I am trying to use helm repos on Artifactory (6.4) and have created both virtual and local repo and then uploaded (using Web UI) some example charts. Then:
Adding the virtual repo:
$ helm repo add helm-test-virtual http://x.x.x.x/artifactory/helm-test-virtual --username user1
Password:
"helm-test-virtual" has been added to your repositories
What's new in the repo:
$ helm repo update
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "helm-test-virtual" chart repository
...Successfully got an update from the "stable" chart repository
Update Complete. ⎈ Happy Helming!⎈
I already have some example helm charts uploaded and reindexed:
$ helm search helm-test-virtual/
No results found
Can anyone hint me why the helm search helm-test-virtual/ says No results found ?
Did you do a helm repo update after you've uploaded the files?
helm client has a local cache and resolves everything from it, including the search.
Related
I've installed airflow on my k8s using helm chart apache-airflow/airflow.
Currently installed airflow version is 2.2.4, what path should be followed from the options below to upgrade airflow version 2.3.0.
The official helm chart version is 1.5.0 and the default airflow version is 2.2.4. When the newer version of helm chart is released, the default airflow version will be set to 2.3.0. Will the helm repo update and helm upgrade command provision the upgrade of airflow? or are there any other similar process or official upgrade process guide?
If the upgrade process to accommodate the airflow default version has to be manual, what process/steps should be followed? N.B. changing defaultAirflowTag value from 2.2.3->2.3.0 in values.yaml is not an option as it causes an exception.
Thanks in advance.
there. I checked my airflow environment on my aws ec2.
It didn't work properly.
So I did lots of methods. It proved all failed.
But finally I found solution.
if you should run below these two commands, you should remove your airflow namespace first.
There is some errors in airflow namespace.
So I found that it worked properly after deleting airflow namespace.
1)kubectl delete namespace airflow
2)helm repo remove apache-airflow https://airflow.apache.org
helm repo add apache-airflow https://airflow.apache.org
3)You should see your brand-new airflow interface.
How I solved :
helm upgrade --install actually works. but I saw a message on airflow web saying some of data migration failed cause of schema changes in airflow 2.3.0. After I drop the mentioned table in postgresql, airflow worked fine.
and changing values.yaml also works.
images:
airflow:
repository: apache-airflow/airflow
tag: latest
pullPolicy: IfNotPresent
I succesfully pushed a git-lfs repo to a Nexus repo and I can see two git blobs on the remote server. I'm wondering if is there the possibility to clone this repo. I tried different ways:
git clone http://..../reponame/info/lfs
git clone http://..../reponame/
creating a new git-lfs repo and pulling (no error, but no pull is actually done)
There's no reference to this topic in the doc here https://help.sonatype.com/repomanager3/nexus-repository-administration/formats/git-lfs-repositories
What's the right way to clone/pull this type of repository from Nexus?
We have Micro services application which has different helm charts for each MS, and there is umbrella chart which adds all these charts and create an integrated chart for simple deployment purpose, we use Nexus3 as a helm repo.
In umbrella chart requirements.yaml dependencies are added with version, repository.
When we specify exact chart version in requirements.yaml, sometimes it fails to find that chart in nexus repo, though it actually exists. After some investigation found that it looks for the version in index.yaml of that nexus repo, if required version is not same as like the one in index.yaml file it fails.
index.yaml contains latest version always, but what if we want to work with different version than latest one ? how can we fix this ?
Below is the snippet of requirements.yaml file
dependencies:
- name: ms1
version: "1.3.0"
repository: http://user:passwd#nexus_host:8081/repository/helm_chart_repo/ms1
- name: ms2
version: "1.3.0"
repository: http://user:passwd#nexus_host:8081/repository/helm_chart_repo/ms2
- name: ms3
version: "1.2.0"
repository: http://user:passwd#nexus_host:8081/repository/helm_chart_repo/ms3
- name: ms4
version: "1.3.1"
repository: http://user:passwd#nexus_host:8081/repository/helm_chart_repo/ms4
If the required version of the Helm Chart exists in the repository, but does not exist in index.yaml, then it means that your index.yaml is not correct. It should contain all the versions that are in the repository.
Check if helm repo index <directory> is always performed after new version is added.
I have created a Helm repository in Jfrog Artifactory using 'Quick Setup'.
Then used the Jfrog helm binary to add the virtual 'helm' repo:
I am able to add custom charts to the virtual repo, install them but not able to install charts from the helm-remote (i.e stable) repo:
When I try to access any such chart, helm throws as error:
On the Artifactory web ui, the helm virtual repo lists all the remote charts, but I'm unable to install them:
I have done "helm repo update" multiple times. And also have deleted ~/.helm folder and "init" again. Still, am plagued by the same problem.
I see the following error in the Artifactory System Logs repeated several times:
2018-02-28 09:18:34,266 [art-exec-2] [ERROR] (o.a.a.h.r.m.HelmVirtualMerger:194) -
Couldn't read index file in remote repository helm-remote :
(was com.github.zafarkhaja.semver.UnexpectedCharacterException)
(through reference chain: org.jfrog.repomd.helm.model.HelmIndexYamlMetadata["entries"]->java.util.LinkedHashMap["grafana"]->java.util.TreeSet[3])
What can be the possible issues here? Thank you for your help.
tl;dr -
This issue was fixed in Artifactory 5.9.0 -
https://www.jfrog.com/jira/browse/RTFACT-15668
The problem is that Artifactory didn't support Helm charts with non-SemVer 2 in version 5.8, and only after version 5.8 came out a non SemVer2 was added to the stable chart repo. (grafana 0.6)
This issue was fixed in 5.9.0, and you should be able to resolve remote chart correctly.
check out Artifactory release notes -
https://www.jfrog.com/confluence/display/RTF/Release+Notes#ReleaseNotes-Artifactory5.9
I git cloned this repository: docker-symfony and followed the installation instructions.
When I visited symfony.dev:81, I saw kibana 4.
My problem is, I don't understand where I should put the Symfony project.
My OS is Ubuntu 14.04
The instructions mention:
put your Symfony application into symfony folder and do not forget to add symfony.dev in your /etc/hosts file.
Then, run:
$ docker-compose up
You are done, you can visite your Symfony application on the following URL: http://symfony.dev (and access Kibana on http://symfony.dev:81)
That "symfony folder" comes from docker-compose.yml#L4
volumes:
- ./symfony:/var/www/symfony
It means you should create a symfony folder in your docker-symfony cloned repo, which will be mounted as /var/www/symfony in the application container.
docker-symfony
code
elk/logstash
nginx
php-fpm
.gitignore
.travis.yml
README.md
symfony <====== Add your folder there
That folder can be a symbolic link to wherever you have your symfony project.