Restarting Green Data Center Blog v2.0

12 years ago I started blogging on green data centers. A that time no one was carbon neutral. PUE was not yet a standard. I thought greening the data center would make things better. 2 years ago I got bored and reduced my writing. Why? Because being carbon neutral and more efficient did not change data centers and it was getting boring. Same old stuff.

Taking a break I got more into construction issue. Then found how bad wireless was constraining innovation. And there were so many other things interesting going on that was not being covered at conferences and data center news. These entities also found the attendance and readership was falling and they added more things. Cloud, SaaS, IOT, and now Edge. But that stuff is boring and irrelevant to the core DC crowd that are where so many of friends work in.

One of my good friends who runs a big data center site and has worked for two of the big companies asked a great question 8 years ago. “How do you do PLM (product lifecycle management) in a data center?” This is something that our departed friend Olivier Sanche would be asking as well.

Over the past 8 years I have tried a variety of approaches to solve the problem of how to do PLM? One of the top problems is few people want to solve the PLM problem, but the benefits are huge.

What is PLM? Wikipedia has this post.

In industry, product lifecycle management (PLM) is the process of managing the entire lifecycle of a product from inception, through engineering design and manufacture, to service and disposal of manufactured products.[1][2] PLM integrates people, data, processes and business systems and provides a product information backbone for companies and their extended enterprise.[3]

So do you want the complete view? You can think of a data center which runs huge amount of services as a product.

I was deep into the technical issues, but nothing works. Why? Because the approach is wrong. There was so much data in silos that was hidden. Or information was not being collected. Groups would not work together. Groups would optimize for their own ways. You could say do the “Toyota Way” but few industries have been successful adopting the Toyota Way.

How to solve a hard problem that is like a big puzzle problem?

A couple of weeks ago I saw a Smithsonion Channel video on Bletchley Park - the Code Breaking’s Forgotten Genius. This video is an hour long and know most of you will not watch this.

A 3 minute video that may be more interesting is on how this same Genius paved the way for the Cloud.

I have much more to write on what it takes to green the data center, and what can be done beyond carbon neutral and a low PUE>

Dell Announces Wireless shielded Rack Enclosure

Electromagnetic Radiation is a complex topic that is heavily regulated, but not necessarily in the data center. Users are figuring out though that there are times when you want to protect what runs in the racks from what gets out and what can peak in.

For the past few years I have had fun connecting with old, old, old EMC (Electromagnetic Compliance) friends who I worked with 30 years ago when I was at Apple. These are friend I worked with to ship a wide range of Apple hardware products that needed to pass FCC and other government regulations. These guys were magicians in figuring out how shielding, cabling, and other changes could be made. This is stuff you do not see as it is hidden inside the box.

Dell has a cool rack they just announced for addressing electromagnetic radiation of server mounted gear.

This not easy to do, but I believe the Dell team has heard the customers and created a solution that customers have been asking for.

Operations IT, a new role for IOT

Stacey Higginbotham on StaceyOnIOT posts on a new role in IT for Industrial and and Enterprise IOT.

As technology comes for more and more of the corporation, from conference rooms to manufacturing plants, there’s a culture clash brewing between the IT department and the folks responsible for operations. Because while those operations might be focused on maintaining the safety and comfort of those conference rooms or managing the safety in that manufacturing plant, when everything gets a computer and a potential connection to the internet, it all becomes part of IT.

In this post Stacey highlights a problem in operations of a BASF plant with WiFI out for 11 months to support the sensor network. Stacey expresses a sense of shock that this happened.

But they didn’t replace it. Odom said the plant has now been without Wi-Fi, and thus this particular system, ever since the IT team took it down 11 months ago. Recently, the plant experienced an equipment failure that cost “something in the six figures,” which would have been detected by the vibration sensors had they been operational. Odom said that her notes to the IT department had not had any effect, but she hoped that the equipment failure — a clear example of how the lack of Wi-Fi was costing BASF money — might inspire a bit of urgency.

That is a particularly egregious example of a corporate IT department not communicating with an operational tech team. But while I was stunned by the idea that a section of a plant might be without Wi-Fi for 11 months, most of the other attendees didn’t seem surprised. As to how anything got done at all, there were other Wi-Fi networks in the plant — just not ones this particular system could use.

This is easy to see happening as the the corporate IT department wanted to show it knows better and choose to shut down the risky hardware. IT addressed a risk, and put the replacement project in the queue to be worked on when they have budget and time.

Would this situation be prevented by the role of Operations IT? Not necessarily. This is why almost all the big data center groups have their own IT group for their operations which is separate from the enterprise IT group. Operations has their own budget and resources to work on the problems to keep their business running. Running office space for enterprise staff is so different than operations that needs to run beyond an 8 hour day and has a direct impact on the business when there is an outage.

More Than 3 Use cases of IOT, building on StaceyOnIOT's post

On Aug 19, 2019 Stacey On IOT posted on three different version of IOT. Where she discussed Consumer, Enterprise, and Industrial use cases. I was going to post, but did not. Stacey followed up today with another post with user feedback and the idea of adding two more use cases for Military and Healthcare.

Across these now five different use cases what is common approach that could be used? An approach that have been using to look at some other systems is a combination of hexagonal architecture, ports and adapters and domain driven design. Ok. Lost all of you there.

Hexagonal architecture is

The hexagonal architecture divides a system into several loosely-coupled interchangeable components, such as the application core, the database, the user interface, test scripts and interfaces with other systems. This approach is an alternative to the traditional layered architecture.

Ports and Adapters is

The main idea of Ports & Adapters is to define the structure of an application so that it could be run by different kinds of clients (humans, tests cases, other applications,…), and it could be tested in isolation from external devices of the real world that the application depends on (databases, servers, other applications, …).

Domain Driven Design is

Domain-driven design is predicated on the following goals:

  • placing the project's primary focus on the core domain and domain logic;

  • basing complex designs on a model of the domain;

  • initiating a creative collaboration between technical and domain experts to iteratively refine a conceptual model that addresses particular domain problems.

OK still lost you. The use cases of consumer, industrial, enterprise, military and healthcare are domains that are at the center of the architecture. The interaction with IOT and users is what drives the ports and adapters to the domain core, then gets passed on to ports and adapters for the infrastructure. This same system design can treat of the different uses as simply different iterations of the same design approach.

This is a complex topic, but can be elegantly simple. But explaining the simplicity takes a lot more words. :-)