Run Once Monthly at 3:45 AM | CronBase
0 45 3 l * ? Once a month, on the last day, at 3:45 in the morning.
* In a Nutshell
The cron expression 0 45 3 l * ? runs Once a month, on the last day, at 3:45 in the morning.. This cadence is critical for tasks that only need to occur after all data for the month has been finalized, such as end-of-month financial reconciliation or generating monthly reports. Running on the last day ensures completeness, preventing errors caused by incomplete data sets.
* When to use this
Use 0 45 3 l * ? when a recurring task needs to run Once a month, on the last day, at 3:45 in the morning.. This schedule is commonly associated with end of month and financial processing and monthly schedules workloads. It uses Quartz Scheduler (6–7 Fields) syntax, supported by Unix cron daemons, cloud schedulers such as AWS EventBridge, and container orchestration platforms such as Kubernetes CronJob.
CronBase parses 0 45 3 l * ? using a dialect-aware rules engine that identifies the Quartz Scheduler (6–7 Fields) format, validates field structure against the Quartz Scheduler (6–7 Fields) specification, and produces the translation above. Next run times are calculated by forward-scanning from the current UTC clock. Learn how CronBase works.
Platform Implementations
Bash
Prerequisites
Unix/Linux host with a cron daemon running (vixie-cron, cronie, or systemd-cron). The script at /usr/local/bin/run-task.sh must exist and be executable (chmod +x).
Configuration
Run crontab -e to open the crontab editor. Cron reads the server's local timezone; prefix with CRON_TZ=UTC before the expression to pin to UTC. Always redirect output with >> /var/log/cron-tasks.log 2>&1 to capture both stdout and stderr.
Gotchas
Cron jobs inherit a minimal PATH — use full binary paths or set PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin at the top of the crontab. For sub-hourly schedules, add flock -n /tmp/run-task.lock before the command to skip overlapping runs if the previous execution is still running.
# Add to crontab with: crontab -e
45 3 l * * /usr/local/bin/run-task.sh >> /var/log/cron-tasks.log 2>&1Last verified:
Nodejs
Prerequisites
Node.js 18+ with node-cron installed (npm install node-cron). Add "type": "module" to package.json to use the import syntax shown above.
Configuration
Pass { timezone: 'UTC' } as the third argument to cron.schedule(). Without this option the schedule uses the Node.js process timezone, which shifts during DST transitions when servers are not pinned to UTC.
Gotchas
node-cron uses standard 5-field cron syntax — not Quartz 6-field. If your job runs longer than the schedule interval the next trigger fires while the previous is still executing. Use a boolean guard or a queue to skip concurrent runs rather than relying on the scheduler to enforce it.
import cron from 'node-cron';
cron.schedule('45 3 l * *', async () => {
console.log('[cron] running at', new Date().toISOString());
// your task logic here
}, { timezone: 'UTC' });Last verified:
Python
Prerequisites
Python 3.8+ with apscheduler installed (pip install apscheduler). For Python 3.12+ use apscheduler>=3.10.
Configuration
CronTrigger.from_crontab() accepts a standard 5-field cron string. Always pass timezone='UTC' to both the BlockingScheduler constructor and the trigger to ensure consistent scheduling regardless of the server's locale.
Gotchas
BlockingScheduler.start() blocks the calling thread indefinitely. For a web application, use BackgroundScheduler instead and call scheduler.start() at application startup. APScheduler logs missed fires if the process is stopped and restarted — configure misfire_grace_time (in seconds) to control how late a missed job is allowed to run.
from apscheduler.schedulers.blocking import BlockingScheduler
from apscheduler.triggers.cron import CronTrigger
scheduler = BlockingScheduler(timezone='UTC')
@scheduler.scheduled_job(
CronTrigger.from_crontab('45 3 l * *', timezone='UTC')
)
def run_task() -> None:
print('task running')
# your task logic here
scheduler.start()Last verified:
Golang
Prerequisites
Go 1.18+ and github.com/robfig/cron/v3 (go get github.com/robfig/cron/v3).
Configuration
Always create the scheduler with cron.New(cron.WithLocation(time.UTC)). Without WithLocation, the library defaults to the server's local timezone. Call c.Start() to begin the scheduler, then block with select {} to keep the process alive.
