• Home
  • Who Am I ?
    • Contact
    • What does Ethereal Mind mean ?
  • Disclosure
    • Disclaimer
    • Comment Policy
    • Privacy Policy
  • Just Three
  • Archive

EtherealMind

Software Defined & Intent Based Networking

You are here: Home / 2015 / Archives for August 2015

Archives for August 2015

Response: Nerd Knobs Are Symptom, Not Cause.

28th August 2015 By Greg Ferro Filed Under: Blog, Response

etm-response-logoIvan makes a very good point that using nerd knobs is bad for your business. He is perfectly correct that too many people do it because they can and without careful consideration of the technical debt incurred.

I would like to add a few more things about Nerd Knobs:

  1. Nerd Knobs are a symptoms of deeper failings in networking. In my view, routing protocols are losing relevance in a multi-path, multi-network world of 2015. Or because routing protocols converge slowly & poorly. We use NATv4 instead if IPv6, PBR instead of SDN and so on.
  2. Nerd Knobs are what happens when vendors give customers what they ask for instead of what they need. Vendors recite the mantra of giving customers what they want. After all that is where the money comes from. But customers are ill equipped to ask for the right things and their businesses want to build a 20-storey building one floor at a time without paying for the foundations or plumbing.
  3. Poor leadership – most IT managers have lost the ability to recognise technical debt and its impacts. The focus on process, ITIL and operations combined with vendor dependency/outsourcing for architecture & validated designs has de-skilled manager decision making competency.

Nerd Knobs are symptoms of much deeper problems/technical debt in the networking market and treat the cause not the symptom.

Summary: Headache tablets treat a headache and are cheap & easy to get – like nerd knobs. Get a good network architect if pain persists. You might have migraine or a flawed network technology.

Why I Hate ITIL So Much

26th August 2015 By Greg Ferro Filed Under: Blessay

In the middle of 2014, Justin Warren  sent me a series of interview questions on the topic of why I hate ITIL so much. The process of writing a response got me … um, motivated and spat a substantial amount of bile.

Here it is. why-i-hate-itil-opt

Why do you hate ITIL so much? How did you arrive at this point of view?

Over the last decade I have worked for tens of companies that use ITIL/ITSM models and all of them are were miserable and unhappy workplaces. When I have worked in companies that don’t use ITIL (and there are a few) I found they were a great places to work where real business value was being created and delivered.

Some years ago I took a step back and made a conscious decision to avoid working for companies that asked for ITIL, ITSM, PRINCE2 and TOGAF experience and my work happiness improved enormously.

It’s about happiness. ITIL = misery and unhappiness. ITSM is operational misery. Who wants that ?

ITIL was originally developed quite a while ago, to address a specific problem. Does it contain anything of value?

Like all business or consulting frameworks, I feel that ITIL has reached the end of whatever useful purpose it might have had. Six Sigma, Total Quality Frameworks, ISO9001 are just a few examples of other frameworks that have passed their useful life and forgotten. Its time for ITIL to die too.

The original purpose of ITIL was to provide a way to manage technology that incumbent managers and executives could not understand and that generation has moved out of the way through retirement or forced engagement.

ITIL contains nothing of value for the future. The fundamental ITIL premise is that technology work can be segmented like machines or work functions in a factory where each task can be assigned to a machine with fixed human resources applied to the task and funding applied to the machine. The concept is valid but simply doesn’t work in real life when the “factory machines and processes” undergo transformation change every three to five years like steam to petrol engines.

In a factory, a machine will last decades before replacement or change is needed. ITIL aping that model has never really worked for IT.

What problems does ITIL cause?

ITIL/ITSM has prevented enterprise change for about 15 years Public Cloud exists because of failed ITIL models that it prevents convergence, new technology adoption and operational flexibility. I’ve worked at public and private cloud deployments and the failed deployments are all due to ITIL practices preventing change. The cloud transformation destroys value in existing operational models at every level and puts ITIL & ITSM in the trash.

Everyone should experience in an ITIL-free project to understand just how enjoyable and exciting it is to deliver something on-time, on budget.

