Metrics & Mayhem

Good tools answer the question so fast you forget you were the one who asked it. Dashboards, runbooks and AI incident summaries are all useful, and all quietly retire the questions you used to ask yourself. Curiosity is a muscle: when your systems stop requiring it, you stop exercising it, and a genuinely novel failure then finds you out of practice. This week: why good tooling de-skills incident...

MCP, the Model Context Protocol, is the open standard for connecting AI models to your tools and data. In the first Tech Tuesday, we explain what MCP is, the problem it solves, and where it fits in 2026, in plain terms. Introduced by Anthropic in late 2024 and now stewarded by the Agentic AI Foundation under the Linux Foundation, MCP is fast becoming the way agents reach real systems, which makes...

Automate incident triage, and your best responders quietly lose the reps that kept them sharp, until the automation paradox meets a novel fault. I once watched one of the best incident teams I have worked with get quietly worse over about six weeks, and every change we made was, on paper, an improvement. We automated triage. The routine got handled, fewer people were woken, and the numbers looked...

The launch pre-mortem: you can't promise a quiet launch, but you can give your team the bad night in advance. Tested, green dashboards, everyone is confident. It wobbled forty minutes in, and the team fell apart, not because the wobble was big, but because nobodyhad told them it was coming. A dad tells his ten-year-old before a cup final: " You're going to have to suffer today. The bo...

Sev1, and someone's typing in capitals. This should never have shipped. Who approved this? The channel tightens, the debugging slows, and now everyone's managing the angry person as well as the outage. I heard someone this week say anger almost always hides fear. In therapy, they call it not hysterical but historical. This episode puts that in the war room: the engineer shouting at 2 am is...

I avoided building one dashboard for about a year. The cost-per-request board for a service I already suspected was showing a bad number. Every deferral had a reason. Every reason was a way of not looking. Joseph Campbell: The cave you fear to enter holds the treasure you seek. Every team has a cave, the part of the system nobody instruments because they're afraid of what it'll say. This episode i...

A duty engineer showed me the page that woke her at 3 a.m. It said: Error rate elevated. Three words and a graph. ThenShe lay there guessing which service, how bad, the big one or the noisy one. That gap is the problem. A context-free alert is an open loop, and open loops fill with fear. This episode is about treating an alert as what it actually is: a message to a tired human at theworst hour of...

Most observability work is reactive, not preventive. This Signal Drop is about the most expensive habit in IT operations: treating the response to the incident as thestrategy. It covers why positioning beats reaction, what the unbuilt position actually costs you, and the merge-time habit that lets you stop paying the heroism tax. One idea, one habit, five minutes.

Your reliability metrics don't climb in a straight line, and most of that month-to-month movement is noise wearing the costume of signal. This episode is about why one bad month isn't a failed strategy, and how chasing every wobble quietly wrecks the thing you're trying to fix. The habit to take away: decide what counts as real before you see the number, not after.

Returning after two months of silence. An ended contract in Abu Dhabi, a war that began a week later, and the integrity decision behind the gap. This Signal Drop is about the line you refuse to sell, the small increments of compromise that erode an IT Ops career, and how to name yours before someone else moves it for you.

Deep Dive (28mins): Auto-remediation is replacing the 3 AM scramble. Gartner says 60% of large enterprises will adopt self-healing infrastructure by 2026. PagerDuty'searly adopters are resolving incidents 50% faster. But the DORA 2024 report found AI tooling correlates with worse delivery performance, for the second year running, and the Catchpoint SRE Report showed toil rising, not falling. W...

Most leaders don't fail because they're bad at leading. They fail because they stay in one mode too long. This Signal Drop breaks incident leadership into four practical modes: Direct, Shield, Coach, Delegate. Pick the wrong one and the room fills with noise. Pick the right one, and people can think. Includes a ten-second habit you can use before your next call.

