Student pilots learn the rhythm of a checklist early: read the item, check the thing, say what you found, move on. Nobody trusts memory when the cost of a missed item is high. The people who keep an organisation’s computers and services running use the same instinct, and the discipline they follow is called ITOps. This post turns it into an audit that fits a quiet afternoon, written for pilots-in-training, flying clubs and small training outfits who suddenly find themselves in charge of booking systems, mail, records and a shared drive.
What an ITOps audit is actually checking
According to the Ops Atlas entry on ITOps, the function exists so services stay up: it takes in requests and incidents, hunts down why faults keep returning, governs change and keeps the estate map honest. That is four jobs, and the audit below is built around them. The same entry separates the function (the people and roles) from the practices they follow and from the technical tools that watch the machines. Keep that split in mind, because most weak setups are weak in one of the three, not all of them.
Run the audit with a pen, not a keyboard. For each item, write one of three marks: yes, partly, no. Then, and this is the part people skip, write a date next to every partly and no. An unmarked gap is just a worry; a gap with a date is a job.
Section one: the front door
In a cockpit you always know which radio frequency to use for which job. In IT you need the same clarity about where a problem gets reported.
- Is there exactly one place where people report a fault or ask for something, whether that is a shared mailbox, a form or a ticket tool?
- Does every report get a reference the sender can quote later?
- Can you tell a fault (a thing that worked yesterday and is broken today) from a request (something new that somebody wants)? The Atlas treats these as separate practices, and the handling really is different.
- Is somebody named as the person who looks at the queue each morning, and who covers when they are away?
If a student has to know whom to text, you do not have a front door; you have a rumour. Mark it no and write a date.
Section two: priorities and promises
Not every fault deserves the same speed. A failed booking page the night before a check ride matters more than a broken colour on a newsletter. The entry describes a working example in which priority comes from combining impact and urgency, and where the clock for each priority only runs during agreed hours. The notable lesson in that example is that one ticket can sit open for more than ninety wall-clock hours while only a handful of those hours are charged to its target, since weekends and public holidays freeze the timer.
- Do you have a short list of priorities, each with a target time, written where users can see it?
- Do you know when the clock runs and when it stops? A target nobody can explain is a source of arguments.
- If a supplier is involved, such as an internet carrier, is their promise to you written down, and does your promise to students leave room for it?
That last point is simple arithmetic. If you promise a student portal repair in eight hours and the carrier only promises four hours to you, the remaining time has to cover your own triage, diagnosis and some spare minutes. The Atlas sets out exactly this kind of budget, and doing the sum on paper once is worth more than any reassuring sentence in a brochure.
Section three: what changed on Tuesday
Aircraft maintenance depends on a record of every change made to a machine. Systems deserve the same. A large share of outages follow a change somebody forgot to mention.
- Are changes written down before they happen, with a date, an owner and a way back if they go wrong?
- Are routine, low-risk changes (adding a user, renewing a certificate) handled by a pre-agreed recipe rather than a meeting?
- Are unusual or risky changes looked at by a second person?
- Is there an emergency route for urgent fixes, with a written review afterwards?
- Does one calendar show planned changes, so two jobs do not collide with exam week or a fuel-price update day?
The Atlas notes that the common myth, that a board must approve every change, is wrong: pre-authorised standard changes are a normal part of the framework. It also cites research suggesting that outside approval has no demonstrated effect on how often changes fail, while peer review does help. For a small training organisation, that translates into a friendly rule: nobody makes a risky change alone, and nobody waits a week to rotate a password.
One more warning from the same entry deserves a place on the checklist. After a bad week the reflex is more forms. The Atlas lists adding process in response to a failure as a known trap. Instead, ask what the failure taught you and change the smallest thing that would have prevented it.
Section four: the inventory of what you own
You cannot maintain what you cannot list. Every maintenance organisation keeps a parts register; every IT function needs an equivalent.
- Is there a list of laptops, tablets, servers, subscriptions and domain names?
- Does each item have an owner, a location or account, and a renewal or end-of-life date?
- When did somebody last compare the list with reality? The entry gives a tiny health check as a model: of ten records, check how many are complete, how many are stale, which have no owner and which are duplicates.
- Is the list limited to what you actually make decisions from? The Atlas is clear that a register containing everything is a common anti-pattern.
Try the ten-record test yourself. Pick ten items at random, find each one in the real world, and count the mismatches. Even a tidy school will usually find one or two. The number is not an accusation; it is the starting figure you will improve.
Section five: the silent alarm
Monitoring is the part most small teams skip, because nothing seems to be wrong. That is exactly the trap. A panel that nobody watches is as good as no panel.
- Is there an automatic check that tells you when the booking site, mail or file storage is down, before a student tells you?
- Where do those alerts go, and is a person actually notified?
- Is the alerting itself treated as production? The Atlas warns that a broken event layer hides outages precisely because it fails quietly.
- Do grouped alerts open one incident rather than thirty? One worked lab in the entry takes thirteen alert rows and turns them into two incidents within a fixed window, then shows later arrivals opening fresh ones.
Test the alarm once a quarter by deliberately tripping it on something harmless, and write down how long it took before a human noticed.
Section six: the paper trail
- When a fix works, are the steps recorded somewhere searchable, not in one person’s head?
- Are repeat faults counted? The same fault three times in a month is a problem to solve once, not an incident to fix three times.
- Do you have a standing half hour each month to read the list of repeats and pick one to remove?
The entry draws a line here too: methods for building a knowledge base sit outside the core function, but the notes the team uses every day inside its ticket tool belong to it. For a flying club, a shared page titled “when the booking page is down” does more good than a perfect manual nobody wrote.
Scoring yourself and picking three jobs
Count your marks. Do not chase a perfect score. Choose the three items with the highest consequences for students and instructors, write an owner and a date beside each, and put the next audit in the calendar for three months out. If you want company while you do it, our earlier walkthrough on preparing methodically for pilot aptitude tests shows the same habit of small, scheduled practice, and the guide to sitting aviation entrance exams is a good reminder that process beats panic whatever the field.
A last practical point. The Atlas also notes the opposite failure: heavy process pushed onto one small team looking after one service. If you are two people and a website, skip the committee: bring a notebook, a diary and the nerve to tick “no” honestly. Do that every quarter and the systems will stay as dependable as the pre-flight habit you are trying to build in the cockpit.

