What do software engineers do during on-call?

Most software engineers know that they are typically tasked with on-call shifts, but new software engineers entering the field may be asking themselves – What do I even do if I get scheduled for an on-call shift? 

This is a common question that often doesn’t get answered until that first on-call shift, and unfortunately that can be overwhelming for a young professional who is nervous about their first on-call shift, let alone their first incident. 

That is why we are writing this blog with an exhaustive representation of on-call responsibilities and best practices, so that you can get a glimpse of what it is like to be on-call and establish peace of mind. 

What are the responsibilities of an on-call software engineer?

At the end of the day, on-call software engineers must be available after hours to ensure continuous coverage of their organization’s tech ecosystem, minimizing losses caused by system downtime, including potential reputational damages. But accomplishing this requires software engineers to take on various responsibilities including: 

Monitoring

Many organizations employ monitoring systems that examine the health and stability of critical systems, and identify any anomalies or vulnerabilities. When these systems are integrated with IT alerting systems, software engineers receive alerts to their smartphones that mobilize them in the event of an incident detected by a monitoring system. 

Incident Response

When an incident does arise and a software engineer is alerted, they must immediately take action. This requires them to acknowledge and investigate incidents to determine their severity, and take the next steps to minimizing their potential impact. 

Troubleshooting

They must also troubleshoot the incident by discovering the root cause of an incident. To identify the causes, software engineers must gather and analyze relevant logs, metrics, and data that can lead them to discovery. 

Documenting

During and after an incident occurs, software engineers must keep detailed notes and records of an incident and the actions they took to resolve it. Not only is this helpful for future incidents, but it improves collaboration and can help with compliance. 

Post-incident Analysis

After an incident, the responders must facilitate a post-incident analysis. This ensures that all of the relevant stakeholders are aware of the incident and can contribute to the continuous improvement of the incident management plan. 

Maintaining clear communication

During an incident, it is important that the rest of the on-call team is informed of the incident and its progress. So, throughout the resolution process, software engineers must deliver incident updates and expected resolution times to improve collaboration and expedite incident resolution. 

Try OnPage for FREE! Request an enterprise free trial.

What will my schedule look like?

Organization’s on-call rotation schedules can vary based on different factors including their size, office hours, and goals. So, it is important for new software engineers to familiarize themself with their organization’s on-call schedule. But in the meantime, these are some of the most common types of on-call rotations: 

Primary and secondary on-call schedules

On-call software engineers should always be available, but in the off chance that they are not when an incident occurs, primary and secondary on-call schedules are there for support. This guardrail ensures that there are always at least two on-call engineers so that if the primary individual misses an alert, then it will be escalated to that secondary one to ensure that the incident is swiftly taken care of. 

Inverse schedule on an escalation policy 

An inverse schedule prioritizes equitability. There are multiple on-call teams that work on a primary and secondary on-call schedule, but the primary and secondary teams alternate, so that one team is not always the first to be alerted. This ensures that on-call engineers do not experience burnout and always respond to critical alerts. 

Follow-the-sun schedules

Larger organizations with multiple locations are able to employ follow-the-sun schedules that allow them to monitor critical systems 24/7 without requiring employees to work after hours. On-call engineers work their typical hours, and then at the end of the work day hand off the on-call work to counterparts at another location in a different time zone.

Expert that is always on-call

Some organizations employ a single expert who knows the ins and outs of an organization’s systems to be on-call everyday. But, this person will only be called in the event of a severe problem that another software engineer cannot fix, so there must also be another on-call schedule in place alongside this one. This ensures that there is always someone to escalate the problem to who will be able to quickly diagnose and resolve severe incidents. 

Who else is involved in after hours incident management?

Though software engineers are oftentimes the main player during an incident, there are many other people involved. Whether it’s communicating outages or delays to clients and stakeholders, or escalating severe incidents to more skilled engineers, you will not have to take on all of the weight during these stressful times. Some of the other people involved during after hours incident management are:  

Secondary on-call engineers

If a software engineer misses an alert or needs a second opinion about an incident, the secondary on-call engineer is contacted. This person is on-call at the same time as the primary engineer, but only takes action when necessary to resolve an incident. 

On-call system experts 

No one person can know everything about a system, so an incident may occur that the on-call software engineer cannot resolve. In this case, they must gather as much information as they can about the incident to share with the system expert, who will be able to solve more technical issues. 

How are on-call software engineers compensated?

You may be wondering how you will be compensated for on-call shifts, considering they are outside of your regular working hours. While there is not one single way that organizations compensate on-call workers, it is important to examine the different ways they are compensated, so that you know what to expect. The following are common ways that software engineers are paid for on-call work: 

On-call pay

Some organizations pay their on-call employees an additional stipend or wage for the entire time that they are on-call. So, whether or not an incident occurs during their shift, the on-call engineer will be paid extra for their time. 

Increased base salary

In some cases, software engineers will not be paid for their on-call services directly, rather they will receive a higher salary to compensate for their on-call responsibilities. 

Per-incident pay 

Software engineers can also be paid per-incident. This compensation method ensures that on-call engineers are paid for their efforts and responsiveness after hours. 

Work flexibility/Remote work

On-call work can disrupt an engineer’s free time and sleep, so many organizations offer a more flexible work schedule to on-call engineers, so that they can improve their work-life balance. Organizations may also employ remote work capabilities to enhance their employee’s satisfaction. 

