MI Modern Infrastructure - docs.media.bitpipe.comdocs.media.bitpipe.com/io_12x/io_123685/item... ·...

Home

Editor’s Letter

Container Your Enthusiasm

#Hashtag: Twitter on #Docker

Colocation With a Side of Cloud

Survey Says: Big Data Decisions

Explained: Inside ChatOps

Ethernet’s Fountain of Youth

Overheard @ ChefCon 2015

The Next Big Thing: It’s Alive!

In the Mix: The Dismal State of IT Security

End User Advocate: Mobile’s Our Problem Now

Citrix Synergy and Modern Infrastructure Decisions Summit

SURVEY SAYS

Big Data Decisions

DATA CENTER NETWORKS


MIModern InfrastructureCreating tomorrow’s data centers

MAY 2015, VOL. 4, NO. 5

THE NEXT BIG THING

It’s Alive!

END USER ADVOCATE

Mobile’s Our Problem Now


Can Docker revive containers as a server consolidation tool?

EDITOR’S LETTER

They’re Baaaack

#HASHTAG

On #Docker

DATA CENTER FACILITIES


EXPLAINED

Inside ChatOps

OVERHEARD

@ ChefCon 2015

IN THE MIX

The Dismal State of IT Security

MODERN INFRASTRUCTURE • MAY 2015 2

JUST WHEN YOU thought you had finally said goodbye to old-school IT, you find out it’s back—with a twist.

Remember operating system-level virtualization? Think Solaris Zones or Parallels Virtuozzo. This method of server virtualization is back with a vengeance, this time as containers, with Docker leading the charge. But whereas the old-school containers were used largely by service providers to eke out every last drop of capacity from their server farm, today’s containers are being pitched as an ap-plication-portability play, says Nick Martin in “Container Your Enthusiasm.” But perhaps that’s selling the tech-nology short? Martin finds several organizations that are actively exploring the use of containers as an alternative to hypervisor-based virtualization à la VMware.

And when it comes to data center infrastructure, not only has the entire world not moved to public cloud, but plenty of organizations are still using that old-school model: the data center colocation facility. That said,

today’s colo isn’t your granddaddy’s colo. In her story, “Colo With a Side of Cloud,” Meredith Courtemanche finds that all but the smallest colo providers have begun to offer some level of cloud services—whether it’s reselling connections to public cloud providers or selling cloud technology themselves.

Ethernet, meanwhile, never left the data center. Still, tomorrow’s Ethernet doesn’t look like the technology you know and love. Not only are we looking at a variety of new speeds (from 2.5 gigabits per second all the way up to 400 Gbps), but Ethernet is poised to benefit from new cabling and connectivity mechanisms, writes Stephen Bigelow in, “Ethernet’s Fountain of Youth.”

“A Cisco staffer once said that only cockroaches and Ethernet will survive the nuclear holocaust—they may just be right,” Bigelow writes. n

ALEX BARRETT is Modern Infrastructure’s editor in chief. Email her at [email protected].

Home

Editor’s Letter











EDITOR’S LETTER

They’re Baaaack

mailto:[email protected]


Today’s containers solve application portability problems. Could they also drive the next wave

of server consolidation? BY NICK MARTIN


CONTAINERS ARE ENJOYING a renewed interest within enter-prise IT, courtesy of Docker. Some analysts have specu-lated they’re the next logical step in server consolidation to replace VMs—a claim even Docker is skeptical about.

The intriguing wrinkle in this new containerized approach is that it’s not new. The idea of containers has been around since the early days of Unix, with the chroot command. Linux containers, the technology upon which Docker’s software was originally built, were introduced in 2008. So, what’s with the sudden surge in interest?

Containerized applications share a common operating system kernel, eliminating the need for each instance to run on its own separate operating system. An application can be deployed in a matter of seconds and using fewer resources than with hypervisor-based virtualization. However, since the applications all rely on a common OS kernel, this approach can work only for applications that share the exact OS version. Docker found a way to address this limitation.

DOCKER LEADS THE WAY

Docker was released as an open source project by dot-Cloud, a platform as a service company, in 2013. Docker relies on Linux kernel features, such as namespaces and

VIRTUALIZATION

HOMEURBANCOW/ISTOCK

Home

Editor’s Letter












Home

Editor’s Letter











cgroups, to ensure resource isolation and to package an application along with its dependencies. This packaging of the dependencies enables an application to run as expected across different Linux operating systems—sup-porting a level of portability that allows a developer to write an application in any language and then easily move it from a laptop to a test or production server—regardless of the underlying Linux distribution. It’s this portabil-ity that’s piqued the interest of developers and systems administrators.

“Prior to Docker, the portability of an application or service was never guaranteed,” said David Messina, a mar-keting vice president at Docker. “Because of the way that Docker containers separate the application constraints from infrastructure concerns, we help solve that depen-dency hell.”

Almost immediately, developers started to notice how this new approach could solve one of their biggest frustrations. One month after launching an interactive tutorial in August 2013, Docker said 10,000 developers tried it out. Within a year, companies such as Red Hat and Amazon added commercial support for Docker—even as Docker executives cautioned users against production use. When Docker announced its 1.0 release in June 2014, the Docker Engine software had already been downloaded

2.75 million times. That number now stands at more than 100 million.

Analysts say Docker’s software is well-timed, arriving as more and more companies invest in cloud computing and in the midst of the current DevOps movement, said Jay Lyman, research manager at 451 Research.

“Docker provides an integrated user interface. It pro-vides a greater level of simplicity. You don’t have to be a Linux kernel expert to use Linux container-based tech-nology with Docker. It broadened the pool of potential developers,” Lyman said.

The intensified spotlight on Docker has also served to highlight its flaws, and it’s possible that it became too pop-ular too soon. At least, that’s the thinking of Cal Leeming, a software engineer and Docker critic who’s voiced his concerns on his blog and through social media. During a six-month trial in a production environment, Leeming said he found Docker’s software and the Docker Hub

n Analysts have speculated whether containers are the future of server consolidation.

n Prior to Docker, the portability of an application or service was never guaranteed.

n Docker’s Engine software has been downloaded more than 100 million times.

HIGHLIGHTS

THE INTENSIFIED SPOTLIGHT ON DOCKER HAS ALSO SERVED TO HIGHLIGHT ITS FLAWS.


Home

Editor’s Letter











Registry slow and frustrating.“It seems clear to me that they were under pressure

from the people giving them funding to get something out the door,” Leeming said. “The reason I wrote about Docker is not to destroy or get in the way of a project that’s going somewhere. But, so many people are trying to treat this like it’s going to be the next damn industry standard. When you see something like that—everyone is talking about it—and you know the solution is flawed, you’ve got to fight back.”

“We were very quickly able to use Docker to build de-velopment and test environments for various developers and become productive right away without interfering with production systems,” said Tom Chernetsky, CTO of Yik Yak, a mobile application company based in Atlanta. “In that way, Docker was a game-changer for us as a fast-growing company.”

The unexpected success of Docker has also brought attention to several competing approaches to container virtualization and spurred others to develop their own. Late in 2014, CoreOS CEO Alex Polvi introduced the com-pany’s new container project, called Rocket, as a direct response to Docker’s “fundamentally flawed” approach.

Docker’s technical approach is not secure, because it requires a central Docker daemon, Polvi said. Rocket, on the other hand, relies on the systemd daemon to create a container.

“It remains to be seen what the official standard for containers is going to be,” 451 Research’s Lyman said. “I think we’ll see something more like what we’ve seen with hypervisors. VMware is the most prominent and widespread, but it’s certainly not the standard, and we’re likely to see a similar thing with Docker and Rocket, and maybe others.”

CONTAINERS AND CONSOLIDATION

Docker’s ability to bring new life to the otherwise stale technology of container virtualization has led some analysts to ask whether the company’s success could also revive the use of containers as a server consolidation tool.

Containers have played a role for many years—especially in Linux shops running a common operating system—as a lightweight approach to ensure isolation of multiple workloads on a physical server. With hypervisor-based

Docker’s High Profile Customers PayPal Spotify Rackspace Bleacher Report Yelp


Home

Editor’s Letter











virtualization, a software layer (the hypervisor) abstracts the underlying physical hardware of a server, allowing for the creation of VMs upon which an operating system and then applications can be installed. Unlike hypervi-sor-based VMs, containers do not aim to emulate physical servers. Instead, all containerized applications on a host share a common operating system kernel. This eliminates the resources needed to run a separate operating system for each application and can greatly reduce overhead.

This reduction in expenses is especially appealing for cloud service providers, which rely on low costs to offer competitive pricing while turning a profit. In most cases, the inner workings of these companies’ infrastructures are closely guarded secrets, but occasionally one will open up about certain aspects of its approach. Google, for example, has said it runs nearly every application within Omega, its homegrown containerization software.

This trend toward containers as an alternative to VMs among cloud providers could have big implications for companies providing VM-management software and ser-vices. VMware introduced enterprise IT to its GSX Server virtualization platform in 2001. Over the next decade the company built an enterprise of its own, raking in profits and becoming the name in server virtualization. Today, it still stands head and shoulders above its nearest com-petition (Microsoft) in the hypervisor business. But the renewed interest in containers has some wondering if the hypervisor party could be coming to an end.

“Containers are effectively a more efficient, lower cost way to do nearly everything you can do with hypervi-sors,” said Marc Staimer, senior analyst at Dragon Slayer

Consulting. “Containers are the next generation of virtu-alization, and they’re a direct threat to VMware.”

VMware certainly isn’t ignoring the new container hype, rolling out updates that streamline the process of moving Docker containers from development platforms to test or production in vSphere or vCloud Air environ-ments. Company executives have said they see Docker as a partner that offers developers a tool—and not as a threat to their server virtualization business. And, of course, VMware’s support for containers running on top of VMs

Vroom, VroomWHILE SOME CONCERNS remain about the use of con-

tainers in production environments, one aspect

that doesn’t appear to raise eyebrows is the per-

formance of applications running as containers.

A July 2014 IBM research report examined the

performance of Linux containers compared to

VMs and found containers offer performance as

good or better than VMs. In some cases, contain-

erized applications performed at levels closer to

native installations. Another benchmark report,

by VMware, found Docker containers performed

at near-native levels. VMware’s report also shows

the performance of Docker containers running

inside VMware VMs was similar to applications

running directly on the VM. n


Home

Editor’s Letter











solves the application portability problem, but does noth-ing to further consolidation.

In fact, Docker doesn’t have any public plans to sup-plant established virtualization vendors.

“In practice, Docker is implemented in a vast majority of cases where containers are running on top of VMs. The notion that containers are a replacement for VMs is certainly not one that we perpetuate,” Docker’s Messina said. “The core values of Docker containers are complete portability of applications … as opposed to higher densi-ties on hardware.”

HOW CONTAINERS FIT IN

Barriers remain to using containers as a consolidation tool. For starters, many of the familiar concerns that crit-ics of hypervisor-based virtualization raised several years ago are being cited again. For example, higher consolida-tion rates raise the risk of widespread workload disruption if hardware fails. Others point to the security concerns inherent in an approach in which a single kernel exploit could affect all containers on a host. And, the orchestra-tion tools and advanced management features available for VMs are—so far—largely missing for containers.

However, Staimer isn’t alone in thinking that what has worked for Google and large cloud providers can work for other companies. For example, another relatively new company, Colorado-based DH2i, is specifically positioning its Windows container management software to consoli-date servers and save money on software licensing.

If Docker’s internal business roadmap reflects its public

hesitance to push containers as a consolidation tool, exec-utive leadership may be taking the company in the wrong direction, says Simon Bramfitt, an analyst at The Virtual-ization Practice.

“It doesn’t seem as though they’re really playing to the strengths of the platform if all they want to do is make developers’ lives a little easier,” Bramfitt said. “I think there is a real need for containerization of workloads in an enterprise environment. And if they’re choosing not to go after that market, then I would wonder if that is more out of a desire to go after safer markets that they’ve identified as facing less competition, from VMware, for example.”

Even some Docker customers, including Yik Yak’s Cher-netsky, say they might consider testing containers as an alternative to VMs.

“The reality is that everything is about mixed use, using a number of different tools and technologies to accom-plish your enterprise IT objective,” 451 Research’s Lyman said. “Containers are just another tool.”

“Docker and containers are not going to replace VMs anytime soon. A lot of the VM use will stay, right along-side containers. But Docker and containers don’t have to topple VMware in the market to continue to have tremen-dous growth and traction. If containers replace just 10% or 15% of VMware VMs that are out there, that’s pretty darn significant, just given how pervasive VMware is in enterprise IT.” n

NICK MARTIN is senior site editor for SearchServerVirtualization. Email him at [email protected].


zzzzzz

Home

Editor’s Letter












Jessica Twentyman

@jtwentyman

Tech things my recent travels have taught me I urgently need to understand better: #docker (and contain-erisation in general), #chef, #puppet

Plamen Kalchev

@PlamenKalchev

Time to get my hands dirty with #docker and #kubernetes Lets hope it won’t hurt :)

Jeroen Moors

@JeroenMoors

It’s interesting how much secrecy there [is] about future features of (opensource) proj-ects around #docker

Justin Parisi

@NFSDudeAbides

#VMware going after #Docker with announcement of #Photon? #ContainerWars #vExpert

Hamza

@aikchar

Is it just me or is #Docker doing things that don’t make it re-ally isolated and run anywhere unmodified? Unlike #FreeBSD #jails.

