I am trying to copy a file in Kubernetes to an existing directory. I am getting the error
caused: mount through procfd: not a directory: unknown
I looked at similar issues, but cannot seem to resolve it. These are the confs
Configmap
apiVersion: v1
kind: ConfigMap
metadata:
name: hdfs-yarn
data:
yarn-site.xml: |
<?xml version="1.0" encoding="UTF-8"?>
...
This is my deployment yaml
volumeMounts:
- mountPath: /opt/druid/conf/druid/cluster/_common/yarn-site.xml
name: hdfs-volume
subPath: yarn-site.xml
- mountPath: /druid/data
name: data-volume
...
volumes:
- name: hdfs-volume
configMap:
name: hdfs-yarn
items:
- key: yarn-site.xml
path: yarn-site.xml
...
Can anyone point out what could be wrong here? Thanks in advance for all the help.
The Kubernetes documentation states:
Volumes can not mount onto other volumes or have hard links to other
volumes
Instead of using one config map for volume hdf—--try using two config maps for each of the volems and let me know if this works.
volumeMounts:
- mountPath: /opt/druid/conf/druid/cluster/_common/yarn-site.xml
name: hdfs-volume
subPath: yarn-site.xml
- mountPath: /druid/data
name: data-volume
...
volumes:
- name: hdfs-volume
configMap:
name: hdfs-yarn
- name: data-volume
configMap:
name:data-yarn
Refer this Volumes and Configmap doc for more information
Related
yaml file like below
apiVersion: v1
kind: Pod
metadata:
name: fortune-configmap-volume
spec:
containers:
- image: luksa/fortune:env
env:
- name: INTERVAL
valueFrom:
configMapKeyRef:
name: fortune-config
key: sleep-interval
name: html-generator
volumeMounts:
- name: html
mountPath: /var/htdocs
- image: nginx:alpine
name: web-server
volumeMounts:
- name: html
mountPath: /usr/share/nginx/html
readOnly: true
- name: config
mountPath: /etc/nginx/conf.d
volumes:
- name: html
emptyDir: {}
- name: config
configMap:
name: fortune-config
defaultMode: 0777
As you can see,I set the configMap volume defaultMode 0777, when I try to modify file in container path /etc/nginx/conf.d, it told me operation is not allowed, but why?
https://github.com/kubernetes/kubernetes/issues/62099
Secret, configMap, downwardAPI and projected volumes will be mounted as read-only volumes. Applications that attempt to write to these volumes will receive read-only filesystem errors. Previously, applications were allowed to make changes to these volumes, but those changes were reverted at an arbitrary interval by the system. Applications should be re-configured to write derived files to another location.
This is my pod:
spec:
containers:
volumeMounts:
- mountPath: /etc/configs/config.tmpl
name: config-main
readOnly: true
subPath: config.tmpl
volumes:
- configMap:
defaultMode: 420
name: config-main
name: config-main
This is my config map:
apiVersion: v1
kind: ConfigMap
metadata:
name: config-main
data:
config.tmpl: |
# here goes my config content
This produces the following error:
Error: failed to create containerd task: failed to create shim: OCI
runtime create failed: container_linux.go:380: starting container
process caused: process_linux.go:545: container init caused:
rootfs_linux.go:76: mounting
"/var/lib/kubelet/pods/dc66ebd1-90ef-4c25-bb69-4b3329f61a5a/volume-subpaths/config-main/podname/1"
to rootfs at "/etc/configs/config.tmpl" caused: mount through procfd:
not a directory: unknown
The container has pre-existing /etc/configs/config.tmpl file that I want to override
To mount as file instead of directory try:
spec:
containers:
volumeMounts:
- mountPath: /etc/configs/config.tmpl
name: config-main
readOnly: true
subPath: config.tmpl
volumes:
- configMap:
defaultMode: 420
items:
- key: config.tmpl
path: config.tmpl
name: config-main
name: config-main
I've managed to make it work:
mountPath must be a directory
using subPath didn't work for me, anyway official doc says "using a ConfigMap as a subPath volume mount will not receive ConfigMap updates", which isn't an option for me
so I guess you can't mount a single file, you always mount a directory but then you can optionally limit which files from the configmap's data you want there via items in a volume.
spec:
containers:
volumeMounts:
- mountPath: /etc/configs
name: config-main
readOnly: true
volumes:
- configMap:
defaultMode: 420
name: config-main
name: config-main
I have created ConfigMap from Openshift 4 Windows 10 CLI:
.\oc create configmap my-cacerts --from-file=cacerts
I can see ConfigMap with name my-cacerts and download binary file cacerts from it using web interface of Openshift 4
Now I mount it (part of my-deployment.yaml)
containers:
volumeMounts:
- name: my-cacerts-volume
mountPath: /etc/my/cacerts
volumes:
- name: my-cacerts-volume
config-map:
name: my-cacerts
Unfortunately /etc/my/cacerts is mounted as a empty folder but not as a single binary file.
How can I mount cacerts as a file and not as a directory?
Update:
If I issue
.\oc get configmap my-cacerts
There is following output:
apiVersion: v1
binaryData:
cacerts: ... big long base64...
kind: ConfigMap
metadata: ...