Overinvestment in Operational Control ITIL is about control, scope and reporting. It’s about minimum viable delivery, lowest possible compliance, least effort, review meetings, and other short sighted activities.

ITIL is not about delivery or excellence. In my experience, ITIL and PRINCE2 prevent excellence through focus on deliverable and cost management.

Take a look around your ITIL-compliant project team and count the secretarial staff (aka project secretaries/managers) doing the reports, raising purchase orders, scheduling meetings, calculating budgets and indulging to “project programme review” instead of delivering useful work on technology.

I once worked on a large project that had more than 25 ITIL/PRINCE2 project managers and only five engineers. Each and every one was convinced that if only we were better managed then productivity would improve. I pointed out the one more engineer would increase output by 20% and they pulled out their reports to deny that fact. I quit a few days later.

Responsibility abandonment – The single biggest problem in ITIL the creation of silos creates “not my problem” attitudes. Workers have no ownership, no incentive and are actively prevented from communicating outside of their budget allocation or silos. This leads to technical debt and cost wastage as teams maximise their performance at the expensive

Poor communication – the creation of silos forced teams to form around narrow disciplines and created structures for leadership and work segmentation. At the same time, it creates villages of isolation where like-minded groups formed around core disciplines and enabled self-referencing and self-reinforcing practices to emerge. For example, storage professionals are often criticised for their inability to adopt new technology but this is a function of silo isolation where large-scale change is not rewarded or allowed.

Poor Management Practices Like any group of people, there are many managers and executives with poor skills and limited competency. Frameworks like ITIL create the illusion of control that results in organisational delusion that IT is efficient and manageable.

What should be done instead of ITIL?

Information Technology is moving to an integrated or converged model at many different levels. Technology disciplines around network, storage and compute are no longer isolated but integrated as cloud, both public and private, replaces the existing service models.

Change is being led by converged infrastructure in the data centre but extends much further – smartphones and tablets are replacing desktops, the death of 3-tier software, the return of developers in the enterprise and much more. The transformation is very wide.

The cloud transformation destroys residual value in existing operational models. ITIL & ITSM are fundamentally about limiting scope, silos of functional separation and change prevention and cannot cope with the amount of change in the next few years that will be driven by convergence.

The future isn’t yet clear but the most viable models today are Agile / Kanban, Devops, Scrum and Sil0-busting teams.

Aside: What is most interesting is that these models are what ITIL was intended to “fix”. By preventing flexible and adaptive work, ITIL wanted to make the workforce into predictable robots. Of course, this feeds the denial by the ITIL cartels.

What do you think of DevOps?

Is it a valid replacement for ITIL? Or is it complementary? Should both be thrown out in favour of something else?

DevOps is, in large part, a reaction to the failure of ITIL and very much about finding a working solution to ITIL’s unsolvable limitations. It removes silos through cooperative activity where each persons commits to delivering something to their peers. It promotes community and communication where ITIL prevents and restricts.

Younger people are especially able to embrace DevOps since they lack indoctrination of the older people like me. This, of course, leads to incumbent “ITIL experts” telling us that they don’t have experience or don’t know how things work. Older people should have the wisdom to see an approaching wave of change.

in terms of replacing ITIL, we already know that this will happen. Public cloud companies like Google and Facebook clearly demonstrate the failure of Enterprise IT and ITIL/ITSM. The fiscal cost of ITIL will be discarded to compete internally. Here is the secret – public clouds are cheaper because they don’t use ITIL.

What follows ITIL isn’t yet clear. Human nature will be to develop yet another framework and the big consulting companies are probably working on it already. I see that it will be based on DevOps that we see today, be empowering to technology practitioners to promote teamwork, ownership and outcomes instead instead of isolation, silos and poor responsibility handling.

What about organisations that have spent a lot of time and money implementing ITIL practices and claimed success? What should they do now?

Find a small project in the portfolio and use Kanban or Agile methods. Get some contractors with Kanban/Agile experience to enculturate that team. The hardest part Agile methods is that employees must become responsible for their output and empowered to deliver. The “factory” model prevents exactly that.