What are the most common technical problems to encounter on-call? 

You may be nervous about the types of incidents you may have to respond to when on-call. So, we have outlined a few common technical problems that occur after hours to help you prepare for your first on-call shift. 

System downtimes

Unexpected downtimes are likely to occur from failures, network issues, and bugs. It is the responsibility of the on-call software engineer to restore functionality and minimize downtimes so that clients don’t experience prolonged downtimes. 

System failures

If a system fails, it could lead to service disruptions for an organization’s clients. So, software engineers must swiftly troubleshoot and resolve the issue to prevent service disruptions and system degradation. 

Security incidents

On-call engineers may have to respond to security incidents, like unauthorized system access or data breaches that can potentially destroy a system’s integrity. Engineers must act fast to eliminate vulnerabilities and protect critical systems, data, and safety. 

Application bugs

Software engineers may identify anomalies like software bugs that can crash critical applications. It is their job to diagnose and resolve the issue to restore application accessibility and performance. 

What do software engineers do when an incident occurs?

Now that we have outlined the basics of being an on-call software engineer, we will uncover what you will actually do during an incident. This can be stressful, but these steps will help guide you through your first incident, and the many more that will follow: 

Identify

The first step to resolving an incident is identifying an incident. Software engineers typically use monitoring tools for incident identification. These tools can be integrated with incident alert management systems that will immediately deliver high-priority alerts to the engineer’s smartphone when an anomaly or vulnerability is detected. 

Triage

Once an on-call engineer is aware of an incident, they must determine its severity and prioritize accordingly. When prioritizing incidents, it is important to consider the nature of the incident and how it potentially impacts service delivery, client satisfaction, and security. 

Diagnose

To begin resolving an incident, software engineers must diagnose the incident and identify its potential root causes. It is important, at this point, for on-call engineers to update relevant parties about the incident and their findings.   

Escalate

If, after prioritizing and diagnosing an incident, the engineer finds it necessary to escalate the issue they must contact the next level of support. There will be an established escalation path that engineers can follow to ensure that they are contacting the right individual. 

Respond

After identifying, triaging, and diagnosing an incident, the software engineer will begin resolving the incident. This includes taking the necessary action steps to minimize potential impacts and quickly restore normal business operations. 

Notify

As briefly mentioned before, on-call engineers are tasked with updating relevant stakeholders on the incident’s progress. Typically they will use their organization’s incident alert management system to quickly deliver status updates to the necessary team members – management, other on-call staff, system experts, communications team etc. 

Document

Throughout an incident, software engineers will maintain logs and notes about an incident and its resolution. This is incredibly important for teams, so that they can prevent similar incidents from happening in the future. 

Post-incident review

After an incident is resolved, the software engineer will need to conduct a post-incident review. This is a collaborative approach that makes the entire on-call team aware of the situation and facilitates continuous improvement for the future. 

Try OnPage for FREE! Request an enterprise free trial.

What are some challenges faced by on-call software engineers?

Being on-call can be stressful for a variety of reasons and software engineers should be prepared to face a few challenges. So, we wanted to outline a few common challenges you may face and then provide a robust solution. The following are some challenges faced by on-call software engineer: 

Unreliable Alerting Methods 

Some organizations still use legacy technologies, like pagers, that are unreliable and cause for unnecessary added stress. These technologies are ineffective at waking up on-call responders and often experience connectivity issues that prevent engineers from receiving critical alerts. This is why organizations must research and employ reliable and up-to-date alerting solutions. 

Lack of Organizational Support 

Unfortunately, unforeseen incidents occur and there is no one on-call who can solve the problem, and organizations don’t have methods in place to contact an expert. This can be very stressful, because the on-call engineer must wait until office hours for the expert to be on-duty. So, there must be efficient communication plans in place for unexpected incidents. 

Lack of Work-Life Balance

Alert fatigue stems from the absence of alert automation and is a huge problem for on-call engineers, because it can disrupt sleep and cause employee dissatisfaction. Without alert automation, alerts are not prioritized, meaning that engineers will constantly receive unactionable alerts that can lead to frustration, exhaustion, and job dissatisfaction. So, organizations must employ automated systems and provide flexibility for on-call workers experiencing poor work-life balance.

OnPage as a solution

OnPage is a trusted, incident alert management solution that, luckily, addresses all of the challenges listed in the last section. Some of the benefits of OnPage, that can tackle these issues and improve on-call management for both an organization and its on-call employees, are: 

Reliable Alerting

OnPage has a smartphone application that delivers loud, high-priority push notifications that bypass the silent switch, ensuring that on-call engineers always wake up to a critical alert, even when their other notifications are turned off. Additionally, OnPage delivers messages through data and Wi-Fi, and allows visibility into message status, so that message senders always know when an alert was delivered, unlike pagers. 

Role-based Messaging

Organizations must schedule on-call experts for all types of incidents, who will only be called in the event of a severe incident within their specialty. With OnPage’s role-based messaging the primary on-call engineer can directly alert the on-call expert without requiring paper schedules that may not reflect schedule changes, or manually entering contact information. 

Eliminates Alert Fatigue

Automatically prioritizing alerts is crucial for eliminating alert fatigue for on-call engineers. OnPage can prioritize alerts based on specific thresholds, and will only deliver high-priority alerts if those thresholds are reached. This allows on-call engineers to improve work-life balance because they are not worried about receiving a large number of unactionable alerts.

FAQs