Plenty of Amazon Outage Blame to Go Around

Michael Vizard

There's been a lot of pointed criticismleveled at Amazon specifically and cloud computing in general as of late following recent high-profile outages that have affected some fairly well-known sites. Much of the criticism is well-deserved. But what seems to be escaping everyone's mind is some of the questionable decisions that the IT organizations made when ultimately deciding to rely on the Amazon public cloud computing service.

The first issue that comes to mind is how overly dependent these organizations are on Amazon. There is nothing in the annals of computing that would suggest that just because an IT service is in the cloud that it is not subject to the same issues that affect IT infrastructure running on premise. While it's true that one might expect that Amazon would have done a better job of getting customer sites back up and running, the fact of the matter is that IT infrastructure is subject to failure. What is just as surprising as Amazon's inability to quickly resolve the problem is the IT organizations that relied on Amazon didn't appear to have much of a contingency plan in place to deal with such an ordeal.

Abhik Mitra, a product manager for Kroll Ontrack, a provider of data recovery and backup tools, says one of the big issues with any public cloud computing service these days is the assumption that somehow the availability of the public cloud computing service is guaranteed. However, when you look closely at the service-level agreements (SLAs) provided by these providers, they typically say things like 99 percent availability. Well, that 99 percent availability, depending on the size of the overall site, might equal two weeks of actual downtime a year. And as a fellow named Murphy likes to constantly remind us, the odds are that at least one of those days is going to affect your business. The other issue that too many IT organizations ignore, adds Mitra, is making sure recovery that goes beyond a 'best effort' clause is spelled out in detail.

Ultimately, it's your organization's data that is stored on the public cloud service. That means the internal IT organization is still responsible for managing it, including putting in place all the backup and recovery processes required. In an ideal scenario, there would be virtual servers either on premise or in another cloud that would be ready to take over for the affected servers on a moment's notice. Of course, that would mean that IT organizations have a proactive approach to data management in the cloud in place.

There's no doubt that Amazon's woes over the past few days are a black eye for public cloud computing. But in general, most internal IT organizations have more outages than an external provider. Those outages just don't get as much publicity because they don't affect multiple customers at the same time. But when those outages do occur, there should be plenty of blame to go around both inside and outside the IT organization that ultimately selected a public cloud computing service without putting the proper controls in place to deal with availability issues that are bound to occur.
 



Add Comment      Leave a comment on this blog post
Apr 26, 2011 5:04 PM CloudInsure CloudInsure  says:
All true on the above. Outages happen in the Cloud as well and users shouldn't expect perfection. However, traditional internal I.T infrastructure is a risk-centric model. Meaning, the company building, maintaining and storing the I.T. infrastructure also accepts the inherent risk associated with operations. That said, when you move to a cloud - All of those I.T. operations, including the inherent risk assumption, are moved to the Cloud provider - Otherwise, why are there SLA agreements in the first place? (Albeit SLA agreements that don't cover much, if anything in the end of the day in terms of a true and meaningful indemnification for a client) What is the key point missed by most is that a company operating their own I.T. Infrastructure has the ability to move that inherent risk of operation off their balance sheets and onto an Insurance product. The cloud is a different story. When Cloud outages happen, the risk is either 1)fully accepted by the Cloud - or 2)the customer assumes all the risk of failure - And that failure is COSTLY. Since we haven't seen Amazon write any checks - meaning checks covering the economic loss of an outage and not just a refund for using their technology - we are assuming the customer is carrying the risk here. So, despite customers paying to outsource their I.T. infrastructure and the inherent risk along with it, they still have to accept the inherent financial liabilities. Does the pendulum seem off to anyone else? Ask your cloud provider for Insurance - OR a check to cover your operational business losses in the event of their cloud failure. Cloud Insurance does exist for this very reason and can be found at www.cloudinsure.com Reply
Aug 28, 2011 5:08 PM regime hyperprot�ine regime hyperprot�ine  says:
I dugg some of you post as I thought they were very helpful extremely helpful. Reply
Oct 28, 2011 9:10 AM crise de couple crise de couple  says:
Deference to article author, some great selective information. Reply
Nov 25, 2011 2:11 PM comment mincir comment mincir  says:
Magnificent website. A lot of useful information here. I am sending it to some pals ans also sharing in delicious. And of course, thank you to your sweat! Reply
Nov 25, 2011 2:11 PM perdre du poids vite perdre du poids vite  says:
I am happy that I discovered this website, exactly the right info that I was looking for! Reply
Nov 25, 2011 2:11 PM regime hyperproteine regime hyperproteine  says:
I am so happy to read this. This is the kind of manual that needs to be given and not the accidental misinformation that is at the other blogs. Appreciate your sharing this greatest doc. Reply
Feb 28, 2012 11:02 PM Savoir Maigrir Savoir Maigrir  says:
No doubt that Amazon is one of the biggest company online. With their Kindle a new age arrive. Reply

Post a comment

 

 

 

 


(Maximum characters: 1200). You have 1200 characters left.

 

null
null

 

Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.