Amazon is the king of online retail.  In getting there, they have engaged in many anti-competitive practices…to destroy and then become brick-and-mortar retail.

And a really well kept secret – is that retail isn’t Amazon’s most successful business.

Amazon Web Services (AWS) is.  That is richest-man-in-the-history-of-the-planet Amazon owner Jeff Bezos’ cloud computer backup company.

Amazon’s Cloud Is Bigger, More Profitable Than We Think

Amazon Web Services: The Secret to the Online Retailer’s Future Success

Amazon Web Services: Holding The Profitability Bag For Amazon

Amazon Web Services turns a massive profit – which allows Amazon online retail to do obnoxious, no-profit things like this:

“Black Friday in brick-and-mortar retail circles – is the day after Thanksgiving, and marks the start of the Christmas shopping season.  It is called Black Friday because most retailers operate in the red (i.e. at a loss) – all year until then.  It is that day they break through into the black.  So obviously, Christmas for them is life-saving crucial.

“Amazon knows this.  So what do they do?:

“‘Amazon is cutting prices ferociously on its products as Christmas shopping picks up, to levels that are break-even or below.’

“Amazon cuts prices to sell at a loss – right in the heart of retailers’ most important time of the year.

“Lather-rinse-repeat for a quarter century – and Amazon gets to a nearly $900 billion market cap.  They are the biggest of behemoths – crowding out more and more and more competitors.”

So AWS – is funding Amazon’s anti-free-market retail murder spree.

But Amazon and AWS are both having a recurring problem: Keeping themselves online and operational.  Which is…more than a little problematic for an online company.  Just yesterday:

Prime Down: Amazon’s Sale Day Turns into Fail Day:

“Amazon Prime Day started 15 minutes ago, and so far, it’s not going well for Amazon. The landing page for Prime Day does not work. When most links are clicked, visitors are sent to an error page or to a landing page that sends readers back to the main landing page….

“This is a huge blow to Amazon and its faux holiday Prime Day. The retailer has been pushing this event for weeks and there are some great deals to be had. It’s not a good look for the world’s largest retailer even though the retailer saw glitches last year, too….

“Updating…3:30pm EDT: It’s 30 minutes past the launch of Prime Day and the landing page and deal navigation page is still down.”

Get all that?  There were massive Amazon Prime Day offline glitches yesterday – and last year too.

It’s not like the day snuck up on them.  Amazon picked the date.  It’s their Day. They knew it was coming – and for at least two years in a row now…they blew it.

For Amazon – this all by its onesies would be a horrible day.  But they weren’t done screwing up.

Amazon Prime Day Glitches Spread to Amazon Web Services:

“Minutes into the made-up Amazon holiday, users began reporting that they were unable to make purchases on the platform. Amazon, which was predicting record sales for this year’s event, was apparently unable to handle the mass amount of traffic the sale generated.

“That’s bad news for Amazon if they’re using their own Amazon Web Services (AWS) cloud server platform, which boasts the ability to scale for just such an event. But worse, the problem seems to have actually spread to AWS.”

AWS Hits Snag after Amazon Prime Day Downtime:

“After reeling from technical glitches that prevented many shoppers from accessing bargains on Prime Day, the company’s biggest shopping day, the company announced another outage that makes it harder for Amazon Web Services customers to control their usage.”

So Amazon’s two main online businesses – both went offline.  On the same day.  For hours each.

Is this the first time this year AWS has had problems?  Heavens no.  This from April:

Suspicious Event Hijacks Amazon Traffic:

“Amazon lost control of a small number of its cloud services IP addresses for two hours on Tuesday morning when hackers exploited a known Internet-protocol weakness that let them to redirect traffic to rogue destinations. By subverting Amazon’s domain-resolution service, the attackers masqueraded as cryptocurrency website MyEtherWallet.com and stole about $150,000 in digital coins from unwitting end users. They may have targeted other Amazon customers as well.”

But certainly AWS has been golden before 2018, right?  Not so much.  This from February 2017:

Massive Amazon Cloud Service Outage Disrupts Sites

Oops.  AWS seems to have a prolonged, pronounced problem remaining up and running.  This from August 2013:

Amazon Web Services Suffers Outage, Takes Down Vine, Instagram, Others with It

Oops.  And also this from the wayback machine – February 2008:

Amazon Web Services Goes Down, Takes Many Startup Sites With It

Oops.  In fact, besides these periodic massive outages…:

“AWS has what I call ‘micro outages’ every day where 10 of your 100 instances fail to start properly, there’s a network partition for some minutes (your instances can’t all talk to one another) etc.”

AWS outages…”every day.”  At a fail rate of 10%.  Why that seems…more than a little problematic.

And all of this renders obviously stupid – anything like this:

Defense Department About to Give AWS Ten-Year, Monopoly Cloud Contract

The Department of Defense (DoD) is actually considering giving a monopoly contract…to chronically down AWS.  That seems…sub-optimal.

Especially considering the DoD is in all other things – the king of backups and redundancies.

Because much of what they do is life-or-death – you can’t make any…anything they use the sole determiner of life-or-death.

When Amazon Web Services Goes Down, So Does a Lot of the Web

Under this possible contract, when AWS goes down – ALL of the Defense Department goes down too.  Which seems…sub-optimal.

There should be multiple cloud backup providers – constituted relatively thusly:

“Defense should have at least two providers providing each portion of the cloud service – so that if one crashes, you have at least one at-the-ready backup.

“So, say, at a bare minimum: Five providers – each providing 40% of the necessary services. For government school victims – that makes 200%. Which means two providers each are providing every part of the total cloud service.

“Backup. Fail-safe. Redundancy.

“Not one provider – all by its onesies, providing all of the service.”

The Donald Trump Administration now thankfully seems to be stepping back from the brink of this very bad AWS monopoly deal.

Here’s hoping they never again step in that direction.