David Lyle

@dlyle65535

#docker has com-pletely integrated itself into my daily dev. Spin it up, try it out, shut it down. #tooeasy

Stephen Masters

@sctrcdr

Must resist jumping into #Kubernetes. I think I should learn more about #Docker first. But it’s so tempting...

Steven Peguero

@misterpeguero

On the Microsoft side of the fence, I’m glad to see they’re trying to develop Nano and Windows support for #Docker! That + #PowerShell = <3.

#Hashtag Twitter on #Docker

https://twitter.com/hashtag/docker?src=hash

https://twitter.com/hashtag/chef?src=hash

https://twitter.com/hashtag/puppet?src=hash


https://twitter.com/hashtag/kubernetes?src=hash


https://twitter.com/hashtag/VMware?src=hash

https://twitter.com/hashtag/Docker?src=hash

https://twitter.com/hashtag/Photon?src=hash

https://twitter.com/hashtag/ContainerWars?src=hash

https://twitter.com/hashtag/vExpert?src=hash


https://twitter.com/hashtag/FreeBSD?src=hash

https://twitter.com/hashtag/jails?src=hash


https://twitter.com/hashtag/tooeasy?src=hash

https://twitter.com/hashtag/Kubernetes?src=hash



https://twitter.com/hashtag/PowerShell?src=hash


IF YOU’RE LIKE most traditional IT organizations, you have a disaster recovery site residing in a colocation center 20 miles from your main data center. But if you’re like most startups, your servers spin up and down on a public cloud running who-knows-what hardware.

Today, with startups outgrowing their clouds and leg-acy data centers aging out of usefulness, colocation and dedicated hosting serve as transitional space for produc-tion workloads, with a range of services beyond power, cooling and hardware to support enterprises.

Conventional organizations build or upgrade on-prem-ises data centers because it’s what they’ve always done, said Josh Hatten, consulting manager of a data center re-location consulting business within Eden Prairie, Minn.’s DataLink. But this conventional approach doesn’t always make sense, he said, in terms of cost, uptime or perfor-mance. He’s seen a general trend toward colocation over the six years since the recession, and IT organizations are increasingly adopting managed services such as infra-structure monitoring and management.

Many outsourcing providers such as CenturyLink and QTS blur the lines between colo, managed services and cloud. IT can customize a suite of cloud services—backup

HOME

DATA CENTER FACILITIES

PESHKOV/ISTOCK

Colocation With a Side of CloudOrganizations have long built and upgraded on-premises data centers, but does that make sense anymore?BY MEREDITH COURTEMANCHE


Home

Editor’s Letter











as a service, disaster recovery as a service, low-latency hybrid cloud interconnects—and colocation space, while tacking on extras such as IT asset lifecycle management. This accommodates existing capital investments and workloads that require direct business ownership and control.

“The whole is bigger than the sum of its parts,” said Patrick Gilmore, CTO of cloud and colocation provider The Markley Group in Boston. Markley recently added ap-plication and server monitoring to its portfolio of services.

THE COLO COMPROMISE

Although most workloads can conceivably run in the cloud, colocation and dedicated hardware won’t disappear any time soon. “You need to adapt to the way companies do things,” Gilmore said. An IT department at a financial, insurance or retail company has evolved its back-end in-frastructure to fit its specific needs over the past 20 years. “They’re not going to change their entire system to save a bit of money.”

Music streaming service Cloud Cover Music, based in Santa Monica, Calif., represents the other side of the equation—a move off public cloud. Managed services on dedicated hardware from Tualatin, Ore.’s Peak Hosting

offered the right balance between owning a data center and pure cloud.

“I get immediate monitoring at machine levels, as if I was in the data center myself,” said John Shiple, CTO of Cloud Cover Music. It’s cost effective because he doesn’t

have to pay for a sys admin to do the monitoring required on Amazon Web Services instances.

The combination of application and server-level mon-itoring services on dedicated hardware enabled Cloud Cover Music to optimize its code and catch problems sooner, before customers felt the effects, he said. This connects developers with operations—without having to restructure into a DevOps IT shop.

“We used ... a number of [cloud] companies on the

n Colocation and dedicated hosting serve as transitional space for production workloads.

n Analysts see a trend toward colocation over the past six years since the recession.

n IT organizations are increasingly adopting infrastructure monitoring and management.

HIGHLIGHTS

BY USING SELECTIVE OUTSOURCING TO HANDLE ROUTINE IT TASKS, THE ORGANIZATION RELIES ON HIGHLY TECHNICAL IT STAFF AT ANY TIME OF DAY, BUT ONLY WHEN NEEDED.


Home

Editor’s Letter











server side [before transitioning to dedicated hardware hosting],” said Carrie Pobre, business development direc-tor at Cloud Cover Music. “Peak was a company that could provide that back-end service and structuring to allow us to scale,” including a move into more advanced business analytics products.

By using selective outsourcing to handle routine IT tasks, the organization relies on highly technical IT staff at any time of day, but only when needed.

“You don’t want to hire someone to set up an OS, do security patches, etc. These aren’t part of your core busi-ness,” said Markley’s Gilmore.

Just as every enterprise has different skillsets and goals, every market has different needs. For Silicon Valley-based Colovore, managed services aren’t as high priority as power and cooling density. The provider offers water-based cooling with rear-door heat exchangers and other facility infrastructure to support 20 kW in every rack, priced by actual power use rather than potential use per rack.

Silicon Valley, one of the largest data markets in the U.S., has different density needs than financial, govern-ment or other traditional enterprise IT markets, said Ben Coughlin, one of Colovore’s founders. The industry verticals are big data, ad serving, bio informatics, gaming, content providers and cloud or SaaS providers.

These companies are at the leading-edge of hardware and software deployments—pushing densities higher— in a region bursting with IT talent. IT teams may not want to build racks and power systems, but they have highly technical staff at their disposal, and a lot of the

Colo LingoWhite space: White or data hall space is the area

where IT equipment resides. Depending on the

colocation contract, the leased space includes

both power/mechanical and white space, or racks

in the white space alone.

Wholesale colocation: Wholesale colocation is

akin to renting an entire new facility. Typically it

means more control—and more expertise—com-

ing from your company.

Retail colocation: While usually associated with

sub 1-MW deployments, retail colo can mean any

size IT infrastructure where the colocation pro-

vider handles much of the data center staffing

and tasks.

Dedicated cloud hosting: Dedicated hosting gives

you whole servers for workloads, with support

services and flexible scaling. The servers’ re-

sources are not shared with other users.

Public cloud hosting: Public clouds abstract the

physical servers from the instances, which share

hardware. This maximizes efficiency but takes

ownership a step farther away from the IT org. n


Home

Editor’s Letter











newest tools to manage apps and infrastructure, Coughlin said.

‘BUY THE BASE’ BUNK?

