| Patch management across time zones works when one approval policy runs on local schedules: updates and reboots happen at the right local time in every office, third-party apps are patched alongside the operating system, and devices that are offline or off the network catch up when they reconnect, with no VPN required. |
Most patching advice assumes one office, one clock and one maintenance window. Companies with offices in several countries do not have that luxury. A window that is convenient at headquarters can land in the middle of a workday somewhere else, and a window that was right in March can be wrong in November. This guide explains why patch windows break across countries, how to build one policy that still runs on local time in every office, how to prioritise zero-days and CVE scores when offices and time zones differ, and how to keep devices that are off the network in the loop.
Why patch windows break across countries
The usual approach is to pick a window, often at night in the head office’s time zone, and apply it everywhere. That works until the first office in another time zone reboots at two in the afternoon. The problem is not the patches themselves. It is that a single window is always right for one office and wrong for the others.
Daylight saving time makes it worse, because countries do not change their clocks together, or at all. In 2026, the United States moved its clocks forward on 8 March and back on 1 November. Brazil abolished daylight saving time in 2019, Colombia, Peru and Argentina do not use it, and most of Mexico stopped using it in 2022, although border municipalities follow the US schedule. Chile does observe it, on its own calendar, with changes in September and April. The result is that the time difference between a New York office and a São Paulo or Bogotá office is not constant: it changes twice a year, and a window written in one office’s time shifts by an hour for the other.
| Country | Clock changes in 2026 | Effect on a fixed window |
|---|---|---|
| United States | Yes: forward on 8 March, back on 1 November. | Shifts by an hour, twice a year, relative to offices that do not change. |
| Chile | Yes, in most of the country: forward in September, back in April. | Shifts on a different calendar from the US. |
| Mexico | No in most of the country since 2022; border municipalities follow the US. | Can differ between offices within the same country. |
| Brazil | No. Abolished in 2019. | Stable, but its offset to US offices changes twice a year. |
| Colombia, Peru, Argentina | No. | Stable, with the same shifting offset to US offices. |
A fixed window therefore creates four recurring problems:
• Reboots during business hours. The most visible failure: users lose work because a patch restarted their computer in the middle of the afternoon.
• Windows that drift. A window that was safely after hours in one season is an hour closer to the workday in another.
• Offices on different calendars. Weekends, holidays and business hours differ by country, and one schedule cannot respect all of them.
• Reports that cannot be trusted. When some offices are patched late because the window did not fit them, compliance numbers hide the gaps.
Defining the window in UTC does not solve it. It simply moves the mismatch somewhere else, because UTC does not match anyone’s workday.
One policy, local windows
The way out is to separate what should be common from what should be local. Approval rules, patch categories and the list of applications to cover are decisions the company makes once. The clock at which they run is a property of each office. In NinjaOne terms, one policy sets the rules, and each device runs the actions on its own local time.
NinjaOne’s patch policies keep two schedules apart. The scan schedule decides when a device checks for available patches, and the update schedule decides when patches are downloaded and installed. For each time setting in a policy, the administrator can pick a specific time zone or have the device run the action according to its own local time. By default, new policies scan at 8 a.m. and update at 5 p.m. local device time, which is worth adjusting if some of your offices work later than five.
Several other controls matter for multi-country environments:
• Approvals stay central. Patches are approved, rejected or overridden by policy, so one team can govern every office without touching each schedule.
• Reboots can be staggered by local window. NinjaOne describes scheduling patch deployments globally while staggering reboot execution according to local maintenance windows.
• Missed windows are not lost. A device that misses its window is queued for the next approved cycle, and an offline endpoint inherits the next available reboot window when it reconnects.
• Maintenance mode quiets the noise. During patching, a device can enter maintenance mode, which suppresses selected alerts so that planned restarts do not generate false incidents.
• Policies can differ by operating system. Each policy defines its own schedule and settings, so Windows, macOS and Linux devices can follow different rules.
A simple illustration, with hypothetical times, shows how this works. A company has offices in New York, Mexico City, Bogotá, São Paulo and Santiago. One policy scans every Wednesday at 6 p.m. local time, installs updates on Thursday at 8 p.m. local time and restarts at 10 p.m. local time. Every office receives patches at the same local hour, no one does UTC arithmetic, and when the clocks change in the United States or Chile, nothing has to be edited.
| Decision | Set once, globally | Set per office or device |
|---|---|---|
| Which patches are approved | Yes, in the approval rules. | Only for exceptions. |
| Which third-party apps are covered | Yes, in the policy. | No. |
| When to scan and install | No. | Yes, on each device’s local time. |
| When to restart | No. | Yes, on each office’s local window. |
| Who sees the report | Yes, one view for all offices. | No. |
NinjaOne also shares what its customers usually do. Most partners schedule a patch scan once a week, and the most common time to scan or start installing is between 5 and 6 p.m. device time. That is a starting point, not a rule, and the right answer depends on when each office actually finishes its day.

