===============================================================
 Log #23   || 11/04/2025 || Work  || Planet Computers Gemini  
===============================================================

We're not ready.

We're not ready for the complexities upon complexities of Cloud-hosted services and garbage when we can
barely manage the legacy client/server model we wallow in now.  We talk a big game about being 'Cloud First (TM), Agile and AI Forward'
but we don't know where our apps live, which hypervisor they're on or even what the backend DB is - "Sequel? No, I think this is the first one
we've used?" (There may or may not be uneasy laughter after this line is delivered). We know nothing of the ports in use, firewall rules
allowing packets, protocols or network segments traversed. We know little to none of this - to say nothing of security - yet we'll happily
send whole environments; VMs, appliances, DBs of all shapes and sizes to someone elses data center where we'll be charged by the Gb as well as the transaction to get it there (as well as more to get it back out) and smile the whole way.  After all...we're Cloud First(TM) now!


We're also not prepared for the fallout of oversubscribed Cloud services.  

Performance issues, you say?  They'll happily sell you a dedicated instance,
calm in the knowledge that you won't come to the conclusion that's exactly what you had just prior to your cloud journey.

Outage or something unavailable?  Why, they'll happily keep you updated with vague, infrequent status page updates, RSS feeds and mass emails.
You may eventually find out root cause from Wired or The Verge.

Finally able to utilize APIs for some of these whizbang cloud services where your data now resides?
Sorry, that endpoint or method is now depreciated.  Also, the handy "Sonic Boom" service is now called 'Octopus Pants' but we aren't telling
anyone officially.


We're in no state to deal with the new ways of shooting ourselves in the foot.  Then stepping on a rake.  Then falling into an open manhole.

By:
Leaving some block or object storage open to the world
Assuming something is backed up because 'it's in the Cloud' (Spoiler: It isn't)
Being affected by a Cloud outage due to only having presence in one region/availability zone/tenant
Provisioning a compute instance with specs designed to fight God then discovering it will cost your entire dept's budget to run
Not being based near an ISP that can directly connect your business to The Cloud, leaving you to fight for bandwidth
Daring to copy data *back* from The Cloud, assuming it would be free of charge
Utilizing an AI or LLM assistant or chatbot who will willingly give away corporate or financial info when presented with the correct
prompts.

That's not to say operating on-prem is paradise, but better the devil you know. ...And the hardware you actually own.