White space in colocation facilities has become increas-ingly commoditized in the years following the recession, to the point where a cabinet costs $600 to $700 per month in some facilities, according to Hatten. Five years

ago, TechTarget research put colocation racks at $700 to $4,000 per month, depending on density. Many coloca-tion providers today rely on lessees to adopt their cloud and managed services offerings.

If you don’t have an existing million-dollar server in-vestment to depreciate, however, don’t necessarily assume that a server refresh into colo space is cheaper than hosted servers.

The adage of ‘buy the base, rent the spikes’ doesn’t

Solving for Skills, Not SpaceGEOGRAPHY AND EXISTING talent and infrastructure also

play a role in IT decisions, especially in emerging mar-

kets. Africa’s burgeoning digital connectivity is still

unreliable for modern business. Bidco, a Kenya-based

retailer with locations in 15 countries on the continent,

outsourced its IT infrastructure management to IBM’s

technical services division, but remain in its existing

data center rather than adopt cloud or colocation

services.

Centralized back-end IT is an enabler for the busi-

ness to continue expanding across Africa, said com-

pany CIO Alkane Patel, but it is not Bidco’s core

business. When it came time for a server refresh initia-

tive, the company decided managed services from IBM,

consulting with in-house IT specialists, would yield the

best results going forward.

“Rather than finding a problem and pulling in part-

ners to fix it, now we have proactive measures before

a failure, with continuous updates and patches,” Patel

said. Infrastructure as a service yields high uptime

with predictable costs.

While Bidco would like to use some cloud services,

such as software as a service versions of certain ap-

plications, the bandwidth isn’t good enough yet to

support it.

“Within the next five to seven years, connectivity

and support for cloud services in Kenya will pick up,”

Patel said. “Managed services are a good step, but [it

would be simpler to] plug-and-play to the Internet and

start using a good platform.” n


Home

Editor’s Letter











always work today, Markley’s Gilmore said. “There’s an assumption that 24/7 servers are cheaper to own than rent,” he said, and that’s only true for the companies with massive buying power, such as Web-scale IT names. Over the past five to 10 years, cloud prices from the likes of Amazon Web Services (AWS), Google Cloud Platform and other providers have dropped enough to make enterprises question buying the base, he said.

Know what you need for steady operation, and how much it will cost, Coughlin said. Low prices typify stan-dard AWS or Google instances, but when you tweak and customize deployments, “the costs on a public cloud will go through the roof,” he said, giving the example of

a company in the gaming industry that cut $5 million in annual costs by leaving AWS and managing its own inno-vative, high-density hardware in the colocation facility.

Understanding the base is also important. IT organiza-tions must frame cost and return on investment discussions about cloud and colocation within an overarching data center strategy that delineates uptime and performance expectations for different workloads, Datalink’s Hatten said. Otherwise, sprawl and inefficiencies will persist. n

MEREDITH COURTEMANCHE is the site editor for SearchDataCenter and has covered semiconductor and IT topics since 2006. You can reach her on Twitter: @DataCenterTT.

http://www.twitter.com/datacenterTT

w


Survey SaysBig Data DecisionsHome

Editor’s Letter











D What are the primary business goals of your organization’s big

data project?*

D Will you look into different vendors, or stay with your current vendor?

*MULTIPLE SELECTIONS ALLOWS; SOURCE: TECHTARGET BIG DATA SURVEY 1Q 2015;

BASED OFF RESPONSES FROM 150 IT AND BUSINESS PROFESSIONALS

SOURCE: TECHTARGET BIG DATA SURVEY 1Q 2015; BASED OFF RESPONSES FROM 150 IT AND BUSINESS PROFESSIONALS

76D Percentage of respondents who say they purchased or upgraded

big data analytics

SOURCE: TECHTARGET BIG DATA SURVEY 1Q 2015;

BASED OFF RESPONSES FROM 79 IT AND BUSINESS PROFESSIONALS

Improve organizational efficiency and profitability

Gain competitive advantages over business rivals

Better understand customer needs, preferences and buying decisions

Drive increased revenue and/or new revenue opportunities

Use system/network log data to help improve IT operations

Track sentiment about our company and products on social networks

43+41+12+4+Q43%

Undecided, still evaluating

all options

4% Sticking with

the same vendor

41% Exploring different vendors

12% Same vendors for some products; new vendors for others

43%

39%

31%

27%

11%

8%


Home

Editor’s Letter











DOES YOUR IT team spend a good chunk of its day in a chat client, collaborating on projects and troubleshooting issues? If so, check whether your chat platform supports ChatOps, so the team can execute external scripts and commands without ever having to leave the chat room.

ChatOps emerged in the early part of this decade at GitHub, the Web-based code repository service. “It started out as how to run a handful of commands and execute them from a chat bot,” said Brian Doll, GitHub’s vice president of strategy. By 2012, the company had open-sourced Hubot, its “customizable, life-embetterment robot,” which sits in chat rooms and listens for commands to execute. Today, Hubot is supported on upwards of 30

chat platforms, including HipChat, Yammer, Skype and IRC. GitHub counts about 3,600 Hubot projects on its site.

IT professionals who use ChatOps sing its praises. “When customers are moving very, very quickly, we need an efficient way to communicate about new features and test new code,” said Eric Brinkman, senior product man-ager at Rackspace, which uses a ChatOps-enabled client from Slack Corp. to support customers of its DevOps ser-vice. Through Slack, support personnel can communicate in real time with customers in a dedicated channel, use Slack’s ChatOps integrations to look at monitoring data from New Relic or run Chef recipes. “It’s become our preferred way to communicate as a team,” Brinkman said.

Some ChatOps clients come with a variety of pre-built integrations. Slack said it has over 70 pre-built integra-tions, with more on the way. At Rackspace, the DevOps team uses Slack integrations with common DevOps workflows such as continuous integration and continuous deployment, autoscaling, and spinning up staging and pro-duction environments—“the highest use case workflows,” Brinkman said. If the integration is unavailable, many modern tools allow you to write your own via application programming interfaces (APIs).

One of the big benefits of today’s ChatOps clients is their ubiquity. Slack, for instance, is available on Win-dows, Mac, Android and iOS. That way, users can not only communicate from wherever they may be, but also take

EXPLAINED

Inside ChatOpsLearn how a ‘life-embetterment robot’ can change how IT teams get work done from inside their chat platforms. BY ALEX BARRETT


Home

Editor’s Letter











action from wherever they may be. It also records the conversations and actions, creating a

de facto audit trail that can be referred to later. “Think of it like a persistent IRC,” Brinkman said. That’s especially important in 24x7 support environments, where people come on and off-duty. “If you leave for a couple of days, you can easily catch up on everything you’ve missed,” he said.

For teams that make heavy use of ChatOps, it has be-come the place from which all work is done, said GitHub’s Doll. “In the same way that DevOps shops frown on you installing something directly on a server rather than through the configuration management system [e.g.,

Puppet or Chef], with ChatOps, it’d be very odd to run a terminal command on a server,” he said. “It’d be seen as sort of rude—like, why are you limiting the visibility of that server?”

