Availability, Resiliency, and Over-reliance on a (THE) Cloud

[CraginS]

 I have a nifty, rather expensive Roomba robot vacuum cleaner from iRobot, named Darth III. If I want it to do anything other than leap out to clean the entire house, I can only control it with a smartphone app connected through my iRobot account. Early this morning I sent it to clean a group of four tooms while no one was in them. Done. An hour later I tried to send it on another assignment. FAIL! The app said Darth III was not connected to the net. Hmmm. I tried to send a tech support note to iRobot using that feature in the app. FAIL! No connection to the servers.
After several attempts at local fixes, such as rebooting the robot and rebooting my router, I gave up ane called iRobot tech support. The recorded message answering the number was informative: due to a problem at AWS (Amazon Web Services, Amazon's HUGE cloud serice provider operation), the iRobot app was not available. Well dang, that means that all of the high-end Roombas and Bravas (robot mops) are dead until further notice. It is now 8 hours later, and my fancy-dancy vacuum cleaner is still a doorstop (that's an old PC reference for the younguns).

Sure enough, a web search for "AWS outage" tells the basic story:
Amazon Web Services outage takes a portion of the internet down with it
Zack Whittaker@zackwhittaker / 11:32 AM CST•November 25, 2020
Prolonged AWS outage has taken down a big chunk of the internet, recovery may take ‘a few hours’
By Jay Peters@jaypeters  Nov 25, 2020, 12:14pm EST
where I learned,
"Many apps, services, and websites have posted on Twitter about how the AWS outage is affecting them, including 1Password, Acorns, Adobe Spark, Anchor, Autodesk, Capital Gazette, Coinbase, DataCamp, Getaround, Glassdoor, Flickr, iRobot, The Philadelphia Inquirer, Pocket, RadioLab, Roku, RSS Podcasting, Tampa Bay Times, Vonage, The Washington Post, and WNYC."
and
Amazon Web Services  »  Service Health Dashboard

So, lesson time. Do you have local operations or services that cannot work if a remote server outside your control, or cloud service, goes T!t$ Up? Think about it. I do have my Shark in the closet upstairs, ready to go. But, being old and lazy, I  think I will try to wait out Amazon.

Gee, I wonder if the Amazon delivery management system is in that part of AWS that has crashed? If so, gonna be a LOT of Amazon orders late today.

(c) 2020 D. Cragin Shelton

The above originally published on Randomness blog

https://cragins.blogspot.com/2020/11/availability-resiliency-and-over.html

Craig

[Steve-Wilme]

@CraginS Our Roomba is currently sat in the garage unused.  We had a more low tech problem though; our dog took exception to it and would attack it when it ventured out of it's charging station to clean.  

[mgorman]

I agree with the headline overreliance on the cloud.  I do not understand why everyone needs everything to be connected to the cloud.  Not just network capable for updates, or to send metrics or other information, but the d@mn control plane.  I bought a pellet grill earlier this year, and I purposely did not buy the "WiFi" model, as I read one review where someone couldn't start the grill because it was downloading a firmware update. 

If the thing could be enhanced with a remote control function from anywhere (mobile device), then add it, but add it as a proxy where the device is attached and after proper auth, etc. you can reach your device from anywhere.  However, when on the same network, connect locally, so that outages, like the big AWS one, are non events.

And then you get to things like Google Nest.  When Google bought Nest, they shut down the developer API, and there is no local connection. So Google can gather all the temperature, run time, et. data on your system, but the only access to it you have is via their app.  There are many examples of that kind of behavior, too, and it drives me nuts.

Thanx for posting, I didn't even know I needed to vent that,