If I issue
.\oc describe pod my-pod
I get
Volumes:
my-cacerts-volume:
Type: EmptyDir (a temporary directory that shares a pod's lifetime)
Your volumes definition is incorrect, config-map does not exist and is invalid, but it seems the API is silently falling back to an EmptyDir here, thus leading to an empty directory.
When you create a ConfigMap using the oc command above, the result will be a ConfigMap that looks like this (note that there is one key called "cacerts"):
apiVersion: v1
kind: ConfigMap
metadata:
name: my-cacerts
data:
cacerts: |
Hello world!
In the volumes section, then use configMap: together with subPath as follows to mount a only a single key ("cacerts") from your ConfigMap:
$ oc edit deployment my-deployment
[..]
spec:
containers:
- image: registry.fedoraproject.org/fedora-minimal:33
name: fedora-minimal
volumeMounts:
- mountPath: /etc/my/cacerts
name: my-cacerts-volume
subPath: cacerts
[..]
volumes:
- configMap:
name: my-cacerts
defaultMode: 420
name: my-cacerts-volume
This then results in:
$ oc rsh ...
sh-5.0$ ls -l /etc/my/cacerts
-rw-r--r--. 1 root 1000590000 13 Dec 3 19:11 /etc/my/cacerts
sh-5.0$ cat /etc/my/cacerts
Hello world!
You can also leave subPath out and set /etc/my/ as the destination for the same result, as for each key there will be a file:
[..]
volumeMounts:
- mountPath: /etc/my/
name: my-cacerts-volume
[..]
volumes:
- configMap:
name: my-cacerts
name: my-cacerts-volume
For the right syntax, you can also check the documentation
For Openshift 4 defaultMode should be specified:
volumeMounts:
- mountPath: /etc/my
name: cacerts-ref
readOnly: true
volumes:
- name: cacerts-ref
configMap:
defaultMode: 420
name: cacerts
After that configMap contents are mapped correctly.
.\oc describe pod my-pod
Volumes:
cacerts-ref:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: cacerts
Optional: false
I've multiple secrets created from different files. I'd like to store all of them in common directory /var/secrets/. Unfortunately, I'm unable to do that because kubernetes throws 'Invalid value: "/var/secret": must be unique error during pod validation step. Below is an example of my pod definition.
apiVersion: v1
kind: Pod
metadata:
labels:
run: alpine-secret
name: alpine-secret
spec:
containers:
- command:
- sleep
- "3600"
image: alpine
name: alpine-secret
volumeMounts:
- name: xfile
mountPath: "/var/secrets/"
readOnly: true
- name: yfile
mountPath: "/var/secrets/"
readOnly: true
volumes:
- name: xfile
secret:
secretName: my-secret-one
- name: yfile
secret:
secretName: my-secret-two
How can I store files from multiple secrets in the same directory?
Projected Volume
You can use a projected volume to have two secrets in the same directory
Example
apiVersion: v1
kind: Pod
metadata:
labels:
run: alpine-secret
name: alpine-secret
spec:
containers:
- command:
- sleep
- "3600"
image: alpine
name: alpine-secret
volumeMounts:
- name: xyfiles
mountPath: "/var/secrets/"
readOnly: true
volumes:
- name: xyfiles
projected:
sources:
- secret:
name: my-secret-one
- secret:
name: my-secret-two
(EDIT: Never mind - I just noticed #Jonas gave the same answer earlier. +1 from me)
Starting with Kubernetes v1.11+ it is possible with projected volumes:
A projected volume maps several existing volume sources into the same
directory.
Currently, the following types of volume sources can be projected:
secret
downwardAPI
configMap
serviceAccountToken
This is an example for "... how to use a projected Volume to mount several existing volume sources into the same directory".
May be subPath (using subPath) will help.
Example:
volumeMounts:
- name: app-redis-vol
mountPath: /app/config/redis.yaml
subPath: redis.yaml
- name: app-config-vol
mountPath: /app/config/app.yaml
subPath: app.yaml
volumes:
- name: app-redis-vol
configMap:
name: config-map-redis
items:
- key: yourKey
path: redis.yaml
- name: app-config-vol
configMap:
name: config-map-app
items:
- key: yourKey
path: app.yaml
Here your configMap named config-map-redis created from file redis.yaml mounted in app/config/ as file redis.yaml.
Also configMap config-map-app mounted in app/config/ as app.yaml
There is nice article about this here: Injecting multiple Kubernetes volumes to the same directory
Edited:
#Jonas answer is correct!
However, if you use volumes as I did in the question then, short answer is you cannot do that, You have to specify mountPath to an unused directory - volumes have to be unique and cannot be mounted to common directory.
Solution:
What I did at the end was, instead keeping files in separate secrets, I created one secret with multiple files.
I'm having hard time configuring mountPath as a relative path.
Let's say I'm running the deployment from /user/app folder and I want to create secret file under /user/app/secret/secret-volume as follows:
apiVersion: v1
kind: Pod
metadata:
name: secret-test-pod
spec:
containers:
- name: test-container
image: nginx
volumeMounts:
# name must match the volume name below
- name: secret-volume
mountPath: secret/secret-volume
# The secret data is exposed to Containers in the Pod through a Volume.
volumes:
- name: secret-volume
secret:
secretName: test-secret
For some reason the file secret-volume is created in the root directory /secret/secret-volume.
It's because you have mountPath: secret/secret-volume change it to mountPath: /user/app/secret/secret-volume
Check documentation here.