GitHub, not surprisingly, is far down the ChatOps path. “We use it for assigning pager duty, incident response, load balancing—literally every operational task that we have we can do via ChatOps,” Doll said. Other teams might need to start slow. “If you haven’t invested in modular tool-ing, with good APIs, then it might be more of a ramp.” n

ALEX BARRETT is editor in chief of Modern Infrastructure. Email her at [email protected].



IN THE IT industry, where lifecycles are often just a few years, few technologies have proven as ubiquitous, resil-ient and extensible as Ethernet. It’s been around for 40 years and no one is writing its epitaph anytime soon.

Ethernet dominates the connectivity landscape—it’s the standard across homes, offices and data centers—and has matured to handle all types of data including files, messages, voice, streaming media and almost every other conceivable content type. But the demands on our networks are constantly increasing, and Ethernet has evolved to provide high levels of bandwidth and efficient utilization that its inventors at Xerox PARC never could have imagined back in 1974. But where is Ethernet headed over the next few years?

SPEEDS AND FEEDS

Bandwidth is king in networking, but it also doesn’t just happen—it takes carefully considered and thoughtfully developed standards to define the physical and logical el-ements needed to implement technologies like Ethernet. Both the Ethernet Alliance and the Institute of Electrical and Electronic Engineers (IEEE) have been instrumental

HOMEAKINDO/ISTOCK


New standards and software platforms promise more bandwidth and versatility

for tomorrow’s Ethernet networks.BY STEPHEN J. BIGELOW

DATA CENTER NETWORKS


Home

Editor’s Letter











in defining and developing standards for Ethernet, and any discussion of network “speeds and feeds” is pointless without including the prevailing standards efforts.

We saw the importance of network standards in 1998 as gigabit Ethernet (GigE) arrived on twisted pair copper cable (IEEE 802.3ab) to challenge legacy 10/100 megabit Ethernet deployments. Today, 1 GigE servers and network interface cards (NIC) are almost universal, but 10 GigE on optical fiber or twisted pair copper cable (IEEE 802.3ae-2002) deployments are growing and expected to outpace 1 GigE by 2016. By 2010, initial standards for 40 GigE and 100 GigE were adopted (IEEE 802.3ba-2010), though the supplemental standards needed to fully support both using twisted pair copper cabling are still in development with Working Groups (such as IEEE P802.3bq and IEEE P802.3bw, respectively). While 40 and 100 GigE deploy-ments have limited use in backbone and provider-class environments, analysts expect 40 GigE deployments to start ramping up with broader adoption in 2018.

Emerging standards will open new bandwidth operat-ing points within today’s familiar range such as 2.5 GigE, 5 GigE, 25 GigE and 50 GigE, and enable substantially faster bandwidth options such as 200 GigE and even 400 GigE.

“It’s not as chaotic as it seems,” said John D’Ambrosia, Ethernet Alliance Chairman. “The Ethernet market is

diversifying, and there are groups working on everything from DC [Power over Ethernet] to 400 GigE.” The benefit of such diversity, said D’Ambrosia, is that a greater selec-tion of available Ethernet technologies allows businesses to tailor networks to best match their short- and long-term

n Ethernet has been around for 40 years and there are no signs of it fading away.

n Ethernet has evolved to provide high levels of bandwidth and efficient utilization.

n Emerging standards will open new bandwidth operating points within today’s familiar range.

HIGHLIGHTS

Ethernet Bandwidth Points

1 GigE

2.5 GigE

5 GigE

10 GigE

25 GigE

40 GigE

50 GigE

100 GigE

400 GigE

n CURRENTLY AVAILABLE

n COMING SOON


Home

Editor’s Letter











business goals while observing budget constraints.Just consider the impact of a broader set of Ethernet

bandwidth options—even at the lower end. For example, establishing 2.5 GigE and 5 GigE standards would enable broader wireless access to support increased adoption of the IEEE 802.11ac Wi-Fi standard. In addition, such modest increments support enhanced bandwidth within cabled networks without the need to replace aging Cat-egory 5e and Category 6 cables with the Category 6a or better twisted pair cabling—saving considerable expense and disruption. While there is no ETA for these initiatives, the Ethernet Alliance, NBASE-T Alliance, and the IEEE 802.3 Next Generation Enterprise Access BASE-T PHY Study Group are all debating the development of these increments to create a single interoperable standard.

NEVER FAST ENOUGH

Meanwhile, not all vendors with networking interests are entirely satisfied with 40 GigE and 100 GigE bandwidth points for high-speed networking. A vendor-led alliance called the 25G Ethernet Consortium is developing a 25 gigabit single-lane and dual-lane 50 gigabit Ethernet link protocol (currently draft version 1.4) to allow two additional options between 10, 40 and 100 GigE. Early members in the Consortium include Microsoft, Mellanox, Google, Broadcom and Arista, which see the move as a way to boost performance and lower the interconnect costs between server NICs and top-of-rack switches. The IEEE is also studying 25 GigE potential through a 25 Gb/s Ether-net Task Force (IEEE P802.3by). The Consortium hopes to

start rolling out compliant implementations of these new standards in early 2016, and the IEEE anticipates ratifying its 25 GigE standard sometime later in 2016.

Speed remains a priority as standards organizations seek to shatter 100 GigE with a plan for 200 and even

400 GigE. Such astonishing speeds should support the most demanding service provider networks. Chances are that a 400 GigE standard will appear sometime in 2017 as the Ethernet Alliance and the IEEE P802.3bs 400 Gb/s Ethernet Task Force both work to develop the technology. D’Ambrosia expects a 200 GigE version of the standard sometime between 2018 and 2020.

This trend toward intermediate speeds might seem odd at first, but such quantum leaps often yield a “trickle down” benefit that accelerates adoption of complemen-tary standards—the underlying technical issues have already been resolved in the faster, more mature standard. “With the 400 gig project, we could potentially make a lot of technology steps to go faster, introduce new fibers, new modulation, more lambdas, or all of the above,” D’Ambro-sia said. D’Ambrosia notes that a 400 GigE approach could potentially employ eight 50 gigabit lanes, four 100 gigabit lanes, and other creative uses of technical innovations that

A NEW CROP OF ETHERNETCONTROLLERS HAVE EMERGED TO ACCOMMODATE A WIDER RANGE OF CONNECTIONS...


Home

Editor’s Letter











have already been established. The net result drives com-ponent manufacturing volumes and lowers costs to make a more diverse body of Ethernet options more interoperable and cost-effective.

CABLES AND POWER AND CHIPS! OH MY!

New speeds and feeds may generate a great deal of ex-citement, but making quantum leaps in Ethernet band-width—or just expanding the number of bandwidth choices available—will depend on new developments in cabling, connectors, and controller chips.

