that many organizations will want the option of managing disparate service types, as well as all their clouds, from a common management platform.  Even if an organization is running multiple, distinct private clouds, it makes sense to embrace a common management platform, to leverage shared policies, processes, and people.

Our next topic in this series – using external Cloud Service Provider offering definitions as a model for internal cloud service types.

Thank you to those of you who were able to make the webinar yesterday, Automate the Delivery of Business Critical Events and Improve Response times by Over 40%, we had a great turnout. For those that were unable to attend, the recording is now available: Download Now.
In the webinar, we discussed how AlarmPoint and BMC automate the process of targeting, delivering and escalating alerts to the person who can best resolve them enabling improved response times accountability and audibility, Covering how organizations can:

• Automate dispatch and incident assignment processes
  • Reduced time to event ownership on average of 85%
• Reduce Mean Time to Repair
  • Cost reduction of up to $1,000,000 / yr by reducing the notification element of incident resolution time by an average of 80%.
  • Increased SLA compliance by reducing time from ticket creation to ticket close by over 40% increasing margins in the upwards of $800,000/yr.

• Eliminate unnecessary calls into help desk
  • Proactive notification saves millions of dollars in manual callouts and drastically reduced calls into helpdesk

  For those of you that missed it, you can now watch a recorded version of the Webinar online. If you have comments or questions, contact your BMC Sales Rep or email sales@alarmpoint.com.

A Bad Cloud, eh?  What in the world is a Bad Cloud?  How can I avoid one, if they even exist?  And for that matter, that brings me back to the original question – Is there such a thing as a Bad Cloud?

In a nutshell?  Yes.

But before we answer the other questions as to what is a Bad Cloud and how to avoid one, let’s give some background behind how a Bad Cloud forms.

Cloud often means different things to different companies.  However, most commonly across progressive organizations, the term describes a visible and executive-sponsored initiative.  If a project is part of a cloud initiative, then it probably also has a better chance of being funded than other projects in the IT realm.  Sure, this isn’t always true, but the point I’m making here is that the positive power of cloud is significant. 

Part of the reason that cloud projects are so popular is simply because buyers are ready to implement clouds, and therefore to build or buy cloud-focused products.  The strong support for cloud is also influenced by the sheer hype around the topic that exists in the industry.  Yet another factor is because it is the most commonly understood concept that unifies many of the solutions that companies have been developing across the Enterprise for years.  And, the list goes on…

Cloud is exciting. Cloud is the Cool Nerd in a teen movie. Cloud reeks of innovation and the cutting edge. But the cloud boost factor (the appeal attributed to cloud for the aforementioned reasons) sometimes causes otherwise intelligent decision makers to make poor choices, and that’s when a Bad Cloud rolls in.

A Bad Cloud can be the result of something as simple as rushing to implement a private cloud without understanding the proper controls of that environment. For example, no change management or tracking of virtual machine provisioning can cause the underlying virtual infrastructure to grow unchecked (commonly known as sprawl).

Luckily, such a situation can be easily remedied with the proper tools and processes put into place. However, a Bad Cloud can also be represented by a poorly formed cloud strategy that encourages employees to use public clouds for any short-term projects and/or to aggressively try new ideas in the public cloud. Such unbridled enthusiasm without foresight can unintentionally cause placement of confidential data – like customer credit card numbers or employee Social Security numbers – into a public cloud.

So, what can you do to avoid a Bad Cloud?  Like with all major decisions, use common sense – think through what are your core needs and stick to addressing those first. Use industry leaders and market sources to help you understand where are the pitfalls. And purchase the right products and solutions – ones that will not only add automated execution to your cloud environment, but also the right governance and controls to keep your good cloud from going bad.

zIIPs. I am sure you have heard about them and how they can save you money.

They can.

But before you jump into the zIIP world, weigh the costs and benefits. Consider whether you can offload MIPS during peak processing times (high return on investment) or during off-peak hours (not so high ROI).

Choose vendor software that can help with offloading work to zIIPs. Be sure that the vendor is providing software that is compliant with IBM standards for specialty engines.

For more information on zIIPs, see this Q&A with Gary Salazar, BMC IMS architect.

Tell us about your experiences with specialty engines.

By popular demand, here's that link to the video we used to launch Mobile BMC Self Service at last week's WWRUG: http://www.youtube.com/watch?v=PRzeyMOjnl4.

Thanks for all the great feedback on the new product and the mobility discussion with Doug Mueller! It was a reminder of why the Remedy community is so vibrant after so many years.

