JANWIERSMA.COM

Cloud & Datacenter strategie.

My move to SDL…

Posted by Jan Wiersma op 22/12/2013

It’s the weekend before the holiday season and just like last year I find my self at an US airport making my way home… just in time for Christmas.

Sitting at the airport lobby, listing Christmas songs, I can’t help to reflect on the past year.

A lot has happened and things changed a lot. I have left OCOM (LeaseWeb/EvoSwitch/Dataxenter) after 2 years in September this year. Something that some of my peers in the market didn’t expect, but was long overdue. For too long I couldn’t identify myself with the way the company was run and its strategy. No good or bad here… just a big difference of opinion on vision and execution.

The last 2 months I have been able to talk to lots of different organizations in an effort to see what my next career step should be. I needed some time to recuperate from my little USA adventure with LeaseWeb & EvoSwitch. It was a great project to participate in but all the travel took a big toll on my personal life and me.

I also learned how passion for your work can be killed and what it takes to be sparked again. And how people are motivated by the Why in their jobs.

I had some good conversations with DCP board members Mark and Tim about my frustration in the lack of progress in the datacenter industry. It felt like I have been doing the same datacenter and cloud talks for the last 3 years, and things still didn’t move. I wanted to have the opportunity to really make a difference.

In November this year I joint SDL as the VP Global Cloud Operations. Again something most of my peers didn’t expect.

SDL’s services are based around language translation, customer experience and social media intelligence. Most of them are geared toward a CMO and its department. The company is making an aggressive play in to the PAAS/SAAS market.

Gartner has identified the potential of CMO’s moving in to cloud computing consumption publishing “The Top Four Impacts of Cloud Computing on Integrated Marketing Applications “.This is based on the prediction that in 5 years from now the CMO’s will spend more then most CIO’s.

Running this PAAS/SAAS environment requires a large-scale infrastructure that is able to handle high volumes of traffic and data storage. Being able to deliver these services on a worldwide scale, with the ability to scale-out, don’t only require technical transformation but also process & organizational.

During my conversations with SDL’s leadership (and specifically its CTO) their vision around infrastructure and service delivery lined up perfectly with mine. It encompass everything I worked on the last 5 years and the associated vision I evangelized for many years; DevOps, Kanban, Infrastructure as a Service (and the true meaning of the ‘service’ part), BigData, etc..

The next months and years I hope to execute on the companies cloud vision and build the needed reliable & scalable infrastructure and a world-class team to build & support it.

In the coming months I will share more information on SDL’s journey.

Have a great Christmas!

Posted in Overig | Leave a Comment »

CIO’s plukken vruchten van server innovaties

Posted by Jan Wiersma op 21/12/2013

Op de afgelopen CIO dag (Nov 25&26) in Amsterdam werd ook ‘The big data center book’ uitgereikt. Hier in werden de Data center & cloud trends voor 2013 & 2014 toegelicht.

Mijn artikel in dit boek beschreef de trends in server hardware innovaties;

Naarmate onze datacenters groter worden en de eisen van webscale- en cloudproviders de markt overnemen, worden servers meer en meer slechts een component van een grotere machine. Hoewel componenten waardevol kunnen zijn, zijn zij niet langer het hele systeem en als zodanig kan hun waarde niet los gezien worden van het datacenter waarin zij gehuisvest zijn. De efficientie van het component wordt daarmee belangrijker dan het bezitten van ‘coole features’. Wat zijn de actule ontwikkelingen die moeten zorgen voor een lager energieverbruik van servers ?

Het volledige artikel is hier te vinden.

Posted in Cloud | Leave a Comment »

EE-IAAS – call to join IAAS energy research

Posted by Jan Wiersma op 20/11/2013

seflab

The cloud market is hot. The IAAS market sees a lot of growth and new IAAS providers seem to entering the market on a daily basis. Enterprise IT shops are exploring the usage of public cloud solutions and building their own private cloud environments.

IAAS distributions like OpenStack and CloudStack seem to have thriving communities.

Organizations that are successful in deploying IAAS solutions, either for customers or internal use, see rapid growth in demand for these services.