Cables and connectors will play a huge role in tomor-row’s Ethernet, and new standards will be developed to further enhance Ethernet connections. For example, 40 and 100 GigE already define optical fiber connections, but the IEEE P802.3bm seeks to define next-generation optical cables that support 40 gigabits up to 40 kilometers over a single-mode optical fiber and 100 gigabits up to 100 meters of multi-mode optical fiber, offering high-perfor-mance Ethernet over greater distances within buildings and metropolitan areas. IEEE P802.3bq seeks to influence both 25 and 40 GigE by defining full-duplex operation over four pair twisted pair copper cabling up to 30 meters. The Ethernet Alliance has already named these Category 8 cables and anticipates availability in 2016.

Ethernet has even made inroads into automotive and industrial environments with recent developments, such as the 1000BASE-T1 1GbE PHY Task Force, which is working on IEEE P802.3bp to support Gigabit Ethernet over copper cabling. At the same time, IEEE P802.3bv is

exploring 1 GigE connectivity using plastic optical fiber up to 50 meters in home and industrial settings, and up to 15 meters in automotive applications.

Another oft-overlooked Ethernet development is Power over Ethernet (PoE), which can provide power to some network gear without the need for a separate power outlet. PoE isn’t a new idea; IEEE 802.3af-2003 and the updated IEEE 802.3at-2009 have both been rolled into the overall IEEE 802.3-2012 Ethernet standard. But the work is on-going, and the IEEE P802.3bt DTE Power via MDI over 4-Pair Task Force is updating the standard for 10 GigE compatibility and supply up to 40 watts of power for larger Ethernet devices. Considering the proliferation of wireless access points and other supporting network devices, PoE development will allow faster and more power-hungry device deployment far from electrical outlets.

And let’s not forget the need for support in the silicon. A new crop of Ethernet controllers have emerged to accom-modate a wider range of connections and deliver a host of advanced network features. Intel’s XL710 controller chip supports quad small form-factor pluggable transceivers capable of handling both optical and copper media for 10 and 40 GigE applications such as active optical cables and modular optical cables along with twinax and active copper cabling.

Advanced controllers like the LX710 add a spate of powerful features to improve how Ethernet traffic is or-ganized and handled. There are the familiar features that have been around a while such as NIC port teaming (also called link aggregation or channel bonding), checksum offload, large send offload (LSO), jumbo packets, and


Home

Editor’s Letter











native support for Fibre Channel over Ethernet (FCoE) storage traffic. Such features are typically deployed with support from the underlying operating system. But chips like the LX710 can provide granular control over inter-rupts such as interrupt moderation rate and message signaled interrupts to lower the overhead on CPU activity and streamline memory activity in conjunction with Win-dows features like receive side scaling (RSS), RSS receive queues and receive side coalescing. Chip-level capabilities like native quality of service and data center bridging help ensure that traffic types receive the appropriate priority—especially with storage network traffic.

Ethernet controllers also support virtualization and cloud functionality directly in the silicon. For example, Intel’s XL710 provides virtual machine device queue to optimize virtual machine data traffic and single root I/O virtualization, which allows a PCIe-based Ethernet con-troller to appear as multiple logical NICs for virtual envi-ronments. Other on-chip features like support for virtual extensible LANs, network virtualization using generic routing encapsulation and the IETF’s Geneve protocol all allow for network integration and support for cloud computing.

BOOSTING ETHERNET FROM THE SOFTWARE SIDE

The future of Ethernet will also be influenced by software developments that seek to enhance network efficiency and improve how network resources are configured and pro-visioned. A big part of this includes a type of software-de-fined networking (SDN) that allows administrators

to manage network services by separating a network’s control and data planes. “Some of the technologies we’re considering include software-defined networking as part of an OpenStack initiative to speed up server platform and application delivery,” said Tim Noble, IT director and advisory board member at ReachIPS.

In effect, the network’s switch ports (data plane) are separated from the logic that decides how to direct the traffic (control plane). The control plane is relocated into software that network managers can configure and control directly, also enabling them to make dynamic decisions about traffic flow based on traffic types and activity patterns. This is much easier than manually recon-figuring traditional switches as networking needs change. As control changes, new rules are communicated to the switches, so SDN promises to keep the network simplified and optimized. “SDN is one of the most exciting areas [of Ethernet development], since the promise of being able to deliver IT infrastructure faster is clearly needed in conjunction with other agile delivery approaches used by the development community,” Noble said.

Other software developments include hypervisor-based network virtualization like VMware’s NSX, which uses a

“ SDN IS ONE OF THE MOST EXCITING AREAS [OF ETHERNET DEVELOPMENT].” —Tim Noble, ReachIPS


Home

Editor’s Letter











network hypervisor to abstract Ethernet traffic from the underlying physical network. Just as server hypervisors convert CPU and memory into computing resources which can be allocated to workloads, network virtualiza-tion allows data centers managers to treat the physical network as a pool of transport resources that can be man-aged and allocated to VMs along with security and other services.

The concept of network functions virtualization (NFV) uses a foundation of server virtualization to simulate net-work components (such as switches) within server-based VMs. The idea is to create network building blocks to which administrators can connect and then manage to create complex, highly-tailored virtual networks. NFV supports a range of virtual network functionality includ-ing firewalls, load balancers, intrusion detection, WAN accelerators and so on. NFV is still in its infancy, but is poised to develop as interest in SDN and network virtual-ization technologies continue to evolve.

One other push toward better network efficiency from the software side is the emergence of HTTP/2 as an up-grade to the venerable HTTP 1.1. Moving to a binary-based protocol and multiplexing TCP connections promises more effective use of available Ethernet bandwidth and faster performance for interactive or content-rich websites. “I love the binary protocol for slimmer, more

efficient, less error-prone delivery as well as a true mul-tiplexed connection,” Noble said. “No more waiting on a slow CDN [content delivery network] service. However, the encrypted payloads will bring new challenges to net-work and IT operations teams.”

Soon, businesses will be able to select an expanding range of bandwidth standards to meet their needs while still limiting cost and risk. At the same time, new protocols and virtualization platforms like SDN, NFV and others will allow network administrators to create and optimize Ethernet networks on the fly to suit the most demanding enterprise workloads. A Cisco staffer once said that only cockroaches and Ethernet will survive the nuclear holo-caust—they may just be right n

STEPHEN J. BIGELOW is a senior technology editor in the Data Center and Virtualization media group at TechTarget. Email him at [email protected].