Take this seed team and embed the practices into other projects. Evaluate strongly and ruthlessly push staff into changing their personal outlook to professional responsibility. In IT infrastructure, nothing should be a network problem, a storage issue, or developers failure. Everyone owns the platform, everyone owns the solution.

Is there something that ITSM and DevOps practitioners are both missing?

DevOps practitioners are busy with delivery and execution at any price while developing their own control/measurement systems through the use of Scrum/Agile methods. DevOps doesn’t have a rigid and inflexible framework that costs money to use nor does it need expensive consultants to advise. They simply don’t care about ITIL or ITSM.

ITSM practitioners are made obsolete by the fundamental convergence technology trend. The ITSM process will be replaced by fluid process defined by software application applying concepts of Kanban or Agile methods. In a very real sense, ITSM will be replaced by applications like products implementing Kanban/Agile like Atlassian, LeanKit and Asana. Technology practioners are smart people and often much smarter than their leadership or business managers.

ITIL takes the view that technology people are too smart for business which requires rigid control to prevent overactivity or over-innovation. DevOps takes the view that the business grows if smart people are managed smarter and will unlock genuine business advantage.

Five Action Items For IT Infrastructure Architects If the Global Economic Slowdown Starts To Bite

25th August 2015 By Greg Ferro Filed Under: Blog

There are signs that the global economy is slowing down. 5-action-economic-slowdown-optThis week, Shipping rates1 from Asia to Northern Europe fell by 25% last week. The stock market in China had sharp falls2. Europe is still wrangling its hand over the Greek bailout and the American economy is starting to get worried because who needs American Dollars to trade with. The IMF3 has been predicting a rout in US stock prices and a decade-long rut for the global economy.

And this week, the US stock market got the wobbles:

Stock markets down 1

 

 

 

 

 

What does this means for IT Infrastructure and Networking specifically ? What can you do to make the most of the time when capital is limited, deployments stop happening and you need to improve efficiency.

Takeaways

  1. The global economy outlook isn’t good, markets are nervous and signs of widespread slowdown are everywhere.
  2. Global companies feel global impacts. Slowdowns in Asia are particularly damaging to as vendor profits and growth was predicted last year.
  3. Action: Delay your spending and fight harder and expect to get discounts and deals.
  4. Work on automation and orchestration project using existing resources
  5. Start planning upgrades and maintenance that will sweat the assets and generate savings for your employer.

1. Get Discounts and Blow the Budget Early

If the global economy continues its decline then IT customers will stop investing in infrastructure as budgets shrink. Vendors will get increasingly desperate/keen to maintain sales volume, to clear stock in the warehouse and make their numbers. In a shrinking economy, stock reductions will improve cash flow and reduce exposure. Sacrificing profits is acceptable and reductions in selling price to close deals.

On the other hand, your own employer may start cutting budgets early to pull back costs. You need to pick the moment to blow the budget before the cuts start. You will need to exercise your white collar skills to get the trigger points just right.

2. Shrinking Inventory and Spare Parts

Reducing stock reduces costs. Vendor will reduce production runs to reduce outlay and reduce warehousing costs. Distributors will reduce their stock holdings to manage cash flow.

And don’t forget that spare parts inventory which will be under pressure to reduce the amount of money sitting on the shelves. Vendors often squeeze too hard on spare parts which can impact service inventory.

3. Plan Upgrades, Maintenance and Automation Work

Your own budgets are at risk of cuts as your employer will be affected to and focus will shift on sweating the capital already spent.

  1. Start thinking about upgrades instead of replacement.
  2. Look at your operational capability – it could be time to work on new processes, procedures.
  3. Work on your software-defined capability – its time to work on monitoring systems and works on the orchestration/automation of your existing hardware. Change projects from “installation” of new products to “deploying software-defined” which will be simpler when the

It might be possible to replace the endless (and mostly pointless) ITIL Project meetings with time spent on automation projects. Time to get started with the Ansible framework for task automation and start a Proof of Concept for bigger SDN projects in 3 to 5 years time.