IAAS services still need IT equipment to run on and datacenters to be housed in.  While there has been a strong focus on making the datacenter facility and IT equipment more energy efficient, not much is know about the energy efficiency of software running these IAAS services.

The Amsterdam University of Applied Science (HvA) lanced an energy consumption lab focused on energy efficient software called SefLab 1,5 years ago.

Several EU Datacenter Pulse (DCP) members donated time and equipment to the research during the launch.

During the start of the SefLab students and researchers got familiar with the research facility and its equipment. They did a webbrowser compare on energy usage. (results on the SefLab website)

Now that the lab is operational, its looking for new area’s of research. One of the focus areas is the development of IAAS distributions like CloudStack, OpenStack, etc.. and their energy efficiency; what are the effects of architecture choices?, is their a difference in energy efficiency between the distributions ?, what power-manager & reporting elements are missing from the distributions ?

The HvA and SefLab are looking for:

  • IAAS distribution users willing to participate in the research. This is possible in several ways ranging from active participation in formulating research questions, managing a subproject, and supervising student projects, to participating in workshops and events of the project.
  • IAAS distribution communities willing to participate in the research. Future development of IAAS energy management components can be tested and validated in the SefLab facilities.
  • Vendors willing to support the effort with knowledge transfer and hardware/software donation.
  • Industry groups willing to support the effort in knowledge transfer network.

The research done by SefLab is really end-user driven, so this is a natural fit with DCP’s strategy.
Details can be obtained by contacting jan.wiersma@datacenterpulse.org

DCP members will be updated on the projects progress using the LinkedIn DCP member group.

See full call for participation here.

Posted in Overig | Leave a Comment »

The need for datacenter education in emerging countries.

Posted by Jan Wiersma op 28/09/2013

<Cross-post with my DCP blog>

I recently had the privilege to travel to some of the emerging datacenter countries like Eastern Europe, Russia, Middle East, Africa and some of the Asian countries.

While I was impressed with the deployment of mobile bandwidth & devices the conversations with datacenter owners and operators scared me. With more and more people getting internet access in these countries, the need for local datacenters also rises. Taking the time to really talk and listen to the stories of some of the local datacenter owners, it reminded me of the way datacenters were build and operated 10 years ago in most western countries.

Zooming in to the way these local operators and owners gather their knowledge, two things stuck me:

  • Most of them aren’t connected to industry groups or communities like Datacenter Pulse, 7×24 Exchange, Uptime Institute, etc… Sometimes this is a cultural or language barrier but most of the time I found this to be an awareness issue.
  • There aren’t as many industry conferences in those locations. It’s hard for conference organizers to get something going from a sponsor & revenue perspective. So where do local operators go to learn?

I think the larger datacenter vendors need to step up and take responsibility.
I think western datacenter owners and operators have a moral responsibility to go out and educate.

We need to send our Dr-Bob’s out and educate them on the benefits of retrofits of old facilities.
We need to send the industry leaders (either vendor or DC-owners) out and educate the emerging countries on the technical opportunities they have for the many greenfield projects begin run in those countries.

This way they know what technology is available before some of the old-fashioned datacenter consultants step in and copy-past our 10-year-old datacenter designs.
This way we protect them from mistakes we already made and they can leapfrog in to a brighter datacenter future.

I know this sounds idealistic be we owe that to these emerging economies, especially from a resource (like energy) efficiency perspective.

So if you have the chance to travel to emerging datacenter countries, take it. Share, lecture, educate, and try to protect the local operators and owners from making the same mistakes we made 10 years ago. They deserve that chance.

Believe me… it will make the world a better place.

Posted in Datacenter Pulse, Groen | Leave a Comment »

Where is the rack density trend going ?…

Posted by Jan Wiersma op 09/06/2013

<English cross post with my DCP blog>

When debating capacity management in the datacenter the amount of watts consumed per rack is always a hot topic.

Years ago we could get away with building datacenters that supported 1 or 2 kW per rack in cooling and energy supply. The last few years demand for racks around 5-7kW seems the standard. Five years ago I witnessed the introduction of blade servers first hand. This generated much debate in the datacenter industry with some claiming we would all have 20+ kW racks in 5 years. This never happened… well at least not on a massive scale…

So what is the trend in energy consumption on a rack basis ?