Zero-days and CVE/CVSS scoring
Time zones complicate urgency too. When a serious vulnerability appears, the question is not only how fast to patch, but when, in each office, it is acceptable to do it. Three terms frame the discussion. A CVE is the public identifier of a vulnerability. CVSS is a severity score from 0 to 10. A zero-day is a vulnerability exploited before a fix exists.
CVSS alone is not enough to rank the queue. A high score with no known attacks can wait longer than a medium score that is already being exploited. That is why catalogs of known exploited vulnerabilities, such as the KEV list, add a second signal. NinjaOne correlates each device’s software inventory with CVE and KEV intelligence continuously, without scan windows, and maps each vulnerability to the patch that resolves it. Teams that already run an external scanner such as Tenable or Qualys can import that data instead. If a device is offline when a vulnerability is disclosed, it is matched against its last known software state, and remediation can proceed when it reconnects.
| Situation | Action | Timing across offices |
|---|---|---|
| Listed as known exploited | Treat as urgent and patch outside the normal cycle. | Start immediately; accept an out-of-window restart where the risk justifies it. |
| Critical or high severity, no known exploitation | Accelerate to the next available window. | Use each office’s next local window. |
| Medium or low severity | Follow the normal cycle. | Regular local windows. |
| Zero-day with no patch yet | Mitigate: reduce exposure, restrict access or disable the affected feature. | Apply mitigations everywhere at once, since they do not need a restart. |
When the vendor publishes the fix, NinjaOne’s approval overrides and scheduled deployments let the team move the patch out of the regular cycle. The discipline for a multi-country company is to decide in advance whether an urgent patch follows local windows or overrides them, and to record the exception when it overrides them.
Devices off the network
Offices in several countries also mean employees who travel and work from home, and devices that spend most of their time outside the corporate network. A patching process that depends on being connected to a VPN or a domain controller leaves exactly those devices behind.
NinjaOne uses a cloud-based agent, so a device receives policies and patches over the internet wherever it is, without a VPN, a domain or on-premises infrastructure. For offices with limited bandwidth, patch caching stores Windows, third-party and custom updates on a local machine so other devices fetch them locally instead of each downloading the same patch, while approvals remain centralised.
Three behaviours matter for devices that come and go:
• Missed scans and updates can catch up. A policy can run missed scans or updates as soon as the device comes online.
• Reboot windows are inherited. An offline endpoint takes the next available reboot window once it reconnects.
• Local time follows the device. When a policy uses device-local time, a laptop that travels from one country to another patches on the clock of wherever it is.
The result is that no device is excluded because of where it happens to be, and the report shows which ones are still pending rather than hiding them.
How NinjaOne does it
NinjaOne describes patch management in four steps. Administrators configure agent policies that enable patching, with schedules, automations, restart options and approvals. They deploy the agent to enroll devices. NinjaOne then scans for operating-system and third-party patches on schedule and installs the approved ones. Finally, administrators monitor status, find failed installations and verify compliance through dashboards, logs and reports.
In scope, NinjaOne patches Windows, macOS and Linux, plus more than 6,000 third-party applications, including those installed through WinGet. On Windows policies, Patch Intelligence AI flags unstable updates so the administrator can apply them manually or reject them. An activity log records who approved, rejected or deployed each patch and when, and compliance reports and a dashboard show the state of every office in one place.
Two practical notes are worth keeping in mind. Windows has the broadest management coverage in NinjaOne, while macOS and Linux support the core monitoring, scripting and automation capabilities, so check what you need for each operating system. And on macOS, a restart is required for patches to be installed, so the restart window deserves as much attention as the install window.
Aufiero Informática distributes NinjaOne in Latin America and works with teams that run offices on both sides of these time-zone differences. A patching assessment starts by listing the offices and their time zones, then defines the local windows, designs the policy, pilots it in one office and sets up the report that covers all of them.
Frequently asked questions
Can we approve patches first?
Yes. Approval is set by policy, so patches can be approved automatically, reviewed manually or overridden one by one before they are installed.
Does it work without VPN?
Yes. NinjaOne uses a cloud-based agent, so devices receive policies and patches wherever they are, without a VPN, a domain or on-premises infrastructure.
Does it cover macOS and Linux?
Yes. NinjaOne patches Windows, macOS and Linux, plus more than 6,000 third-party applications. Windows has the broadest management coverage, so check feature support for each operating system you run.
Aufiero Informática distributes NinjaOne in Latin America.

