Amazon, which has had its share of outages (including one that lost customer data), recently had its network go down in Dublin. The company afterward issued a major mea culpa. That's Latin for "Oops, we goofed, and we're really, really sorry."
Like it did after another recent outage, Amazon went to great pains to explain what happened. Last time around, the outage was caused by a network configuration error, where traffic was routed to network connections incapable of handling it all. This time, it turns out an electrical problem -- a down transformer -- was to blame. Then another technical problem stopped a backup generator from kicking in. Without power, the UPSes quickly went down, and that was all she wrote.
Like Microsoft, Amazon is giving customers a rebate equal to 10 days of use, and promises to add more redundancy.
Posted by Doug Barney on 08/30/2011 at 12:47 PM7 comments
Boy, that cloud sure does sound great. All those admins can get pink slips and the air conditioners in the server room can be redeployed at the CEO's summer house. What could possibly go wrong?
But an insightful reader pointed out that if your WAN is slow (and many are), your cloud apps will be sluggish, as well. And as more and more application processing moves over these wires, latency becomes, well, more latent. A 10-gig Ethernet used to be overkill and was needed only by service providers. As the cloud continues to grow, we may need 10 gigs everywhere.
Are you happy with WAN speeds? Let me know at [email protected].
Posted by Doug Barney on 08/30/2011 at 12:47 PM7 comments
Some of the same folks that built Amazon's EC2 are now behind a startup called Nimbula, presumably a takeoff on the word "nimbus." (By the way, there is a storage company called Nimbus and an open source Infrastructure as a Service project also called Nimbus.)
Nimbula hopes to become the Microsoft Windows of the cloud with an operating system called Director 1.5. This newest version is designed to support clouds that are distributed all over the world.
The idea is to give IT and service providers a single view and management of a system that may span many machines, countries and continents. And for the end-user, the distributed system also appears as one single entity.            Â
Posted by Doug Barney on 08/16/2011 at 12:47 PM0 comments
Software licensing is way more complicated than it needs to be, and moving to the cloud, especially using your existing apps, offers a whole new wrinkle.
Microsoft, which is "all in the cloud," wants its customers equally in, and is tweaking its Software Assurance volume licensing program to ease the transition. The basic idea is through "license mobility" you can use what you already paid for to run on your servers and move that software to the cloud.
Analyst firm Directions on Microsoft analyzes license mobility, and their analyst John Cullen spoke to Microsoft watcher and Redmond magazine columnist Mary Jo Foley about all the gory details. Licensing comes easy to Cullen, who for half a decade crafted volume programs in Redmond.
According to Cullen, mobility is an attempt to lure IT to the cloud, but also a lifeline for Software Assurance, which could end up irrelevant as computing shifts off site.
The Microsoft side of the equation is not the most complicated part. The tricky area is continuing to pay Microsoft fees while at the same time negotiating new fees with a hosting company. It is unclear whether, in the final analysis, you'll save or lose money on this deal.
While I may have mentioned the Microsoft side is a bit less hairy than with hosters, it ain't exactly second grade math. Here's an example from Cullen:
"A scenario where you 'win' (licenses let you do more in the cloud than on-premises): We're running one SQL Server workload on a dual proc on-premises server licensed with two SQL Enterprise proc licenses. You can move the workload up to a multitenant hoster with a quad proc box, at times using more proc 'horsepower' than you did when on premises, and yet you only need to allocate ONE of your two SQL Enterprise proc licenses to do so."
Not exactly nuclear science, but not simple either, especially when you have multiple servers, multiple apps and myriad VMs to match. Break out your HP EasyCalc 300 to figure all that out!
Has the cloud made licensing easier or harder? You tell me at [email protected].
Posted by Doug Barney on 08/16/2011 at 12:47 PM1 comments
Very few people I know buy a car without taking a test drive (my mother is an exception here).
Service and communications provider Logicalis clearly wants your business, and wants you to head to either Irvine, Calif. or Farmington Hill, Mich. to land it. That's where you'll find two so-called "Cloud Centers of Excellence."
The idea is that IT can see for itself that cloud infrastructure is not much different from in-house infrastructure: They both have networks, servers and storage (hopefully with fully vetted restore). Besides hoping to make the customer confident in the service provider, the centers are also meant to show that cloud functionality can be built on top of existing IT systems.
Both centers are based on the HP CloudSystem Matrix, a software/hardware suite to build and manage clouds. You can learn more here.
Posted by Doug Barney on 08/02/2011 at 12:47 PM1 comments
Microsoft just released the Microsoft Assessment and Planning Toolkit 6.0, a product formerly aimed largely at migrating desktop and server apps to newer versions of the exact same desktop and server apps. The new kit now works with Office 365, helping IT figure out if their shops have enough computing power to support Office 365.
This to me sounds odd, since cloud apps are supposed to require fewer on-premise desktop and server firepower, not more. In Microsoft's defense, Office 365 does require recent versions of Office -- and recent versions of Office do need some serious PC power.