Gotchas
robfig/cron v3 uses standard 5-field cron expressions by default. To add second-level precision (6 fields), pass cron.WithSeconds() to cron.New() — but this changes field positions, so never mix syntaxes in the same scheduler. Always check the error returned by c.AddFunc() — a malformed expression is silently ignored without it.
package main
import (
"fmt"
"time"
"github.com/robfig/cron/v3"
)
func main() {
c := cron.New(cron.WithLocation(time.UTC))
_, err := c.AddFunc("45 3 l * *", func() {
fmt.Println("task running at", time.Now().UTC())
// your task logic here
})
if err != nil {
panic(err)
}
c.Start()
defer c.Stop()
select {}
}Last verified:
Java
Prerequisites
Spring Boot 2.7+ (or Spring Framework 5.3+) with spring-context on the classpath. Annotate your @SpringBootApplication class with @EnableScheduling — without it, @Scheduled methods are silently ignored.
Configuration
Spring @Scheduled uses a 6-field Quartz-style expression: [sec] [min] [hr] [dom] [mon] [dow]. The expression 0 45 3 l * ? is derived from 0 45 3 l * ? — a leading 0 is prepended for the seconds field, and ? replaces the unconstrained day field.
Gotchas
Standard Unix cron has 5 fields; Spring requires 6. Pasting a 5-field expression directly into @Scheduled shifts every field by one and the job misfires silently. Use ? for either dom or dow — Quartz does not allow both to be *.
import org.springframework.scheduling.annotation.Scheduled;
import org.springframework.stereotype.Component;
@Component
public class ScheduledTask {
@Scheduled(cron = "0 45 3 l * ?")
public void runTask() {
System.out.println("task running at: " + java.time.Instant.now());
// your task logic here
}
}Last verified:
Kubernetes
Prerequisites
Kubernetes 1.21+ (CronJob API is GA). kubectl configured with cluster access. Container image must be pullable from within the cluster.
Configuration
Apply with kubectl apply -f cronjob.yaml. Check status with kubectl get cronjobs and inspect run history with kubectl get jobs. concurrencyPolicy: Allow is set because this schedule fires infrequently — parallel runs are acceptable.
Gotchas
Without startingDeadlineSeconds, a missed job (e.g., due to cluster downtime) triggers as soon as the controller recovers. Kubernetes 1.25+ supports timeZone: UTC in the spec to avoid timezone ambiguity. Keep successfulJobsHistoryLimit and failedJobsHistoryLimit low to avoid accumulating stale Job objects in the cluster.
apiVersion: batch/v1
kind: CronJob
metadata:
name: scheduled-task
spec:
schedule: "45 3 l * *"
concurrencyPolicy: Allow
jobTemplate:
spec:
template:
spec:
restartPolicy: OnFailure
containers:
- name: task
image: alpine:3.19
command: ["/bin/sh", "-c", "echo 'task running'"]Last verified:
Frequently Asked Questions
What specific action does the `0 45 3 L * ?` cron schedule perform?
This schedule is configured to execute a task precisely once per month, specifically on the final day of the month, occurring at 3:45 AM. It is designed for monthly recurring operations.
How does this schedule handle timezones and Daylight Saving Time changes?
The schedule operates based on the server's configured timezone. During Daylight Saving Time transitions, the exact execution time might shift by an hour relative to UTC, so always verify based on your specific server's settings.
How can I verify that my task is running according to the `0 45 3 L * ?` schedule?
You can verify the schedule by checking your system's job execution logs. Look for entries corresponding to the scheduled task around 3:45 AM on the last day of each month.
Are there any potential issues or gotchas with running a task on the last day of the month?
The primary consideration is ensuring that all monthly data processing is complete *before* this task runs. If subsequent processes rely on this task's output, confirm that dependencies are met by the last calendar day.
What is a common alternative or variant for this monthly schedule?
A frequent variation involves running the task on the first day of the month instead, ensuring it triggers after the previous month's cycle is fully closed and before the new month's activities begin.
More schedules like this
Explore End of Month →* Try any expression
Standard, Quartz, AWS EventBridge, Jenkins, named schedules (@daily, @hourly…)
* Keep Exploring