I have created a custom Splunk forwarder image.
Image name: vrathore/splunkuniversalforwarder
I have verified that the log is pushing to the server. I am using dummy log present in my host (c/Users/var/log). If I run this Docker command:
docker run --name splunkforwarder -d -v /c/Users/var/log://var/log/messages -p 8089:8089 -p 8088:8088 -e SPLUNK_SERVER_HOST=splunk-prodtest-gsp.test.com:9997 -e
FORWARD_HOSTNAME=kubernetes vrathore/splunkuniversalforwarder
Now I wanted to use the same image in Kubernetes pod, where 2 container will share their log folder with my Splunk forwarder image.
spec:
revisionHistoryLimit: 5
strategy:
type: RollingUpdate
rollingUpdate:
maxUnavailable: 10%
maxSurge: 10%
replicas: 1
template:
metadata:
name: %APP_FULL_NAME%-pod
labels:
appname: %APP_FULL_NAME%
stage: %APP_ENV%
component: app-kube-pod-object
spec:
containers:
- name: %APP_FULL_NAME%-service
image: %DOCKER_IMAGE%
imagePullPolicy: Always
envFrom:
- configMapRef:
name: %APP_CONFIG_MAP%
command: ["catalina.sh", "run"]
ports:
- containerPort: 8080
imagePullSecrets:
- name: %DOCKER_REPO_REGKEY%
selector:
matchLabels:
appname: %APP_FULL_NAME%
stage: %APP_ENV%Kubernetes is new to me. How can I share the log folder between the containers?
You need to define an emptyDir type volume and attach it to both containers. Assuming that the logs from the app are under /var/log/myapp/ (I have added the second container as well)
spec:
revisionHistoryLimit: 5
strategy:
type: RollingUpdate
rollingUpdate:
maxUnavailable: 10%
maxSurge: 10%
replicas: 1
template:
metadata:
name: %APP_FULL_NAME%-pod
labels:
appname: %APP_FULL_NAME%
stage: %APP_ENV%
component: app-kube-pod-object
spec:
containers:
- name: %APP_FULL_NAME%-service
image: %DOCKER_IMAGE%
imagePullPolicy: Always
envFrom:
- configMapRef:
name: %APP_CONFIG_MAP%
command: ["catalina.sh", "run"]
ports:
- containerPort: 8080
volumeMounts:
- name: logs
mountPath: /var/log/myapp/
- name: uf
image: vrathore/splunkuniversalforwarder
...
volumeMounts:
- name: logs
mountPath: /var/log/myapp/
imagePullSecrets:
- name: %DOCKER_REPO_REGKEY%
volumes:
- name: logs
emptyDir: {}
selector:
matchLabels:
appname: %APP_FULL_NAME%
stage: %APP_ENV%Also, I would recommend looking for an alternative solution, with Collectord and Monitoring Kubernetes/OpenShift you can tell Collectord where to look for logs and you don't need to run a sidecar container https://www.outcoldsolutions.com/docs/monitoring-kubernetes/v5/annotations/#application-logs, just one Collectord daemon will do the work.