In IT Operations, accountability isn’t blame. It’s ownership. When nobody owns the outcome, decisions wobble, incidents drag, and teams default to theatre. This Signal Drop is about making ownership explicit under pressure: one owner, one outcome, one next check. Blameless doesn’t mean ownerless. Clear accountability creates calm, faster decisions, and better reliability.

In IT Operations, you can get pulled in ten directions by praise, criticism, opinions, and noise. This Signal Drop is about staying steady under pressure by curating your input. Who gets a vote in your head? Who do you trust to tell you the truth? And what data do you use to stay grounded when everyone has an opinion? Clear circle, clear truth sources, clear decisions. More signals soon.

In IT Operations, your team mirrors you under pressure. If you bring chaos, you get chaos back. If you bring a calm signal, you get performance. This Signal Drop is about leading incident calls and high-pressure moments with clarity, pace, and decision support. A ten-second reset can change the whole room. One breath, one headline, one next step. Calm people think. Panicked people guess. Guessing...

Most leaders listen to the loudest voices. In IT Operations, that’s dangerous. The earliest warning signals often sit with the quietest person on the call, a night shift handover, or an engineer who doesn’t feel safe to interrupt. This Signal Drop is about inclusion as a performance control. If your culture listens only to the loudest voices, you’re running production without telemetry. The bill a...

Perfect releases are a comfort blanket. Modern systems fail by default. The real question in IT Ops isn’t how to prevent every outage; it’s how fast you regain control when one happens. In this Signal Drop, I break down the mindset shift from “prevention at all costs” to runtime control: toggles, flags, progressive rollout, and clear decision support. A two-hour rollback is chaos. A 20-second togg...

Why good ops conversations fail: we start halfway through the story and expect everyone to keep up. In IT Operations and observability, this shows up everywhere: incident calls, exec updates, KPI reviews, staffing conversations. Engineers want detail. Leaders want meaning. If you don’t frame the first minute properly, both sides walk away guessing, and under pressure people guess wrong. This Signa...

Did your team actually hear you? Not did you say it. Not did you explain it. But did it land the way you think it did? In this Signal Drop, Allan breaks down why communication fails under pressure in IT Operations, incident response, and leadership. When stakes are high, people don’t listen better — they listen worse. Bias kicks in, details get lost, and teams walk away doing different things. Thi...

Trust isn’t a soft skill in IT Operations. It’s operational. In this Signal Drop, Allan looks at trust through a real-world Ops lens: hiring integrity, vendor behaviour, and why overselling destroys credibility fast. From candidates breaking trust before day one to vendors promising what production can’t deliver, this episode cuts to a simple truth: trust can’t be marketed. It’s earned through con...

A raw Signal Drop recorded on the back of a very early morning, thinking about patience, consistency, and the boring habits that actually keep systems alive. Most outages I have ever seen weren’t surprises. They were someone skipping a simple check because they were chasing something shiny. Reliability comes from doing the basics even when you are tired, bored, annoyed, or ready to throw your lapt...

A brutally honest Signal Drop recorded after a 3 am incident call, when I’d been awake nearly twenty-four hours and my bullshit tolerance was gone. This isn’t the dramatic version of burnout people talk about. It’s the quiet one. The sneaky one. The one that shows up when your work and your life drift out of alignment, and you’re too busy to admit it, nights like this force clarity. Pressure doesn...

A short Signal Drop on cutting the leadership noise and getting back to the work that actually stabilises teams and systems. Real leadership is the quiet, consistent stuff: hygiene, ownership, honest signals. If it doesn’t improve clarity, stability, or reliability, it’s probably theatre. Let the team shine. Hold the line.

A short Signal Drop on pressure, presence, and a moment from Fernando Alonso’s first F1 race that reminded me how fragile everything is. We bury ourselves in overthinking and forget to show up in the moment we’re actually in. Leadership isn’t about control , it’s about clarity, honesty, and being there when it matters. No fluff. Just the signal.

A short introduction to Metrics & Mayhem, what the show is about, why it exists, and what you can expect each week. A simple welcome before we start the weekly Signal Drops and deeper leadership episodes.