Really glad the instructions on the fleet docs site are just wrong..

helpful

I have been seeing a similar issue in Rancher 2.6. if you force an update in the GUI it will get the correct state. It seems like there is some kind of bug in fleet whereby sometimes the Rancher server is not correctly updated with fresh status from downstream cluster.

What are the firewall requirements for fleet agent? I get this error on one of my isolated downstream clusters Internal error occurred: failed calling webhook "webhook.cert-manager.io": Post "https://cert-manager-webhook.cert-system.svc:443/mutate?timeout=10s": dial tcp <ip-address>443 connect: connection refused

I am sorry if this is a dumb question, and I have spent a long time searching to find an answer. But..... I have a bunch of helm charts in a GitHub repo. They have not been compressed or processed, just the raw folders and yaml files. I have now, in the fleet repo created a folder and placed a fleet.yaml in it.

svc1-
         chart.yaml
         values.yaml
         templates
                service.yaml
svc2 - 
         chart.yaml
         values.yaml
         templates
                service.yaml

I have set the helm.chart to the chart name, and the repo to the repo that contains all the charts. All repos are private. Fleet can read the fleet repo, but then when it comes to reading the helm charts from the helm chart repo I get either a 404 error, or an index.html not found error. Previously, I have been using Flex, and it had no problems with the helm repo structure. Below is a sample of my fleet.yaml, perhaps someone can tell me what I am doing wrong.

namespace: manager
helm:
  chart: svc1
  repo: <https://github.com/xxx/helm_repos>
targetCustomizations:

- name: office
  helm:
    values:
      deployment:
      tag: 1.1.16

👋 Hello, team!

hi i was wondering what is the branching strategy for fleet. If I have x number of applications, should we have one repo for all the applications or should we be having x repo's corresponding to each applications. Any suggestions for fleet is welcome. Also, when we define gitrepo objects in rancher again should we put as one gitrepo or nultiple gitrepo objects

Hi. We are facing some issues where fleet occasionally causes massive cpu spikes. Would LimitRange in all the namespaces used by fleet be a viable way to mitigate this problem?

Is there a way to tell fleet to not delete resources when the bundle gets deleted?

Hey, is it possible to patch a resource that is being installed from an external helm chart? e.g I have a 3rd party helm chart installed via helm, which generates a Service resource which I'd like to update once it's created.

How do you force fleet to update a deployment when image tag is always the same, and updated in the registry

Is it possible to move a already deployed bundle from on path to another without having to uninstall it? So basically I have a git repo with ``helm/charts/a-chart` and I'd like to move it to another repository and have just

a-chart

. However, I don't want it to be uninstalled and reinstalled afterwards as the only change here is where it is coming from.

would helm annotations differ

huh, I added helm chart to fleet.yaml, but it doesn't deploy, but marks as successful

Ah found the blunder 🙂

How does Rancher Fleet deal with bundles of the type 'Job'; since by definition they are almost entirely immutable, in ArgoCD I would set a TTL on the Job so that it can simply be run again after it is deleted by K8s itself; in Rancher Fleet it just ends up in a state 'missing' without doing anything about it (which seems entirely contrary to the purpose of GitOps through Fleet)

you mean manifest Job @narrow-ghost-32021

Job is a one time work afaik

@ambitious-plastic-3551 Yeah a manifest for

Kind: Job

(I'm Jean-Philippe's colleague). Exactly because Job's are one-time, they're almost entirely immutable.

hi , I am trying to use fleet for the first time for helm charts. I have this in my logs.

time="2022-09-20T12:26:46Z" level=error msg="error syncing 'fleet-default/teleport-cluster': handler bundle: execution error at (teleport-cluster/templates/deployment.yaml:45:28): clusterName is required in chart values, requeuing"

Further info: helm: # Use a custom location for the Helm chart. This can refer to any go-getter URL or # OCI registry based helm chart URL e.g. "oci://ghcr.io/fleetrepoci/guestbook". # This allows one to download charts from most any location. Also know that # go-getter URL supports adding a digest to validate the download. If repo # is set below this field is the name of the chart to lookup chart: teleport-cluster # A https URL to a Helm repo to download the chart from. It's typically easier # to just use

chart

field and refer to a tgz file. If repo is used the # value of

chart

will be used as the chart name to lookup in the Helm repository. repo: https://charts.releases.teleport.dev # A custom release name to deploy the chart as. If not specified a release name # will be generated. releaseName: teleport-cluster # The version of the chart or semver constraint of the chart to find. If a constraint # is specified it is evaluated each time git changes. # The version also determines which chart to download from OCI registries. version: 10.2.2 # Any values that should be placed in the

values.yaml

and passed to helm during # install. valuesFiles: - values.yaml # Override immutable resources. This could be dangerous. force: false # Set the Helm --atomic flag when upgrading atomic: false clusterName: prm Any help appreciated. Tia

I should have mentioned this first. I get 0/0 clusters available

Hello, I managed 200+ EKS clusters with the same app deployed on all of them. I need to customize a field in the chart values. Is there a way to refer to cluster label in the values? for instance

- hosts [<http://clustername.google.com|clustername.google.com>,<http://clustername.subdomain.google.com|clustername.subdomain.google.com>]

how about using labels inside manifest

apiVersion: batch/v1
kind: Job

spec:
  template:
    spec:
      containers:
      - name: gitops-jobs
        image: public.ecr.aws/.....
        command: 
          - /bin/sh
          - -c
        args: 
          - | 

            aws s3 mb <s3://dh>-<project_name-data>-cur;
            aws s3 mb <s3://dh>-<project_name-data>-us;
            aws s3 mb <s3://dh>-<project_name-data>-sh;  
      restartPolicy: OnFailure

where <project_name_data> is a label on each cluster

You can use cluster labels as variables in fleet. Have a look at https://fleet.rancher.io/gitrepo-structure/ and look for global.fleet.clusterlables and how to use it, for example, in your helm values.yaml file. I think it should fit your use case.

Thanks Juan, Do they work in K8S manifests? how would I use it to replace <project_name-data> in that k8s manifest?

Any chance I can change a GitRepo's path, or even its repository without uninstalling it? Basically I'd like to: 1. Move

/path/to/current

to

/new/path

in the same

GitRepo

2. Move from

GitRepo A

to

GitRepo B

Hello guys, can you help me with fleet-agent. when i install 0.3.11 version it give me this error.

level=error msg="Failed to register agent: looking up secret cattle-fleet-system/fleet-agent-bootstrap: unable to load root certificates: unable to parse bytes as PEM block"

whent install 0.3.9 it gives error about certificate unknown authority. I have used this documentation to install http://fleet.rancher.io/agent-initiated/

Hi, I am trying to create a namespace, as a first test using fleet in production. But I am having issues even with something as small as that

time="2022-09-27T13:28:23Z" level=error msg="error syncing 'cluster-fleet-default-rezuma-main-b67ebb76fdb7/admin-jobs-daskhub-gitops-main-rezumamainjobs-templates': handler bundle-deploy: failed to create resource: namespaces \"rezuma-admin-jobs\" not found, requeuing"

any idea how to start troubleshooting that?