Comments
Cloud Migration Turbulence Continues, Survey Says
Newest First  |  Oldest First  |  Threaded View
JoeS227
50%
50%
JoeS227,
User Rank: Apprentice
6/30/2014 | 11:25:29 AM
In regards to the pain point of adequate hands-on support expertise for the cloud
I have been on the incoming tech support call side of a medium size VMWare public, private and hybrid cloud vendor for years.  The most important aspect of any cloud support is making sure you have the right tools and your people have the right training and critical thinking.  Let me elaborate some.

If your cloud tech support group is split into groups that are typical of corporate support infrastructure, then you are going to be inefficient, clumsy and play the typical support group blame game more often than not.  You cannot use tradtional groups that have the blinders on, manage their own little world and have no responsibility to the overall IaaS infrastructure. You will usually find Networking, SAN, OS and Hypervisor (VMWare, KVM/Openstack, etc...) support groups, each with their own monitoring and diagnosis tools, and each passing the tech support ticket around.

To be efficient, and address problems in a timely manner, you have to have a multi-discipline team of 3rd level heavy-hitting support techs. Those level of techs are expensive to hire. They need a CCNA level (at minimum) understanding of networking and routing, they need a solid understanding of the SAN infrastructure and its limitations, and more importantly, the tools to graph the performance and IOPS loads.  They need to be versed in monitoring and diagnosis of hypervisor issues, and be able to understand the trending and history in the vCenter graphs, as well as the Operating System built-in monitoring tools.  They need a medium understanding of Linux and Windows.

Let's talk about tools.  To pull together an efficient cloud troubleshooting team, you need multiple performance monitoring and graphing tools, and, they need to provide trending data for up to three months or more.  These tools need to be available to the team in the form of tools that can be accessed from the desktop, and tools that are displayed on large TV/monitors like a NOC would use.  We use the 42" TV's to monitor the health of our primary data center routers, the overall health of our SAN's and the overall health of our hypervisor clusters and the bare metal they run on.  The techs have access to all those tools and more, but with the ability to go very granular in the data they look at, and the ability to look back at historical trend data.

I simply cannot stress enough how important it is to have access to these tools and the data.  We each use (at least) three large monitors at our workstations, because we need to see the SAN/Volume performance, Network performance, Hypervisor performance and OS perfomance at the same time to make quick, accurate correlations when running down a problem in the cloud.  Some of the tools we use are Orion Solarwinds, Cacti, SAN vendor monitoring/graphing, vCenter, top (linux), task manager (Windows), ESXTOP (hypervisor command line) and an understanding of how to use each one and how to correlate problems seen in one area, to problems cropping up in another area.

Add to this, customer expectations that you know something about databases, IIS, Apache, Exchange, Sharepoint, various CMS systems, Active Directory and firewalls/ports/ACLs, among others, and you can see why it is hard to find cloud troubleshooters that know how to do it right.

 
SachinEE
50%
50%
SachinEE,
User Rank: Ninja
6/28/2014 | 6:08:01 AM
Re: Cloud computing not there yet
Quite a number of enterprise users are reporting unexpected challenges in their move to the cloud and would like more visibility into their workloads. Early customers of cloud services have found that the realities of operating a cloud footprint can include unpleasant surprises. Using the clouds to move workloads must be monitored and managed to work and the clouds are believed to give easy scalability which depends on a watchful eye identifying and correcting problems. Unfortunately, many IT organizations lack the skills they need to perform these tasks. Using clouds is a new technology with a relatively new pricing of models to their mode of operation. Research is still underway to improve this technology and to find out if it is what everyone needs.
DennisD810
50%
50%
DennisD810,
User Rank: Apprentice
6/26/2014 | 10:03:18 AM
Re: What "stalled/failed" means...
Sure, "failed" are clear failures-- where the attempt to move to workloads to the public IaaS provider was abanndoned . "Stalled" represents a delay beyond initial expectations-- which could in turn lead to failure, or an eventual move public IaaS at a later date. Both represent failures of a kind-- in that critical handshake between CSP and IT. Both represent breakages in that ciritcal partnership.
Laurianne
50%
50%
Laurianne,
User Rank: Author
6/25/2014 | 3:46:23 PM
Re: VMware overall, not vCHS in particular
Thanks Dennis. Can you clarify the stalled/failed point as well?
DennisD810
50%
50%
DennisD810,
User Rank: Apprentice
6/25/2014 | 1:57:07 PM
VMware overall, not vCHS in particular
As one of the EMA analysts involved with the research, I would like to clarify that the VMware data refers not to VMware's own cloud service (vCloud Hybrid Services) in particular, but to public IaaS providers in general leveraging VMware.  So the overall positive results there reflect a much broader base. 
pcharles09
50%
50%
pcharles09,
User Rank: Moderator
6/24/2014 | 9:32:43 PM
Re: What "stalled/failed" means...
During initial deplyment, most people will end up in the stalled phase. You typically have to give it some time before you can deem the effort a failure or just give up.
Charlie Babcock
50%
50%
Charlie Babcock,
User Rank: Author
6/24/2014 | 9:22:18 PM
What "stalled/failed" means...
InformationWeek followed up trying to clarify how the "stalled/failure" percentage was derived, what type of question was asked. The answer: "Respondents had the option to select which vendors they tried, including options for adopted, stalled, tried but failed, etc. The numbers reported are the number of stalled/failed vs. adopted for each platform, with 100% representing the total population who engaged with that platform." Still unanswered: how do we know whether some of the "stalled" customers at time of survey didn't continue to try to use the platform and eventually succeed, thus joining the "adopted" %? Wouldn't be unsusal in learning phase.

 


The Business of Going Digital
The Business of Going Digital
Digital business isn't about changing code; it's about changing what legacy sales, distribution, customer service, and product groups do in the new digital age. It's about bringing big data analytics, mobile, social, marketing automation, cloud computing, and the app economy together to launch new products and services. We're seeing new titles in this digital revolution, new responsibilities, new business models, and major shifts in technology spending.
Register for InformationWeek Newsletters
White Papers
Current Issue
InformationWeek Tech Digest - July 22, 2014
Sophisticated attacks demand real-time risk management and continuous monitoring. Here's how federal agencies are meeting that challenge.
Flash Poll
Video
Slideshows
Twitter Feed
InformationWeek Radio
Live Streaming Video
Everything You've Been Told About Mobility Is Wrong
Attend this video symposium with Sean Wisdom, Global Director of Mobility Solutions, and learn about how you can harness powerful new products to mobilize your business potential.