Kubernetes cronjobs

You know the scenario, you’re wanting to run jobs either at certain points in a day or throughout the data every N timespans (i.e. every 5 mins).

Kubernetes has you covered, there’s a specific “kind” of job for this, as you guessed from the title, the CronJob.

An example app.

Let’s assume you created yourself a job – I’m going to create a simple job that just outputs the date/time at the scheduled time. I’ve written this in Rust but to be honest it’s simple enough that this could be any language. Here’s the Cargo.toml

The application is just a standard console application named crj (for cronjob or cron rust job, I really didn’t think about it :)).

[package]
name = "crj"
version = "0.1.0"
edition = "2024"

[dependencies]
chrono = "0.4"

Here’s the code

use chrono::Local;

fn main() {
    let now = Local::now();
    println!("Current date and time: {}", now);
}

See I told you it was simple.

Docker

For completeness, here’s the Dockerfile and the steps to get things built, tagged and pushed

FROM rust:1.89.0-slim AS builder

WORKDIR /app
COPY . .

RUN cargo build --release

FROM debian:bookworm-slim

RUN apt-get update && apt-get install -y ca-certificates && \
    rm -rf /var/lib/apt/lists/*

COPY --from=builder /app/target/release /usr/local/bin/crj

RUN chmod +x /usr/local/bin/crj

ENTRYPOINT ["/usr/local/bin/crj/crj"]

Next up we need to build the image using (remember to use the image you created as well as the correct name for your container registry)

docker build -t putridparrot/crj:1.0.0 .

then tag it using

docker tag putridparrot/crj:1.0.0 putridparrotreg/putridparrot/crj:1.0.0

Finally we’ll push it to our container registry using

docker push putridparrotreg/putridparrot/crj:1.0.0

Kubernetes CronJob

All pretty standard stuff and to be honest the next bit is simple enough. We need to create a kubernetes yaml file (or helm charts). Here’s my cronjob.yaml

apiVersion: batch/v1
kind: CronJob
metadata:
  name: scheduled-job
  namespace: dev
spec:
  schedule: "*/5 * * * *" # every 5 minutes
  jobTemplate:
    spec:
      template:
        spec:
          containers:
            - name: scheduled-job
              image:  putridparrotreg/putridparrot/crj:1.0.0
          restartPolicy: Never

My cronjob has the name scheduled-job (I know, not very imaginative). We apply this file to Kubernetes as usual i.e.

kubectl apply -f .\cronjob.yaml

Did it work?

We’ll ofcourse want to take a look at what happened after this CronJob was set up in Kubernetes. We can simply use the following. You can set the namespace used, such as dev in my case.

kubectl get cronjobs --all-namespaces -w

you’ll see something like this

NAMESPACE   NAME            SCHEDULE      TIMEZONE   SUSPEND   ACTIVE   LAST SCHEDULE   AGE
dev         scheduled-job   */5 * * * *   <none>     False     0        <none>          9s
dev         scheduled-job   */5 * * * *   <none>     False     1        0s              16s
dev         scheduled-job   */5 * * * *   <none>     False     0        13s             29s
dev         scheduled-job   */5 * * * *   <none>     False     1        0s              5m16s

In my case the job starts (ACTIVE) and then completes and shuts down. Then 5 minutes later it starts again as expected with this cron schedule.

On the pods side you can run 

kubectl get pods -n dev -w

Now what you’ll see is something like this

NAME                           READY   STATUS              RESTARTS   AGE
scheduled-job-29257380-5w4rg   0/1     Completed           0          51s
scheduled-job-29257385-qgml2   0/1     Pending             0          0s
scheduled-job-29257385-qgml2   0/1     Pending             0          0s
scheduled-job-29257385-qgml2   0/1     ContainerCreating   0          0s
scheduled-job-29257385-qgml2   1/1     Running             0          2s
scheduled-job-29257385-qgml2   0/1     Completed           0          3s
scheduled-job-29257385-qgml2   0/1     Completed           0          5s
scheduled-job-29257385-qgml2   0/1     Completed           0          5s
scheduled-job-29257390-2x98r   0/1     Pending             0          0s
scheduled-job-29257390-2x98r   0/1     Pending             0          0s
scheduled-job-29257390-2x98r   0/1     ContainerCreating   0          0s
scheduled-job-29257390-2x98r   1/1     Running             0          2s

Notice that the pod is created and goes into a “Pending” state. Then “ContainerCreating” before “Running” and finally “Completed”, but the next run of the cronjob creates a new pod name. Therefore, if you’re trying to log the pods i.e. kubectl logs scheduled-job-29257380-5w4rg -n dev – then you’ll get something like the below, but you cannot -f (follow) the logs as the next time the job runs it creates a new pod.

Current date and time: 2025-08-17 15:00:09.294317303 +00:00