Readers of my Dutch datacenter blog know I have been watching and trending energy development in the server and storage industry for a long time. To update my trend analysis I wanted to start with a consumption trend for the last 10 years. I could use the hardware spec’s found on the internet for servers but most published energy consumption values are ‘name-plate ratings’. Green Grid’s whitepaper #23 states correctly:

Regardless of how they are chosen, nameplate values are generally accepted as representing power consumption levels that exceed actual power consumption of equipment under normal usage. Therefore, these over-inflated values do not support realistic power prediction

I have known HP’s Proliant portfolio for a long time and successfully used their HP Power Calculator tools (now: HP Power Advisor). They display the nameplate values as well as power used at different utilizations and I know from experience these values are pretty accurate. So; that seems as good starting point as any…

I decide to go for 3 form factors:

..minimalist server designs that resemble blades in that they have skinny form factors but they take out all the extra stuff that hyperscale Web companies like Google and Amazon don’t want in their infrastructure machines because they have resiliency and scale built into their software stack and have redundant hardware and data throughout their clusters….These density-optimized machines usually put four server nodes in a 2U rack chassis or sometimes up to a dozen nodes in a 4U chassis and have processors, memory, a few disks, and some network ports and nothing else per node.[They may include low-power microprocessors]

For the 1U server I selected the HP DL360. A well know mainstream ‘pizzabox’ server. For the blade servers I selected the HP BL20p (p-class) and HP BL460c (c-class). The Density Optimized Sever could only be the recently introduced (5U) HP Moonshot.

For the server configurations guidelines:

  • Single power supply (no redundancy) and platinum rated when available.
  • No additional NICs or other modules.
  • Always selecting the power optimized CPU and memory options when available.
  • Always selecting the smallest disk. SSD when available.
  • Blade servers enclosures
    • Pass-through devices, no active SAN/LAN switches in the enclosures
    • No redundancy and onboard management devices.
    • C7000 for c-class servers
    • Converted the blade chassis power consumption, fully loaded with the calculated server, back to power per 1U.
  • Used the ‘released’ date of the server type found in the Quickspec documentation.
  • Collected data of server utilization at 100%, 80%, 50%. All converted to the usage at 1U for trend analysis.

This resulted in the following table:

Server type

Year

CPU Core count

CPU type

RAM (GB)

HD (GB)

100% Util (Watt for 1U)

80% Util (Watt for 1U)

50% Util (Watt for 1U)

HP BL20p

2002

1

2x Intel PIII

4

2x 36

328.00

   

HP DL360

2003

1

2x Intel PII

4

2x 18

176.00

   

HP DL360G3

2004

1

2x Intel Xeon 2,8Ghz

8

2x 36

360.00

   

HP BL20pG4

2006

1

2x Intel Xeon 5110

8

2x 36

400.00

   

HP BL460c G1

2006

4

2x Intel L5320

8

2x 36

397.60

368.80

325.90

HP DL360G5

2008

2

2x Intel L5240

8

2x 36

238.00

226.00

208.00

HP BL460c G5

2009

4

2x Intel L5430

8

2x 36

368.40

334.40

283.80

HP DL360G7

2011

4

2x Intel L5630

8

2x 60 SSD

157.00

145.00

128.00

HP BL460c G7

2011

6

2x Intel L5640

8

2x 120 SSD

354.40

323.90

278.40

HP BL460c Gen8

2012

6

2x Intel 2630L

8

2x 100 SSD

271.20

239.10

190.60

HP DL360e Gen8*

2013

6

2x Intel 2430L

8

2x 100 SSD

170.00

146.00

113.00

HP DL360p Gen8*

2013

6

2x Intel 2630L

8

2x 100 SSD

252.00

212.00

153.00

HP Moonshot

2013

2

Intel Atom S1260

8

1x 500

177.20

172.40

165.20

* HP split the DL360 in to a stripped down version (the ‘e’) and an extended version (the ‘p’)

And a nice graph (click for larger one):

power draw server trend

The graph shows an interesting peak around 2004-2006. After that the power consumption declined. This is mostly due to power optimized CPU and memory modules. The introduction of Solid State Disks (SSD) is also a big contributor.

