Since about 6:30 am EST, the internet infrastructure company, Cloudflare, has been widely malfunctioning. This has impacted the FTT forum. I expect the issue to be resolved some time today.
Bob
Bob
Cloudflare still intermittently down 18 NOV 2025
- Thread starter deluxestogie
- Start date
The underlying problem is (for those of you old enough to remember) a Y2K kind of stupid problem. As the year 2000 approached, two generations of computer programmers suddenly realized that for decades, they had been encoding the dates of transactions using only two digits, like 95 to represent 1995. Doing it that way saved (previously) precious memory. Oops!
Here is Cloudflare's explanation:
It took me over a half-hour to realize that the issue was not just my own computer. My operating system had updated itself shortly before the outage. And my browser had just updated itself. Fortunately, my email server still worked, and I could directly email @Knucklehead, who informed me that he was experiencing the same outage. Since we live in different states, that eliminated a lot of possible explanations. I looked at downdetector.com, but discovered that its own website was also down.
The internet is fragile, with too many single points of failure.
Bob
Here is Cloudflare's explanation:
After the company investigated the “spike in unusual traffic,” Cloudflare’s spokesperson provided a more detailed update, telling Ars, “the root cause of the outage was a configuration file that is automatically generated to manage threat traffic. The file grew beyond an expected size of entries and triggered a crash in the software system that handles traffic for a number of Cloudflare’s services."
It took me over a half-hour to realize that the issue was not just my own computer. My operating system had updated itself shortly before the outage. And my browser had just updated itself. Fortunately, my email server still worked, and I could directly email @Knucklehead, who informed me that he was experiencing the same outage. Since we live in different states, that eliminated a lot of possible explanations. I looked at downdetector.com, but discovered that its own website was also down.
The internet is fragile, with too many single points of failure.
Bob
marksman01
Active Member
The alarms I configured for work notified me first thing this morning. 
Thanks for the update.
Thanks for the update.
Not the net itself, just a lot of stuff people have built on it after listening to marketing departments make promises their engineering teams can't keep (usually because of their financial departments). Building in fault tolerance requires more effort and doing some things "the old fashioned way" that's no longer in vogue.
I've fought this battle numerous times; sometimes I won, sometimes not, and in nearly very case my recommendations were eventually validated. At my last employer I told them we should develop our systems to be cloud provider agnostic, and in fact should run across at least two separate entities - AWS and Google Cloud, for example. Nope! Too difficult, too expensive, everything went into AWS. We spread stuff across multiple regions, but were still heavily dependent on the VA region, because spreading those functions across regions was either "too expensive" or "too difficult". Everything was great, right up until AWS VA region imploded one day. By then it was not my area of control, and the C-suite execs that told me not to build more resiliency in had no memory of ever hearing of a better way. Naturally.
Fast, cheap, good. Pick two.
Not too long ago, NASA proposed the mantra: Better, Faster, Cheaper.
This all gets kind of tiresome. I'm so old that I had to hand-code (in 8080a assembly language, in 1977) the serial port of my first computer, just to connect a serial printer.
Bob
EDIT: Sitting in my study, I can see my faded blue 8080 Programming For Logic Design book (©1976) on a corner shelf.
Last edited:
I loved the 8080. I'll give up my 1979 Intel MCS-80/85 User's Guide when they can pry it from my cold, dead hands. I don't use it any longer, but it's part of a small stack of books that bring back many, many fond memories.
Dale
(who once figured out how to fix a non-erasable 8751 problem by only flipping bits from 1 to 0...)
Downdetector.com: As of 6pm EST 18 NOV 2025.
There are still 360 reports of outages.
Bob