May 11, 2011
Let’s pretend for a moment that you are the owner or technical lead on a web application that recently captured the public’s attention and swelled in popularity—to the point that serving those visitors or customers without building a new data center would be impossible.
Instead of resorting to that up-front investment, what you need, of course, is infrastructure that will scale with your traffic or transactions—a resource that will allow you to avoid the cost and expertise required to maintain a system robust enough to handle the flood. What you need, in other words, is a cloud-based infrastructure provider—preferably one that offers an attractive guarantee of uptime and continual reassurance that no matter what, that data is backed up and replicated to death in the event of disaster.
So, there you have it; you're all set to serve users and wash your hands of the whole infrastructure problem. In one relatively swift move you’ve shed the need for expensive, cumbersome hardware and can roll ahead with your business.
What could be more convenient? The resources simply scale along with demand, you pay for that demand as it happens, and outside of your maintenance of the core business applications, you can sit back and relax.
According to CTO and founder of the cloud management firm Rightscale, Thorsten von Eicken, a former colleague of AWS chief Werner Vogels, this assumption is part of what sparked some of the trouble for those who had their business lifeblood in the cloud. Those guarantees and assurances backing any cloud computing outlet were as good as gold, weren't they?
Following the initial Amazon Web Services outage, von Eicken wrote that although many customers were able to resort to a solid “Plan B” in the case of an outage, some were not adequately prepared for such an event. He claims that “because of Amazon’s reliability has been incredible, many users were not well-prepared leading to widespread outages. Additionally, some users got caught by unforeseen failure modes rendering their failure plans ineffective.”
So is this to say that if you experienced extensive, damaging outages and data loss the burden ultimately falls on you due to lack of disaster recovery plans? Not necessarily. However, what Von Eicken and other notable experts on the cloud movement suggest is that there might have been some over-confidence in Amazon’s ability to take care of everything. That aside, one has to wonder if even AWS as a whole had been a little too confident.
Now let’s go back to the scenario with you at helm from before. You’ve prepared yourself in any way you thought was necessary or appropriate given Amazon’s very solid track record of performance and uptime, the extensive service level agreements (SLAs) and their several success stories of mission-critical cloud operations.
Nonetheless, you woke up this morning (if you were lucky enough to sleep following the outage) to the following message that Amazon sent out to some of its customers today:
A few days ago we sent you an email letting you know that we were working on recovering an inconsistent data snapshot of one or more of your Amazon EBS volumes. We are very sorry, but ultimately our efforts to manually recover your volume were unsuccessful. The hardware failed in such a way that we could not forensically restore the data.
What we were able to recover has been made available via a snapshot, although the data is in such a state that it may have little to no utility...
If you have no need for this snapshot, please delete it to avoid incurring storage charges.
We apologize for this volume loss and any impact to your business.
Amazon Web Services, EBS Support
This message was produced and distributed by Amazon Web Services LLC, 410 Terry Avenue North, Seattle, Washington 98109-5210
You read this a few times. It doesn’t really sink in at first since after all, wasn’t everything protected under, like 50 layers of different protection and duplication efforts on Amazon’s side?
You see key phrases and nothing else… “failed in such a way that we could not forensically restore the data…”
“..data is in such a state that it may have little to no utlilty…”
And your favorite line of all after you’ve had a few moments to really think about it:
“If you have no need for this snapshot, please delete it to avoid incurring storage charges…”
At these lines, you cock a brow over your left eye, which started twitching occasionally a few days ago in the wake of the outage and now appears to be possessed.
Seriously?… This apologetic letter about your complete and total loss of my data and you’re warning me that I am going to be incurring storage charges?
(insert select profanities here)…
While certainly Amazon could have massaged this message, lending a spoonful of sugar for such medicine, this is not the only way that communication has played a significant role in the increasingly bad press the normally stable infrastructure provider is receiving today.
As von Eicken noted of the initial outage, Amazon receives an “F” for its ability to effectively communicate with users throughout the first signs of trouble. Many are now claiming that failure extends to the data loss matter at hand now. At a time like this, however, good communication is needed more than ever—it seems that either they have no idea of the extent and cause of the loss or they are afraid to let people know how bad it is and how far it extends. Either way, this does not bode well for the company as it has opened the door to a bumrush of negative speculation.
The problem is, in Amazon’s defense, this is the first major problem with multiple, compounding failures that it has ever experienced. There have been latencies and delays in select zones in the past but nothing on this level—not even close to it.
In addition to opening the door for widespread criticism and speculation, it has also allowed competitors (not to mention the six or more backup/recovery companies that are rapid-firing press, barely able to contain their excitement over this outage and loss) to claim dominance—to state that they are immune from such disasters.
But we all know, no one is immune. If Amazon suffers this type of problem, Rackspace could suffer the same issue. If Rackspace trips momentarily, so could Microsoft’s services.
And the point that no one brings up here is that this same problem—and worse—could happen in your very own data center if you chose not to hop on the cloud bandwagon. And it could be far more destructive and expensive.
We’ll spare Amazon for a moment and say that this was written in haste. After all, they have been raked over the coals since news of sporadic data destruction broke…and broke in a very public way.
Just as cloud computing hit the mainstream media outlets in a big way over the past year, so too did news of the problems that could arise when you push your core business into the ether.
The announcement today that data was not only lost or temporarily unavailable—that instead it was actually destroyed—certainly doesn’t bode well for the future of mission critical applications being exclusively hosted on cloud computing infrastructure. It is unfortunate for IaaS providers that this should happen right at that much-anticipated golden moment of growing comfort with cloud computing, but perhaps we should consider this event in light of a few points that major media outlets aren’t talking about.
By the way, the media that I’ve encountered this morning paints the picture of what happened with the Amazon cloud in some rather black and white terms. Mainstream outlets necessarily be condemned for this however—after all, it’s not easy to come up with live news that is approachable for the folks that just a few months ago learned that clouds were more than just the puff upstairs while still painting a picture of the outage that is technically dense and thus in better context.
With that said, the media really doesn’t have a leg to stand (nor does anyone else at this point) when Amazon has been (notoriously) uncommunicative about what actually happened. Aside from sporadic updates following the initial outage and some updates that were dense but not necessarily revealing, the public, not to mention the users whose data may have been chewed up, are in the dark.
If an infrastructure provider of any size has a problem like this, the first item on the agenda should be communication. This not only protects them from wild speculation across media outlets, but it also protects the very notion of the cloud as a reasonable solution for everyone—no matter who they choose to rent hardware from.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 10, 2013 |
Australian visual effects company, Animal Logic, is considering a move to the public cloud.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 08, 2013 |
For engineers looking to leverage high-performance computing, the accessibility of a cloud-based approach is a powerful draw, but there are costs that may not be readily apparent.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.