Azure readiness can also be gauged through the new tool.
Have you checked out either Office 365 or Azure? Send feedback to [email protected].
Posted by Doug Barney on 08/02/2011 at 12:47 PM5 comments
The U.S. General Services Administration (GSA) spends a heckuva a lot of money every year, and now some of that scratch is heading Google's way. The GSA just finished migrating its old e-mail system to Google Apps for Government. The migration included 17,000 users and took half a year.
Now the GSA wants to set the scene for other federal agencies to move to cloud e-mail. Ultimately, nearly a million government mailboxes could migrate to the Internet. This could be a massive test case for the feasibility of large-scale cloud projects.
Has the cloud proven itself to you yet, and if not, why? Shoot your answers to me across the ether to [email protected].
Posted by Doug Barney on 08/02/2011 at 12:47 PM4 comments
OpenStack, an open source platform for building clouds (both public and private), just turned the ripe young age of 1. The party, held in London, was a largely European affair; they may have even served gelato instead of ice cream.
OpenStack has a number of commercial supporters, chief among them Rackspace, Citrix and Dell. It's pretty cool to see a pure open source system gain this kind of traction against multibillion dollar corporations. Can't wait to see what OpenStack does once it learns how to actually walk and talk!
Posted by Doug Barney on 07/26/2011 at 12:47 PM2 comments
Many of you love the cloud and have moved apps and data over. Others are skeptical and curious, which may be why you read this newsletter and blog.
For enterprises, the cloud can be safe if you have the right provider, the right contract provisions, and the right backup and recovery strategy. (Recovery is a lot harder than backup, eh what?) For individuals, the cloud is still a risky thing -- just ask the man who calls himself Thomas Monopoly.
Tom M. trusted Google so much that he moved nearly his entire online/computing life to the Google cloud. Eleven days ago, the automated systems at Google concluded that Tom had violated their terms and conditions.
Instead of a human contacting Tom, a Google robot computer not only shut down Tom's account but purged all the data. We're talking messages, his e-mail address, student records, voicemail, banking data, bookmarks, shared documents, contacts, his calendar, nearly 5,000 videos and pictures, and articles. Oh, and his Web site, for which he paid Google, also went the way of the dodo.
Upset, Tom for weeks had been unable to talk to a real Google person. Finally, he talked to a real human at Google who went to bat for him, but the response was basically that Tom's account was closed and, therefore, so was his case. See ya.
Tom, as far as I know, remains in cloud limbo. Does this scare the bejeepers out of you? Yeses and nos readily accepted at [email protected].
Posted by Doug Barney on 07/26/2011 at 12:47 PM5 comments
It wasn't too long ago that the online version of Exchange went down longer than all of Mike Tyson's opponents put together.
Whoops. Exchange went down again more recently, this time for three hours. That may not seem like much, but the marathon record now stands at under two hours and four minutes. Apocalypse Now, a long movie in its day, was only 153 minutes long, and the 1989 Tour de France was won by only eight seconds.
And if your boss is staring down your neck wondering how much work you're doing, three hours without e-mail is an eternity.
Have you had any cloud outages? And if so, how did you react? Send me an e-mail at [email protected] -- and if it shows up late, I'll blame it on Exchange. (And if you know who beat whom in 1989 and how, also message me at [email protected].)
Posted by Doug Barney on 07/26/2011 at 12:47 PM6 comments
It's no real surprise that virtualization and the cloud are very much related. After all, most cloud services run off of virtualized platforms. This puts VMware in the cat bird's seat, and the company is taking full advantage. Its latest move is vSphere 5, a massive virtualization and cloud suite.
Much of the work focuses on administration. The goal is for IT to set core policies and then have vSphere essentially manage itself from then on. Sounds too good to be true, so we'll be tracking this little promise.
Performance is another key design goal, with vSphere 5 supporting as many as 32 virtual CPUs and VMs able to address a terabyte of memory. That's pretty big-time. VMware's cloud tools now drive off of vSphere 5, including vCloud Director and vCenter Site Recovery Manager.
Posted by Doug Barney on 07/19/2011 at 12:47 PM6 comments
Like VMware, virtualization powerhouse Citrix wants a big chunk of the cloud. (Do clouds have chunks or just wisps?) Citrix has certainly built its share of cloud tools, but now it is also buying its way into the market. Case in point: Cloud.com, which Citrix just picked up for an undisclosed sum.
Cloud.com is in the infrastructure-as-a-service market and its stuff runs on all the major hypervisors. That fits Citrix's strategy to a T as Citrix supports its own Xen hypervisor as well as Hyper-V. Cloud.com's key tool is the open source Java-based CloudStack, which turns datacenters into a service, and self-heals or at least protects itself through robust failover.
Anyone one there ever give CloudStack a whirl? If so let us know how it went by writing [email protected].
Posted by Doug Barney on 07/19/2011 at 12:47 PM1 comments