A CISCO STAFFER ONCE SAID THAT ONLY COCKROACHES AND ETHERNET WILL SURVIVE THE NUCLEAR HOLO CAUST.



w

Overheard @ ChefCon 2015Home

Editor’s Letter











“ If you’re playing Dungeons & Dragons and you have five fighters heading in to the Tomb of Horrors, you know you’re going to die in a minute.” ADAM JACOB, Chef CTO, on the value of diversity in DevOps teams

“ The goal of change management is to make sure nothing ever changes.” JEZ HUMBLE, Chef vice president

“ We’re not generalists—we’re hyper-connected specialists.”ADAM JACOB, Chef CTO, on what makes a DevOps engineer

“ You didn’t ask about Docker. In Silicon Valley, no conversation is complete until you talk about what it means for Docker.” AMULYA SHARMA, DevOps engineer, GE Software

“ A human is at best Three Sigma, never Six Sigma. That means they can only be right 97% of the time.” MATT MERCHANT, global managing director, DevOps for GE Capital, on the value of automating security and compliance policies with DevOps.


Home

Editor’s Letter











MANY PEOPLE THINK that IT infrastructure provides a sup-porting role to the business. Like indoor plumbing, and unlike cool cutting-edge applications, infrastructure is seen as necessary and even critical, but not something that usually provides unique differentiation and competitive value. But that’s about to change, as IT starts implement-ing more “data-aware” storage in the data center.

As IT shops grow to become service providers to their businesses, they come to pay more attention to what is actually valuable to the businesses they serve. But when business folks are asked directly about what IT should and could do for them, they can list out confused, contrary and naïve desires that have little to do with infrastructure (assuming minimum service levels are met). Still, the best IT shops are finding that by looking deeper at what their infrastructure can do “autonomically” they can deliver greater value to the business.

Today, IT storage infrastructure is becoming smarter about the data it’s holding. New types of processing ca-pabilities from big data solutions provide the motivation to see what can be done with formerly disregarded data sets. Technological resources are getting denser and more powerful—convergence is the new buzzword across in-frastructure layers—and core storage is not only getting much faster with flash and in-memory approaches, but can take advantage of a glut of CPU power to locally per-form additional tasks. Storage-side processing isn’t just for accelerating latency-sensitive financial applications anymore, but thanks to new kinds of metadata analysis, can help IT create valuable new data services.

In the past, metadata (i.e., data about data) was used primarily for ensuring ownership and securing access to important files, and in more object-based archives, to en-force longer term data retention policies (keep for at least X years, delete after Y years). If we wanted to know any-thing else about masses of data, we often had to process it all directly. In fact, this became one of the motivations for the scale-out Hadoop/HDFS architecture. But now all of our data is growing bigger and bigger—more objects, files and versions, larger data sets, increased variety in struc-ture and format, and new sources of data arriving daily. Instead of powering through all the bits of our growing data every time we want to know something, it can really pay off to produce and keep more metadata about the data

THE NEXT BIG THING

It’s Alive!Storage isn’t just a bunch of dumb disks anymore. BY MIKE MATCHETT


Home

Editor’s Letter











we are storing and use that distilled info instead. Putting these thoughts together—new forms of storage

can automatically create more kinds of metadata about the data it is storing, and then use that metadata to di-rectly provide not only intelligent, but also fast and highly efficient data services. Here are some examples of what can be done today by some currently available storage solutions:

n Fine-grained Quality of Service. A host system can provide certain metadata at the object/file level which then can direct storage infrastructure (e.g., array, storage network) to independently ensure delivery at a preferred level of performance. Oracle’s FS1 array, for example, has a Dynamic QoS Tiering capability that can track which data bits should receive priority service and flash acceleration on a file-by-file basis. With this detailed information, important files in Oracle databases and applications effi-ciently and automatically receive the best-aligned storage services for optimal database performance.

n Fine-grained Data Protection—In addition to perfor-mance, metadata can also ensure fine-grained data protec-tion. As another example, an evolving application-aware paradigm in virtualization environments is to provision storage at the virtual-machine level. If the storage array supports the hypervisor’s application programming inter-faces (e.g., Tintri, VMware VVols), storage receives meta-data that it uses to provide and enforce per-VM storage policies, for example minimum RAID type or number of required copies.

n Content Indexing and Search—For large data stores of unstructured “text-ful” data, indexing all the content to create a metadata database to power search capabil-ities provides a great opportunity to derive value out of data that might otherwise just occupy space. In the past, active archives performed searches on aging static data, but today data can be indexed as it is ingested in to the primary storage. Examples include solutions like Tarmin GridBank, or the increase in use of search engines like Lucene/Solr (e.g., LucidWorks) across large data stores outside of any specific application development efforts.

n Social Media Analysis—One can create metadata that tracks which user has accessed and/or edited each piece of data over time. Users can find out who in the organization has had an interest or concern in any given identified con-tent, look for group collaboration patterns or generate rec-ommendations based on finding content that other users have also accessed. As an example, Data Gravity’s storage puts their high availability passive secondary controller to work maintaining and serving this kind of analysis of user/usage based metadata.

n Active Capacity and Utilization Management—IT admins can see deep into dynamic storage infrastructure behavior when metadata statistics are created that include resource utilization metrics, client file operations, IOPS, and other system management metrics. Recently an-nounced Qumulo, for example, can help admins see what is actively going on in the storage system down to the file level. This makes it easy to see which files and directories


Home

Editor’s Letter











are or have been hot at different times and which clients are hitting which sections of the file structure across bil-lions of files. The storage is actually helping analyze its own behavior and alignment to workloads.

n Analytics and Machine Learning—With the growing computing power found in modern arrays, data process-ing and analytics tasks can be hosted directly in the array to run extremely close to data. As mentioned above, the idea driving Hadoop and HDFS is to localize processing (through massive parallelization) over big data sets. But we also see emerging technologies that bake advanced sta-tistical analytics and even machine learning into a wider swath of more common storage infrastructure. Much like search, we foresee advanced metadata extraction and

transformation that will allow us to automatically catego-rize, transform, classify, score, visualize and report on our data, simply by storing it in the right place). In the future, maybe our data will even tell us what we need to do with it.

Metadata can help make our infrastructure act intel-ligently about the data it’s holding. Infrastructure that becomes both more data- and self-aware could help us stay on top of our growing data challenges. We think that having IT simply storing bits blindly in the face of such data growth is a fool’s game, and the real value of having all this new data is to leverage as much out of it as we can.

MIKE MATCHETT is a senior analyst and consultant at Taneja Group. Email him at [email protected].



Home

Editor’s Letter











INFORMATION TECHNOLOGY IS one of those fields that is always changing. There are, and always have been, new ideas and products to change the way we do business. Experienced industry folks sometimes say the industry is moving faster now than it ever has previously. That’s probably true, but I also think the Internet makes it seem like things are moving fast because news about advances travels almost instantaneously at this point.

Security is one of the most rapidly changing areas, espe-cially given the massive security breaches such as those at Target and Home Depot, as well as high-profile software bugs in OpenSSL and NTP. As I’ve thought about security

I’ve realized that it’s the one thing that doesn’t ever seem to change in IT. We never have enough security. The secu-rity we have is applied in the wrong places. And vendors don’t make it easy to do the right thing when it comes to security.

