Showing posts with label uptime. Show all posts
Showing posts with label uptime. Show all posts

Wednesday, September 15, 2010

CloudFucius Confused: Cloud Costing Companies?

Konfuzius-1770 You heard that right.  Yeah, I know – wasn’t cloud computing supposed to save money?  All that economies of scale, pay-as-you-go, shared infrastructure, reduced CapEx and OpEx and so on has gotten real cloudy according to a recent survey.  Research company Vanson Bourne, commissioned by Compuware, found that large European companies are losing £600,000 a year due to poor performance of cloud based applications – that’s close to $1 Million USD!  These are e-commerce sites and Internet based business applications.

They surveyed 300 European IT Directors and found that 57% are slowing down, stopping or even shutting down cloud based applications until they can solve this issue.  They noted that while many organizations have large investments in cloud based systems, cloud performance management may hamper or slow the adoption rate moving forward.  84% said that they expect more accurate SLAs than simple availability especially with business critical apps.  Many respondents have experienced degraded performance to their site when a neighboring customer has a huge surge in demand.  There is finite resources available and when one customer grabs a bunch of it, everyone else can suffer. 

Organizations need to look at end to end performance from an end user perspective and not just what the cloud provider is telling them.  There are many variables when determining availability and sometimes it’s difficult to determine if it’s the cloud provider, the customer’s own infrastructure or the Internet itself.  Compuware's Richard Stone said, "Security concerns are obviously still important but it's clear that performance is now becoming the 'day in, day out' business inhibitor that has to be solved,"…"The good thing is people are aware of the issue and understand that the end-user experience can't be compromised."  In another recent survey from Gomez, statistics show that almost one-third of consumers will abandon a slow site within 1 or 5 seconds.  I could relate this to an impatient society but I’d have to include myself.  That same survey found that 39% said speed rather than functionality was more important to a web site.

And one from Confucius: Better a diamond with a flaw than a pebble without.

ps

The CloudFucius Series: Intro, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21

Related:

Technorati Tags: F5, infrastructure 2.0, integration, cloud connect, Pete Silva, security, business, education, technology, application delivery, cloud, context-aware, infrastructure 2.0, automation, web, internet, blog

twitter: @psilvas 

Digg This

Thursday, August 12, 2010

CloudFucius Counts: Cloud Outages

Konfuzius-1770 Well, maybe CloudFucius doesn’t but a couple websites do.  Matching the same uptime guarantees that you have with your physical infrastructure can be hard to accomplish.  While some companies, mostly smaller enterprises, have put all their systems in the cloud, most deployments are ‘extensions’ of existing infrastructures to provide fault tolerance, disaster recovery and all the other situations where the cloud is advantageous.  We all know that the Internet, in and of itself, can have good days and bad – but we somewhat expect this from time to time.  We understand that there are innumerable events that can drastically effect the internet and most of those are out of our control.  While it may irritate us that we can’t get to our favorite website or work application, we generally move on to the next or wait a few minutes and try again.  However, patience and understanding are usually not the headlines used when describing an IT admin when critical business systems are unavailable.
Uptime can also be a nebulous thing – connection is available but the servers are down; servers up but fiber cut; site available but database/authentication/search or other functionality are unavailable.  We hope that the provider has redundancy and other measures to ensure systems, infrastructure and data centers are available.  We’ve expected this for all the years we’ve put servers in co-lo facilities and hosting data centers and should ask the same ‘availability’ questions of our cloud providers that we’ve demanded from our raised floor vendors. 
Now there are a couple websites where you can check things like cloud outages and cloud security incidents.  CloudFail.net is a site that monitors service updates from the leading cloud vendors and aggregates those feeds into a simple blog format.  It allows you to ‘keep an eye’ on cloud providers and get compiled information regarding outages.  An August 11th entry shows the entire thread of a Google Mail issue, tracking it from the initial 4:10am occurrence to the final resolution alert at 2:48pm.  Cloutage.org is an Open Security Foundation project that documents known and reported incidents with cloud services along with providing current news on cloud security.  They provide a couple charts right from the main page showing the latest Cloud Incidents (outage, dataloss, etc) along with the latest Media Reports surrounding cloud security.  These sites should help cloud prospects along with cloud users stay informed about what happening in the haze.
And one from Confucius: Be not ashamed of mistakes and thus make them crimes.

ps

The CloudFucius Series: Intro, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16

Resources:

Technorati Tags: F5, infrastructure 2.0, integration, cloud connect, Pete Silva, security, business, education, technology, application delivery, cloud, context-aware, web, internet

twitter: @psilvas
Digg This