Obviously people will argue that:

  • the performance for most systems is task specific
  • and blades provide higher density (more CPU cores) per rack,
  • and some systems provide more performance and maybe more performance/Watt,
  • etc…

Well; datacenter facility guys couldn’t care less about those arguments. For them it’s about the power per 1U or the power per rack and its trend.

With a depreciation time of 10-15years on facility equipment, the datacenter needs to support many IT refresh cycles. IT guys getting faster CPU’s, memory and bigger disks is really nice and it’s even better if the performance/watt ratio is great… but if the overall rack density goes up, than facilities needs to support it.

To provide more perspective on the density of the CPU/rack, I plotted the amount of CPU cores at a 40U filled rack vs. total power at 40U:

power draw rack trend

Still impressive numbers: between 240 and 720 CPU cores in 40U of modern day equipment.

Next I wanted to test my hypotheses, so I looked at a active 10.000+ server deployment consisting of 1-10 year old servers from Dell/IBM/HP/SuperMicro. I ranked them in age groups 2003-2013, sorted the form factors 1U Rackmount, Blades and Density Optimized. I selected systems with roughly the same hardware config (2 CPU, 2 HD, 8GB RAM). For most age groups the actual power consumption (@ 100,80,50%) seemed off by 10%-15% but the trend remained the same, especially among form factors.

It also confirmed that after the drop, due to energy optimized components and SSD, the power consumption per U is now rising slightly again.

Density in general seemed to rise with lots more CPU cores per rack, but at a higher power consumption cost on a per rack basis.

Let’s take out the Cristal ball

The price of compute & storage continues to drop, especially if you look at Amazon and Google.

Google and Microsoft have consistently been dropping prices over the past several months. In November, Google dropped storage prices by 20 percent.

For AWS, the price drops are consistent with its strategy. AWS believes it can use its scale, purchasing power and deeper efficiencies in the management of its infrastructure to continue dropping prices. [Techcrunch]

If you follow Jevons Paradox then this will lead to more compute and storage consumption.

All this compute and storage capacity still needs to be provisioned in datacenters around the world. The last time IT experienced growth pain at the intersection between IT & Facility it accelerated the development of blade servers to optimize physical space used. (that was a bad cure for some… but besides the point now..) The current rapid growth accelerated the development of Density Optimized servers that strike a better balance between performance, physical space and energy usage. All major vendors and projects like Open Compute are working on this with a 66.4% year over year in 4Q12 growth in revenue.

Blades continue to get more market share also and they now account for 16.3% of total server revenue;

"Both types of modular form factors outperformed the overall server market, indicating customers are increasingly favoring specialization in their server designs" said Jed Scaramella, IDC research manager, Enterprise Servers "Density Optimized servers were positively impacted by the growth of service providers in the market. In addition to HPC, Cloud and IT service providers favor the highly efficient and scalable design of Density Optimized servers. Blade servers are being leveraged in enterprises’ virtualized and private cloud environments. IDC is observing an increased interest from the market for converged systems, which use blades as the building block. Enterprise IT organizations are viewing converged systems as a method to simplify management and increase their time to value." [IDC]

With cloud providers going for Density Optimized and enterprise IT for blade servers, the market is clearly moving to optimizing rack space. We will see a steady rise in demand for kW/rack with Density Optimized already at 8-10kW/rack and blades 12-16kW/rack (@ 46U).

There will still be room in the market for the ‘normal’ rackmount server like the 1U, but the 2012 and 2013 models already show signs of a rise in watt/U for those systems also.

For the datacenter owner this will mean either supply more cooling&power to meet demand or leave racks (half) empty, if you haven’t build for these consumption values already.

In the long run we will follow the Gartner curve from 2007:

image

With the market currently being in the ‘drop’ phase (a little behind on the prediction…) and moving towards the ‘increase’ phase.

More:

Density Optimized servers (aka microservers) market is booming

IDC starts tracking hyperscale server market

Documentation and disclaimer on the HP Power Advisor

Posted in Cloud, Nieuwe technologie | Getagged: | Leave a Comment »

 
Volg

Ontvang elk nieuw bericht direct in je inbox.