4. Resellers will Struggle

During a slowdown/recession, the middle man is always squeezed hardest and resellers may struggle to maintain their cashflow. Monitor your reseller for signs of financial stress such as invoice factoring, repeated payment requests, stressed engineers/experts with cheap laptops.

In reseller-land, sales slow down but costs stay the same.

The EtherealMind View

There are no guarantees that an economic slow down will happen. In the last few days, the market has rebounded from a near 10% drop but remains lower of the year.

The effects I discuss here will take some time, perhaps six to twelve months, to impact but its worth thinking about in advance. After all, your employers financial performance is what determines your future pay rise.

Reference


  1. Asia-Europe Container Freight Rates Have Fallen Off a Cliff – gCaptain ↩
  2. China’s downturn is not a crisis – Business Insider ↩
  3. Global Market Routs Came With Ample Warnings – Real Time Economics – WSJ ↩
  4. China’s stockmarket: China’s stockmarket keeps plunging but cooler heads prevail | The Economist

Response: Snappy for Whitebox Switches | Ubuntu Insights

24th August 2015 By Greg Ferro Filed Under: Response

Ubuntu announces that it wants to operate in the IOT space including network switches and devices.

“Traditional switch manufacturers and the disruptive new entrants both want to move to software-defined solutions, and Ubuntu Core offers a well-defined neutral core that supports the widest range of hardware at the best price,” said Dustin Kirkland, product manager for Ubuntu Core and Ubuntu Server at Canonical. “With Ubuntu Core on the switch, customers choose their preferred hardware and their preferred switch control software independently, and they can extend the functionality of the switch with amazing apps.”

Couple of interesting data points here.

  1. There are two Linux distributions that matter – RedHat & Ubuntu.
  2. RedHat is popular with Enterprise and Service Provider Markets.
  3. Ubuntu is popular with new generation application/cloud companies. For example, it is the most popular distribution on Amazon EC2/AWS.

What I don’t see clearly, is whether an Ubuntu-sponsored version for networking makes any sense.

  1. There are specialist commercial Linux distributions developed by a number of vendors that offer specific capabilities.
  2. Most enterprises would prefer to use RedHat/Fedora based distribution
  3. I recognise that Debian/Ubuntu is popular with cloud providers and that using Ubuntu on network may be attractive.

What do you think ?

IoT World: Snappy for Whitebox Switches | Ubuntu Insights:

Response: Arista EOS & Quality

24th August 2015 By Greg Ferro Filed Under: Response

This video from Ken Duda at Arista is, perhaps, the best explanation of Arista’s success with customers. As an engineer, I found this talk inspirational.

  1. No bonuses for hitting ship dates. This avoids “good enough” code getting shipped.
  2. Sure there are money problems associated with this but Arista believes quality is better.
  3. You write the code you own it. No escalation, no process between technical and developers.
  4. Culture of quality is pervasive and “backbone of a high quality product”
  5. Get sales to buy into quality instead of features and ship dates.
  6. Arista uses Linux but keeps all their code in user space processes. (compare this with Cisco’s NX-OS)
  7. EOS architecture of SysDB is key competitive advantage.
  8. There is no Quality Assurance team because QA doesn’t

I got to watch about half of the videos from Network Field Day  10 last week but this presentation is outstanding. Two key takeaways for me:

  1. You can build a network company with a culture of quality and be successful.
  2. Software architecture and process is competitive advantage compared to legacy vendors.

 

  • 1
  • 2
  • Next Page »

Network Break Podcast

Network Break is round table podcast on news, views and industry events. Join Ethan, Drew and myself as we talk about what happened this week in networking. In the time it takes to have a coffee.

Packet Pushers Weekly

A podcast on Data Networking where we talk nerdy about technology, recent events, conduct interviews and more. We look at technology, the industry and our daily work lives every week.

Our motto: Too Much Networking Would Never Be Enough!

Find Me on Social Media

  • Facebook
  • Instagram
  • Linkedin
  • RSS
  • Twitter
  • YouTube

Return to top of page

Copyright Greg Ferro 2008-2017 - Thanks for reading my site, it's been good to have you here.

Opinions, Views and Ideas expressed here are my own and do not represent any employer, vendor or sponsor.Full disclosure