Google's Data Loss: The Corrolary

Ok - the news that has been making the rounds everywhich way since last night is Gmail's loss of data for 150,000 users.  At a glance, 150,000 seems like a large number, maybe it is.  But lets take a look at the bigger reality and more importantly - the ratios, and the null hypothesis.

Gmail has over 100 *MILLION* users.  Lets round it to an even 100 to keep the math round.  So what percentage of 100 million is 150,000?  Answer:  0.15% - or Fifteen HUNDREDTHS of a percent if you prefer..

Or said another way (again keeping it 100 million to keep the math easy)...  that means everyone is chewing Google's ass for *NOT* losing the data of 99,850,000 people!!!

# of people who did NOT lose data:      99,850,000
# of people who did lose data:                  150.000

Get the picture?

They're acting like Cloud computing has completely failed - when in truth, if there was 0.15% loss of data, then conversely...  there was a 99.85% RETENTION of data.  If we wanted to round it up we'd call that 99.9%. Can we REALLY expect perfection when Google when our own personal hard drives fail and we lose data at home?

And since there are MORE than 100 million users, the percentage lost is actually lower than I'm calculating here.

Now consider the amount of stink being generated by 150,000 people.  *IMAGINE* if Google had had a 100% data failure?  Over 100 million people world wide pissed off.... *NOW* that would have been something and it may have really sounded the death knell for cloud computing, and probably the end of Google.

But that isn't what happened - they retained 99.9+% of their data.  And the likelihood that you would have been effected by the data loss is (get your conspiracy theories ready): 1 in 666.

Am I put off cloud computing by the 1 in 666 chance I'll lose my cloud data?  Let's again compare to our desktop computers.  Depending on the brand of hard drive you have, there is a 2 - 14% (statistic comes from PCWorld) that it will fail.  Get this...  even if I use the lowest figure, 2% chance of failure (1 in 50), far more risky than 0.15% (1 in 666).  If you don't get the numbers - try it this way.  Out of every 50 computer users, there is 1 hard drive failure (loss of everything stored locally).  Out of 666 Gmail users there was 1 loss of e-mails (loss of only your e-mails).

Using that statisic - it is technically SAFER to have your data in the cloud!!!

And thus we have the current news spirals that creating negative press, when what Google has accomplished is a 99.9% success rate!  Mind boggling!  Consider 99.9% of over 100 million people served successfully.  That is nothing short of miraculous.

But there is a moral to the story - there is NO SUCH THING as a perfect data storage medium (they all fail -- even paper decays and burns!) - ergo, conventional wisdom still applies...  back up your data!  Gmail supports POP, install Thunderbird (free e-mail client), configure it download but not delete.  Then you run it periodically and you mail gets backed up to your local drive.  Ta da.  Free Gmail backup, provided by Gmail, with a free e-mail client to take advantage of it.  What more could you ask for?  It's all free!  Like Gmail itself.  Sheesh!

I'll finish with a joke I heard today.  Google is offering everyone who lost data a 100% refund on what they paid to store their e-mails.

Update:  1/3rd of affected people have had their e-mails restored. 50,000 people back in action.  And clearly my guesstimate of 100 million users was WAY low.

"No fix yet, but Google's revised its estimate as to how many users might have been affected by the issue -- 'less than 0.08%' -- which means we're probably looking at closer to 150,000 individuals, rather than 500,000. We're assuming that the revised estimate means that the initial count wasn't precise, and not that customers are ditching Gmail in droves."

No comments: