Your phone buzzes at 2:47 AM. Your heart rate spikes before your eyes even focus on the screen. Itâs another PagerDuty alertâthe third one this week. You stumble to your laptop, troubleshoot for an hour, and finally get back to bed around 4 AM. Your alarm goes off at 6:30, and youâre expected to be fully functional for a 9 AM meeting.
This is on-call life for thousands of IT professionals. And itâs making people sick.
On-call duty is a necessary reality for anyone managing critical infrastructure. Servers donât care about weekends. Network outages donât wait for business hours. Someone has to be available when things break. But the way most organizations handle on-call has created a mental health crisis thatâs driving experienced professionals out of the industry entirely. Whether youâre a help desk tech just starting out or a seasoned sysadmin, the on-call question will come up.
If youâre carrying a pager (or more likely, a phone that never stops vibrating), this guide is for you. Weâll break down why on-call stress has become so severe, what the research actually says about its effects, and concrete strategies that IT pros are using to protect their wellbeing without abandoning their careers.
Why On-Call Stress Has Reached Crisis Levels
The problem isnât being available for emergencies. Itâs how most organizations have implemented on-call systems.
The Alert Fatigue Problem
Alert fatigue happens when the sheer volume of notifications desensitizes you to all of them. One study found that for every reminder of the same alert, attention by the responder dropped 30%. When everything is urgent, nothing is.
According to PagerDutyâs research, most incident responders receive over 10 alerts per shift. Most of those alerts canât actually be acted on. Thatâs the equivalent of a car alarm that goes off constantly. Eventually, you stop hearing it.
The downstream effect? Important alerts get missed or delayed. Engineers become cynical. And the psychological toll of constant interruption compounds into chronic stress.
Sleep Disruption Isnât Just Inconvenient
Being on-call doesnât just mean occasional late nights. Research published in Sleep Medicine Reviews confirms that on-call work contributes to disrupted sleep, fatigue, and increased risk of incidents. The study notes that sleep disruption from on-call work can impact next-day performanceâand critically, performance immediately after waking from a call.
Think about that. You get paged at 3 AM, spend 45 minutes fixing something, and go back to sleep. But the problem you âsolvedâ at 3:45 AM? You might have introduced new issues because you were operating at reduced cognitive capacity. Itâs a cycle that feeds itself.
The research also shows that on-call work adversely affects the sleep of family members, including partners. So itâs not just your rest being disruptedâitâs everyone in your household.
The Always-On Psychology
Even when the phone doesnât ring, youâre not truly off. The anticipation of a potential alert creates a low-grade stress response that persists throughout your on-call shift. You canât fully relax during dinner. You hesitate to start a movie. You check your phone compulsively, even when it hasnât made a sound.
A 2025 report by Catchpoint found that nearly 70% of SREs indicated on-call stress impacted burnout and attrition. Chronic sleep disturbance, constantly switching between contexts, and psychological pressure were significant factors.
This isnât weakness. Itâs biology. Your nervous system canât distinguish between âmight need to respond to a critical alertâ and other threats. The stress response activates regardless.
What Good On-Call Actually Looks Like
Hereâs the uncomfortable truth: most on-call implementations are broken by design. But organizations that get it right prove it doesnât have to be this way.
Shift Length Matters
Googleâs SRE handbook recommends limiting on-call shifts to 12 hours. Thatâs not arbitraryâitâs based on research showing that longer shifts lead to exhaustion and mistakes.
More importantly, Google sets a target of no more than two distinct incidents per 12-hour shift. If a team is consistently exceeding that, itâs a signal to fix underlying systems, not to just push engineers harder.
Compare that to organizations where on-call means 7 days straight, with no limit on incidents. The difference in outcomes is predictable. This is especially important in DevOps roles where incident response is a core expectation.
Separation of Duties
One major source of on-call stress is combining development work with incident response. When youâre trying to ship features while also watching for alerts, you get constant context switching. Research from Cortex shows this leads to reduced development efficiency, more stress, and pathological work behaviors that affect the whole team.
The recommendation? Keep on-call duties and development duties separate. During your on-call shift, thatâs your primary responsibility. During development time, youâre protected from pages unless thereâs a true escalation.
Fair Compensation
Letâs address the elephant in the room: many IT professionals are expected to be on-call without additional compensation. This is especially common for salaried exempt employees, where companies argue that on-call is just âpart of the job.â
But as Atlassianâs guide on on-call pay notes: âEveryone who is on-call should be compensated, whether theyâre paged or not. Holding a pager is inconvenient. It means your non-work time isnât completely yours. Itâs not fair to ask anyone to do that for free.â
Real-world on-call compensation varies wildly. Some places offer $3/hour flat rates. Others pay $350+ weekly stipends or $100/hour for senior roles. According to incident.ioâs guide, organizations should choose compensation that feels fair based on service criticality and typical incident frequency.
If your organization offers no on-call compensation, thatâs a red flag about how much they value your time and wellbeing.
Practical Strategies for Individual IT Pros
You canât always change organizational policy. But you can implement strategies that reduce the personal toll of on-call duty.
Build Your Off-Switch Routine
The research on on-call work emphasizes the importance of âswitching offâ after a call. This doesnât happen automatically. Your nervous system needs deliberate signals that the threat has passed.
After resolving an incident:
- Write it down immediately. Document what happened and what you did. This externalizes the information so your brain isnât churning on it.
- Set a physical boundary. Close the laptop. Put it in another room if possible.
- Do something incompatible with work. Take a shower. Make tea. Step outside briefly. Physical actions signal to your body that the work context is over.
- Avoid screens for 20-30 minutes before trying to sleep again. The blue light and mental stimulation make it harder for your system to wind down.
These sound simple because they are. The challenge is actually doing them at 4 AM when you just want to collapse.
Protect Your Non-On-Call Time
If youâre on a rotation, your off-rotation time is precious. Treat it that way.
- Actually disconnect. Turn off Slack notifications. Set your status to unavailable. Let your team handle things without you.
- Donât check âjust in case.â This reinforces the always-on stress response. If something needs you, theyâll call.
- Schedule activities that require presence. Dinner with friends. A workout class. Something where you canât have your phone in hand.
One pattern that destroys recovery: feeling guilty about being unavailable. If youâre working a healthy rotation and itâs someone elseâs shift, they have it handled. Thatâs the whole point of rotation.
Negotiate Better Terms
If your on-call situation is unsustainable, you have more leverage than you might think. Experienced engineers who can handle incident response are valuable. Use that.
When negotiating:
- Get specific about compensation. Ask what the on-call policy is during interviews. If thereâs no formal policy, that tells you something.
- Request caps on consecutive days. Best practices recommend avoiding back-to-back shifts.
- Ask about incident targets. Whatâs the expected page volume? If leadership canât answer, the alerting system is probably broken.
Sometimes the best solution is finding an organization with healthier practices. Many companies have dramatically improved their on-call cultureâlook for those in your job search.
Build Recovery Into Your Week
This goes beyond the immediate after-incident routine. You need systematic recovery if on-call is part of your regular responsibilities.
Physical recovery: On-call nights, even without incidents, involve disrupted sleep. Plan for catch-up. If you know youâre on-call Friday-Sunday, donât schedule anything demanding for Monday morning.
Mental recovery: After an intense on-call rotation, your cognitive resources are depleted. Plan some lower-intensity work for the following days. Donât try to do your most challenging tasks when youâre running on fumes. That tricky DNS issue can wait until youâve slept.
Social recovery: On-call isolation is real. You canât commit to plans. You have to bail on things. After your rotation ends, prioritize time with people you care about.
What to Do If Your Organization Wonât Change
Not every company cares about sustainable on-call practices. If youâve raised concerns and nothing changes, you have decisions to make.
Document Everything
Keep records of:
- Alert volumes and times
- Sleep disruption incidents
- Impact on your work quality
- Any health effects youâve noticed
- Requests youâve made and responses received
This documentation serves two purposes. First, it gives you ammunition for conversations with managementââIâve been paged 47 times this month, with 23 of those between midnight and 6 AM.â Second, if you decide to leave, youâll have a clear picture of why.
Understand the Legal Side
The Fair Labor Standards Act has specific provisions about on-call time. The key distinction is whether youâre âengaged to waitâ (must be compensated) versus âwaiting to be engagedâ (may not require compensation).
If youâre required to remain at a specific location or have severe restrictions on what you can do during on-call time, you may legally be entitled to compensation even as a salaried employee. This is worth understanding, especially if youâre being asked to do extensive unpaid on-call work.
Consider the Long-Term Cost
Hereâs a calculation worth doing: what is unsustainable on-call actually costing you?
- Healthcare expenses from stress-related issues
- Career impact from burnout affecting your work quality
- Relationship strain from chronic unavailability
- Lost opportunities from being too exhausted to learn new skills or pursue certifications
Sometimes the right answer is to accept lower total compensation at an organization with healthier practices. A sysadmin role that pays $10K less but doesnât destroy your health might be the better deal.
Building Healthier On-Call Culture (For Managers)
If youâre in a position to influence on-call policy, hereâs what actually works.
