James Glenos retweeted
It's hard to describe the massive scale of the waves along the Oregon coast this weekend. The people on the cliff on the left side of the video are standing about 75 feet above the water. Just an incredible thing to witness. Keep watching- the 4th wave is the largest! #orwx
James Glenos retweeted
It’s always a great time chatting with @John_Capobianco. Every conversation feels like a profound lesson, from a friend, who is enormously intelligent and passionate.
Keep teaching and inspiring us John. We need you, and more folks like you.
James Glenos retweeted
Roan got it's first significant snow of the season. I'm glad I didn't skip it.
This platform is nuts. See: this thread
Let’s take this gentleman’s argument at face value that anyone can throw money at rockets and get something like Starship, Falcon or Dragon.
Perhaps he can explain to everyone how SpaceX is able to catch giant starships from orbit with chopsticks and deliver 85% of all Earth payload to orbit with Falcon, while Blue Origin, which started before SpaceX and was funded by the immensely wealthy Jeff Bezos, is not yet able to launch reliably?
Also, why is it that Boeing and Lockheed received several times what SpaceX did from the government for development of their rockets and spacecraft, but the Boeing capsule is still not operational and ULA delivers less than 5% of SpaceX payload to orbit?
I am genuinely curious about these questions.
55
James Glenos retweeted
Don't lose your grip on the dreams of the past. You must fight just to keep them alive!
James Glenos retweeted
No one tells you that parenting is just relearning the world through someone who thinks worms are friends & birds are miracles. It’s the most healing thing I’ve ever done. My daughter looked out the window this morning & said, everything is green & growing. I told her, you too. And something inside me whispered, so are you. Now I’m watching her hold flowers up to the sun while the light bends like it recognizes her. It’s funny, every spring I think I’m teaching my child about the world & every spring she proves she’s the one teaching me how to see it.
James Glenos retweeted
IT network guys gonna love this post 👇
My summary attempt at what actually happened:
• The FAA runs its own private telecom network, FTI. It links radar sites, remote radio sites, towers and ATC centers. It's not the internet, though it rides partly on leased carrier fiber.
• L3Harris has operated FTI for 20+ years and is replacing 5,000+ legacy analog circuits. Its site says ~70% of 4,400 sites are done.
• "Copper to fiber" misses the point. The real change is circuit-switched → packet-switched. A remote site used to hang off a dedicated line. Now it gets a router that joins the shared network.
• During one such cutover, that new router began flooding the network: 1,000+ connection attempts every 4–5 seconds.
• Likely mechanism (not yet confirmed): neighboring routers' control planes get overloaded, routing sessions flap and routes churn. Radar and radio feeds riding the network go unstable. Rerouting can't fix it, because rerouting is the part that's broken.
• Nobody could find the source for ~45 min, so the FAA grounded the whole country out of caution. Isolating that one site fixed it.
• No evidence of a cyberattack. Still unknown: a defective router, or a wrong install or configuration?
• It's happened before. In 2009, a misconfigured Harris replacement router disrupted the FAA's flight-plan system.
• ~30% of sites are still to go. The lesson: a single edge device should never be able to destabilize the core. That means control-plane protection, limits on how far one failure can spread, and staged cutovers with a tested rollback.
👀
My summary attempt at what actually happened:
• The FAA runs its own private telecom network, FTI. It links radar sites, remote radio sites, towers and ATC centers. It's not the internet, though it rides partly on leased carrier fiber.
• L3Harris has operated FTI for 20+ years and is replacing 5,000+ legacy analog circuits. Its site says ~70% of 4,400 sites are done.
• "Copper to fiber" misses the point. The real change is circuit-switched → packet-switched. A remote site used to hang off a dedicated line. Now it gets a router that joins the shared network.
• During one such cutover, that new router began flooding the network: 1,000+ connection attempts every 4–5 seconds.
• Likely mechanism (not yet confirmed): neighboring routers' control planes get overloaded, routing sessions flap and routes churn. Radar and radio feeds riding the network go unstable. Rerouting can't fix it, because rerouting is the part that's broken.
• Nobody could find the source for ~45 min, so the FAA grounded the whole country out of caution. Isolating that one site fixed it.
• No evidence of a cyberattack. Still unknown: a defective router, or a wrong install or configuration?
• It's happened before. In 2009, a misconfigured Harris replacement router disrupted the FAA's flight-plan system.
• ~30% of sites are still to go. The lesson: a single edge device should never be able to destabilize the core. That means control-plane protection, limits on how far one failure can spread, and staged cutovers with a tested rollback.
1
131
James Glenos retweeted
Replying to @etscrivner
I joined Microsoft in the 2013–14 timeframe and was immediately struck by the quiet competence that ran through the place. Engineers, security people (including the physical security team), lawyers, developers, and others were simply very good at what they did. In many niche areas, the person sitting in the role was, quite literally, “the” person in the industry for that domain.
Incident and outage calls reflected the same culture. They were calm, quiet, and professional. I still remember jumping onto a Saturday outage bridge and listening to a group of relatively junior engineers work methodically through their runbooks. When they reached the end of the procedures without a resolution, they pivoted without drama into deep technical troubleshooting well beyond the documented steps.
Security incidents were handled the same way. A call might bring 50 people together, yet an L63 or L64 incident manager would run it with clarity—walking through what was known, identifying the gaps, and tasking items across the company.
There were roughly three kinds of people who made this work. Junior college hires were often brilliant, highly motivated, and full of energy. Mid-career people, many with kids and living on the Eastside in places like Sammamish, served as the steady anchors of their teams—engineers, PMs, and others with a decade or more of experience who were simply good at their jobs. Then there were the longer-tenured employees, some with 30 years at the company. Microsoft valued that longevity. These people knew where systems had come from and what was under the hood; they were respected for what they knew, regardless of level, and were regularly pulled into meetings with senior leaders to walk through hard problems.
Even senior leadership operated in the same spirit. I once got a weekend call from a very well-known Microsoft VP. He was direct and courteous: “Is this X? I’m going through this issue and see that you own a document I’d like to read. Can you give me access? Thanks for the help on a weekend—I really appreciate it.”
Around 2020, that culture began to change, and I left not long afterward. I watched the company part ways with many of the people who had solved some of its most significant problems over the previous 20 years. A large number of my former peers and colleagues left as well. It is no longer the company it once was.
James Glenos retweeted
Your whole weekend does NOT need to be productive. You are allowed to just have a damn Saturday.
James Glenos retweeted
Please I'm begging yall switch to DoT or DoH. Stop using plaintext dns (port 53). Your ISP sees every single website (domain) you visit/query
Some network providers block DoT port (853) but DoH is impossible to block.
In terms of providers, you could go with Cloudflare (they claim they wipe their logs every 25 hours) or for better privacy Quad9 (A swiss non-profit) or Mullvad (no logging at all). You gotta trust someone but way better than trusting your ISP
James Glenos retweeted
Test your routing knowledge !
automateyournetwork.com/arca…
@AutomateIP made this as his example game for the upcoming AutoCon6 workshop
It's a blast!
James Glenos retweeted
WOW!
Sky on fire sunrise over Boone, NC with a fresh blanket of snow on the ground!
Frame it!
James Glenos retweeted
.@Cloudflare is the agentic cloud.
They got the primitives right before agents were even a thing:
Workers → execution
Durable Objects → coordination
D1 → persistence
Everyone’s racing to build agents.
Cloudflare already built the ground they’ll run on.





