Security isn’t a priority, and remains almost an after-thought. Budgetary restrictions are often blamed for security shortfalls, but there may be more to the matter than that simple reasoning. We select products based on features and functions, glamorized by sales folks. Security issues aren’t considered during the purchasing process; it’s often looked at in hindsight. At that point all we can do is hide the system behind a firewall. Maybe we make the firewall more intelligent to discern good traffic from bad. We limit access to the system by users, and obscure the name of the system in hopes that bad actors don’t find it, or don’t know what it is when they do.

All of these are bandages, applied mostly to demon-strate a good faith effort to auditors. But these security controls are in the wrong place, and don’t address the real problem. The real issue is that software has no security. Software developers don’t care about security because the people that hire them don’t think about security. Specifica-tions are written by non-technical folks who don’t include anything about security, don’t specify encryption and don’t make user security a priority (namely decent passwords and two-factor authentication). And implementations

IN THE MIX

The Dismal State of IT SecurityThe one constant in IT has always been security—or a lack thereof. BY BOB PLANKERS


Home

Editor’s Letter











are done by developers who are not properly trained in the technology, especially encryption which is hard to get right.

A great example of this is my bank, which doesn’t let me use a password with a semicolon, percent sign or a space. Why not? They’re likely trying to protect against a type of attack called SQL injection, which uses these characters to subvert the way the application works with the database it uses. What it really means, though, is that their application has weak security. They are weakening my passwords and compromising my safety because they don’t sanitize their application’s input.

There are thousands of examples of bad security baked into products and services. Our data centers are filled with products that ship with default passwords, all communi-cation protocols enabled, and no firewalling or IP-level access control. I applaud the few vendors that are trying to make it easier to secure their products and those that ship their products in more secure configurations. We need more of this, and we need to call attention to it. Vendors should always be working to make security easier to im-plement, at all levels of our infrastructure.

As an industry we also need to insist on security up-front. Some of the first questions to salespeople should be about their security features, things like encryption of data, password handling, two-factor authentication and

IP restrictions. We need to insist on real answers, too, and not the typical answers from sales folks who have likely never been asked these questions. Finally, we need to stop buying and using products that don’t have security fea-tures baked in. In the end the bandaging we’re doing now isn’t scalable or supportable—we need to fix our security problems at their source. n

BOB PLANKERS is a virtualization and cloud architect at a major Midwestern university.

WE NEED TO STOP BUYING AND USING PRODUCTS THAT DON’T HAVE SECURITY FEA TURES BAKED IN.


I BEGAN MY IT career 20 years ago this month doing desktop support, and by 1997 I was using Citrix to centrally deliver and manage Windows applications and desktops. Over the years I progressed into SMS (now SCCM), Remote Desk-top Session Host (RDSH), VDI and desktop virtualization, but my core focus has always been Windows applications and desktops.

That didn’t change even as the iPhone came out in 2007 and the iPad hit in 2010. Even then, I viewed mobile apps as someone else’s problem. Sure, Citrix, VMware and, to some extent, Microsoft all entered the mobile application space with various products to manage mobile applica-tions and devices, but I always viewed those as something different and separate from Windows applications.

But is it time for my attitude to change?As long as there have been iOS and Android devices,

we’ve delivered Windows desktops and apps to them via Citrix and VMware technologies. But I always considered that an edge case, because desktop apps on mobile devices aren’t as useable as desktop apps on desktop devices.

But my blasé attitude towards mobile devices is only possible because of my job as an industry watcher. Ac-tual IT professionals responsible for supporting real end users don’t have the same luxury. Their position is: “Our users need these apps, and they want to use them from mobile devices, so we have to figure out how to make that happen.” So IT pros are not so much “desktop admins” as “end-user computing application admins,” and they must support whatever the users and business need.

Two months ago I wrote about app transformation, a technology that can make desktop apps feel more like mobile apps when accessed from mobile devices. But even that is more like traditional RDSH and VDI—it’s about Windows apps on mobile devices. And what about all these other enterprise mobility technologies like mobile device management and mobile application manage-ment? Is it finally time to unify our approach to traditional desktops as well as mobile devices and apps?

I’m starting to think the answer is yes.Both Citrix and VMware have made strides in this di-

rection. Citrix bought Zenprise and has a XenMobile suite

END-USER ADVOCATE

Mobile’s Our Problem NowIs it finally time to unify desktop and mobile application management? BY BRIAN MADDEN

Home

Editor’s Letter












Home

Editor’s Letter











of mobile device and native mobile app management (in addition to its Citrix Receiver clients for mobile devices). VMware acquired AirWatch, which it’s integrating with the rest of its desktop virtualization solutions.

Meanwhile, several newer vendors have entered the market with products aimed at bridging the gap between desktop and mobile environments. For example, a com-pany called Workspot that previously focused just on mobile apps has recently released desktop application support. Now its customers can buy a single tool that de-livers Windows apps to desktop and mobile clients while

also managing mobile devices and apps all through the same system for a more convenient portal.

I’m not sure whether I’d want to use the same vendor for both desktop and mobile app management and/or whether I’d want to use the same product for both. I do know that I want to start thinking about my mobile devices and desktop devices as part of the same conversation. n

BRIAN MADDEN is an opinionated, supertechnical, fiercely indepen-dent desktop virtualization and consumerization expert. Write to him at [email protected].

mailto:bmadden%40techtarget.com?subject=

Home

Editor’s Letter












Modern Infrastructure is a SearchDataCenter.com e-publication.

Margie Semilof, Editorial Director

Alex Barrett, Editor in Chief

Adam Hughes, Managing Editor

Phil Sweeney, Managing Editor

Patrick Hammond, Associate Features Editor

Linda Koury, Director of Online Design

Joe Hebert, Production Editor

Rebecca Kitchens, Publisher, [email protected]

TechTarget, 275 Grove Street, Newton, MA 02466 www.techtarget.com

© 2015 TechTarget Inc. No part of this publication may be transmitted or reproduced in any form or by any means without written permission from the publisher. TechTarget reprints are available through The YGS Group.

About TechTarget: TechTarget publishes media for information technology professionals. More than 100 focused websites enable quick access to a deep store of news, advice and analysis about the technologies, products and processes crucial to your job. Our live and virtual events give you direct access to independent expert commentary and advice. At IT

Knowledge Exchange, our social community, you can get advice and share solutions with peers and experts.

COVER PHOTOGRAPH AND PAGE 3: URBANCOW/ISTOCK

Follow

@ModernInfra

on Twitter!

http://searchdatacenter.techtarget.com/

http://reprints.ygsgroup.com/m/techtarget

https://twitter.com/ModernInfra

MI Modern Infrastructure - docs.media.bitpipe.comdocs.media.bitpipe.com/io_12x/io_123685/item... ·...

Documents

Transcript of MI Modern Infrastructure - docs.media.bitpipe.comdocs.media.bitpipe.com/io_12x/io_123685/item... ·...