Measure What Matters
You canât improve what you donât measure. Track:
- Alert volume per shift: Anything over 2-3 actionable alerts per 12 hours is a red flag.
- Time to acknowledge and resolve: This tells you about system health, not just responder speed.
- Noise ratio: What percentage of alerts required no action? High noise means bad alerting configuration.
- Responder feedback: Ask people how theyâre doing. Actually listen.
Invest in Alert Quality
Every unnecessary alert has a cost. Not just the time to investigate, but the cumulative psychological toll on your team.
Good alerting practices include:
- Alerting only on conditions that require human intervention
- Consolidating related alerts rather than sending dozens of separate notifications
- Regular review and pruning of alerting rules
- Clear runbooks so responders know what to do
Create Real Recovery Time
Compensatory time off after heavy incidents isnât a perk. Itâs basic management. If someone was up until 4 AM handling an outage, expecting them to be in the office at 9 is counterproductive.
Build flexibility into your policies:
- Automatic late starts after overnight incidents
- Additional PTO after particularly rough rotations
- Clear expectations that recovery time is legitimate, not slacking
Lead by Example
If leadership is never on-call, or takes their rotation less seriously than IC engineers, youâre sending a message. The best on-call cultures have managers who participate in rotation and prioritize alert quality because they feel the pain directly.
Signs You Need to Make a Change
Not all on-call situations are salvageable. Watch for these warning signs:
Physical symptoms: Chronic fatigue, frequent illness, sleep disorders, weight changes, or other health impacts that correlate with your on-call schedule.
Mental health deterioration: Persistent anxiety, dread about upcoming shifts, symptoms of burnout, or depression.
Relationship damage: Partners frustrated by your unavailability. Missed important events. Inability to plan anything.
Performance decline: Making more mistakes. Struggling to focus. Dreading work in general, not just on-call. If youâre questioning whether to leave your job, this might be why.
If several of these resonate, the on-call situation is causing real harm. Thatâs not sustainable, and waiting longer wonât make it better.
The Bigger Picture
On-call stress isnât just an individual problem. It reflects how the tech industry often treats human beings as infinitely available resources.
But things are changing. The best organizations have figured out that burnt-out engineers make mistakes, miss alerts, and eventually leave. Healthy on-call practices arenât just nice. Theyâre operationally superior.
When youâre evaluating job opportunities, ask about on-call. When youâre negotiating compensation, include on-call pay. Building skills with tools like Shell Samurai or pursuing DevOps knowledge gives you more options to choose better situations.
Your career matters. But so does your health, your sleep, your relationships, and your capacity to enjoy life outside of work. Sustainable on-call is possible. It just requires organizations that prioritize itâand IT professionals who refuse to accept anything less.
FAQ
How much on-call compensation should I expect?
Compensation varies significantly by role, company, and region. Entry-level positions might offer $50-150 per week of on-call. Senior engineers or those at companies with mature practices might receive $300-500+ weekly, plus additional pay for actual incidents. Some companies offer time-and-a-half for hours worked during incidents, or compensatory time off. If a company offers nothing, thatâs worth factoring into total compensation discussions.
Is on-call required for all IT jobs?
No. Many IT rolesâparticularly in development, QA, project management, and certain specialized positionsâdonât include regular on-call responsibilities. If avoiding on-call is a priority, focus on roles at larger organizations with dedicated operations teams, or positions explicitly scoped to business-hours work. Ask directly during interviews. Our career advice guide covers how to evaluate these tradeoffs.
How do I bring up on-call concerns with my manager?
Come prepared with data. Track your alert volume, incident frequency, and sleep disruption over 2-4 weeks. Frame the conversation around operational outcomes, not just personal comfort: âIâve been averaging 4 incidents per night shift, which is affecting my next-day productivity and increasing error rates.â Propose specific solutionsâbetter alerting, schedule adjustments, or rotation changesârather than just identifying problems.
Whatâs a healthy on-call rotation?
Googleâs SRE practices suggest no more than 25% of time on-call, with maximum 12-hour shifts and a target of 2 or fewer incidents per shift. Many organizations do weekly rotations with 4-6 people, meaning youâre on-call roughly one week per month. Anything more frequent than that, or with dramatically higher incident volume, warrants scrutiny.
Can I refuse on-call duty?
This depends on your employment agreement and local laws. If on-call was clearly part of your job description and compensation package, refusing outright may not be viable. However, you can negotiate: request schedule changes, additional compensation, or reduced frequency. If on-call was added to your role after hiring without corresponding changes, you have stronger ground to push back. Document everything and know your companyâs policies. Cybersecurity roles in particular often have heavy on-call expectations, so factor this into your career planning.