And of course thanks to Daniel Bloom, Phil Bautista, Lenny Warren, Joel Sender, Kelly Deaver, and everyone who put on an excellent event!

See you next year!

-Dan and the Aeroprise team

Guest post by Mike Sniezek, Product Manager

Can you imagine a computer that could run multiple operating systems and share those resources when and where they were needed? A computer that has virtually unlimited computing and real memory resources and that can be sliced and diced into customized environments? One that is proven to be impervious to computer viruses and be truly available 24 by 7 even when there is a hardware failure? This greenest machine on the planet is known by all as a mainframe.

Unfortunately, many in the world of technology have a picture of a colossal piece of IRON spanning entire office floors, manned by dozens of workers huddled over monitors. The truth is you would walk right by a modern mainframe or try to open it thinking it was a refrigerator.

The mainframe has become a server for Linux and Java as well as z/OS.  When I hear of the needs of cloud computing, I ask “Has that not already been done?” The mainframe is a mega-server whose power grid consumption is for computing, not cooling. I have seen the blade computing centers and server farms...what I saw was a big room big room full of machines with dozens of technicians huddled over monitors...go figure.

We have heard a lot of stories about the experienced mainframe workforce shrinking because people are retiring, and we generally think of the retirees as technicians. But what happens when the data center director or CTO retires? A person who did not grow up on the mainframe often gets promoted into that position of authority.

Mainframe novices are often intimidated by the mighty mainframe, and executives are no exception. Making the mainframe friendly to executives with no mainframe background is tricky.

At first glance, the mainframe appears to be costly in terms of floor space, software, and personnel. Perhaps an experienced mainframe DBA gets a higher salary than a less-experienced distributed systems DBA, but the mainframe DBA is likely to be managing terabytes of data – mission critical data that the day-to-day business depends on.

Recently, I heard a director from a large US company lamenting the fact that his executives were anti-mainframe. This attitude exists despite that fact that his environment of 130,000 MIPS which runs most of their business is being maintained and supported by a team of only 60 technicians.

The mainframe may not seem as exciting as a server, but it is definitely faster and more secure. And the sheer amount of transaction processing power is exponentially higher for a mainframe. And the mainframe invented virtualization!

Executives may take the mainframe for granted because the mainframe sits there and does its work without a lot of fanfare. As long as everything goes as planned, the executive may not even think about the mainframe until a software bill is due. In the unlikely event of an outage, the executive sees the impact of the mainframe on the bottom line.

Share your stories of how you have introduced executives to the mainframe. I’m sure you have some good ones!

Download Audio (13:27)

"We have to be very careful about confusing measuring things with trying to immediately improve them," says Colin Fletcher.

A universal law of business states, “What gets measured gets the attention.”  IT must ensure that the right metrics are used as key performance indicators (KPIs) so that appropriate actions are taken by IT management and IT staff to meet the targets set by these KPIs.

How long have you been measuring the same thing and getting the same results? Does what you're measuring result in positive change in the business? Have you evaluated what you're measuring against what your customers want? How well are you meeting your improvement goals? Have you taken time to dream about what your business would be like if you could do anything?

Join us as we talk with Colin Fletcher, Solutions Marketing Manager, BMC Atrium at BMC Software, about Developing Metrics that Matter. Colin shares insights and tips about how to develop metrics that matter in your IT environment.

Bio

Colin Fletcher is the Solutions Marketing Manager for BMC Atrium at BMC Software, where he serves as a trusted advisor, evaluating business
and technical challenges from a BSM perspective.

Questions

1. Everyone wants IT to be a dependable, high-quality service to business users at a reasonable cost. But all too often it's not that way. IT groups are often overwhelmed with all the things they have to do. So at what point do you start? Are key performance indicators (KPIs) a way making this whole measurement thing a win for the IT group?
   
2. Do you think these days it's useful for IT groups to start from scratch on that perspective as a refresh?
3. You have said that developing KPIs is similar to working inside out. Can you expand on that
4. What is the difference between the perception of what is keeping people from achieving their goals, and the reality?
5. Is there potentially greater value in taking a broader, deeper look at what you're basing your KPI on, beyond the six-sigma approach, and beyond how accurate and how fast you've done things?
6. What is the value in establishing KPIs incrementally? Can you talk a little about what that would look like in an organization?
7. How do IT service management solutions fit into this picture?

Resources

BMC Industry Insight: Developing Metrics That Matter (PDF)