Hello everyone, We're trying to gather more information and deeper understanding how IT companies are optimizing costs and availability for their Kubernetes workloads specially in multi-cluster, multi-cloud environments. We are kindly asking you to participate in our short survey (5 minutes max). The information provided will help us tremendously and allow us to better contribute to CNCF community and projects related to Kubernetes and multi-cloud workloads. Here is a link to the survey: https://ncv.microsoft.com/Ta6MGGOGY8 Thank you for your time and shared insight! Cheers, Kubar.io team (CC: @incalculable-artist-61842)

Hey all, I have a quick question about rancher 2.5.7. I've removed a node to fix some issues with the kernel. But now I can't join it again.

I've tried to run the commands again to join but the cattle get stuck in ' Registering the node'

Do you know about this issue?

Hi, I have the following use-case; we are running an RKE2 cluster using Cilium with the Cilium BGP Control Plane enabled. So we can announce IP’s to Services in our on-premise environment, instead of having some external Load Balancer solution. We now want to install Rancher on this cluster, but as far as I see the Helm chart doesn’t have an option to configure a Service type LoadBalance and only relies on an external load balancer (which routes traffic to any of the cluster nodes running the rke2-ingress-nginx-controller). Any suggestions on how to use BGP for load balancing to the rancher service, or is using Cilium BGP Control Plane actually not a good idea?

Hi, our rancher (2.7.3) installation all of a sudden stopped creating namespaces for some users with a 403 response by rancher webhook. The user has the required RBAC Bindings on the target server. Any hint what the cause could be?

If Rancher is deployed on a single-node k3s "cluster" is there any point in having 3 replicas of the rancher pods which is the default? Or is it set to that because it's assumed rancher runs on a multi-node cluster?

if one instance dies, the extra pods may ensure that there is no actual outage while the failed pod restarts

Right.. maybe scale to 2 than.. seems like waste of cpu/memory to have 3 replicas on a single node so no HA anyway

It's a balancing game. You can also set it up to automatically create more replicas when usage is high and the go back to a minimal number when its not.

Hello ! Is it possible to bypass TLS verification when setting up a downstream cluster and when downloading the docker-machine-driver plugin?

i upgarde the kubernet version v1.25 and it show the error of longhorn unable to get engine image longhornio/longhorn-enginev1.4.2 engineimage.longhorn.io \"ei-f9e7c473\" not found" node=kgg time="2023-06-01T124931Z" level=warning msg="Error syncing Longhorn volume longhorn-system/pvc-555c0618-a058-4530-b0eb-4ad3309ae786" controller=longhorn-volume error="failed to sync longhorn-system/pvc-555c0618-a058-4530-b0eb-4ad3309ae786: unable to get engine image longhornio/longhorn-enginev1.4.2 engineimage.longhorn.io \"ei-f9e7c473\" not found" node=kgg time="2023-06-01T124931Z" level=fatal msg="Error starting manager: admission webhook \"validator.longhorn.io\" denied the request: failed to set the setting taint-toleration with invalid value : cannot modify toleration setting before all volumes are detached

can any one help me

Hello everyone! Maybe you can help me. I am migrating from Docker Desktop to Rancher Desktop. Im trying to build a container installing ssh and curl (docker build .) with the commands you see below ENV HTTP_PROXY=http://proxy... ENV HTTPS_PROXY=http://proxy... RUN apk --no-cache add ca-certificates RUN apk update && apk add curl openssh-client Everything works well in Docker Desktop, every package is dowloaded On the other hand, using Rancher I am unable to download anything. => => # fetch https://ftp.acc.umu.se/mirror/alpinelinux.org/v3.17/main/x86_64/APKINDEX.tar.gz => => # WARNING: Ignoring https://ftp.acc.umu.se/mirror/alpinelinux.org/v3.17/main: operation timed out ...... What do I need to configure for Rancher (WLS - rancher-desktop) to connect to the internet ? Thanks so much

Hello, which is the best practice to accelerate rancher 2.6.11 interface display ? I have a huge latency when showing deployments or pods of a specific namespace. Is it recommanded to enable incrementral loading and Manual refresh ? thank you in advance

Hi all, I'm using Rancher to create an EKS on AWS and I'm having the following issues that started to happen and don't know what could be the root cause

│ Error: Getting Catalog V2: Timeout getting Catalog V2 Client at cluster ID c-rc9p8: Bad response statusCode [500]. Status [500 Internal Server Error]. Body: [lost connection to cluster: failed to find Session for client stv-cluster-c-rc9p8] from [<https://rancher.cardlytics.com/k8s/clusters/c-rc9p8/v1>]
│ 
│   with module.resources.module.k8s_eks_cluster.rancher2_catalog_v2.autoscaler,
│   on .terraform/modules/resources.k8s_eks_cluster/modules/eks-rancher/cluster-autoscaler.tf line 9, in resource "rancher2_catalog_v2" "autoscaler":
│    9: resource "rancher2_catalog_v2" "autoscaler" {
│ 
╵
╷
│ Error: Getting Catalog V2: Timeout getting Catalog V2 Client at cluster ID c-rc9p8: Bad response statusCode [500]. Status [500 Internal Server Error]. Body: [lost connection to cluster: failed to find Session for client stv-cluster-c-rc9p8] from [<https://rancher.cardlytics.com/k8s/clusters/c-rc9p8/v1>]
│ 
│   with module.resources.module.k8s_eks_cluster.rancher2_catalog_v2.artifactory,
│   on .terraform/modules/resources.k8s_eks_cluster/modules/eks-rancher/catalog.tf line 2, in resource "rancher2_catalog_v2" "artifactory":
│    2: resource "rancher2_catalog_v2" "artifactory" {
│ 
╵
╷
│ Error: Getting Catalog V2: Timeout getting Catalog V2 Client at cluster ID c-rc9p8: Bad response statusCode [500]. Status [500 Internal Server Error]. Body: [lost connection to cluster: failed to find Session for client stv-cluster-c-rc9p8] from [<https://rancher.cardlytics.com/k8s/clusters/c-rc9p8/v1>]
│ 
│   with module.resources.module.k8s_eks_cluster.rancher2_catalog_v2.eks,
│   on .terraform/modules/resources.k8s_eks_cluster/modules/eks-rancher/catalog.tf line 16, in resource "rancher2_catalog_v2" "eks":
│   16: resource "rancher2_catalog_v2" "eks" {

Hi Team We are using rke 1.3.10 to create k8 clusters locally in our onprimisis. We are encountering continous slowmess in the cluster resource creation, deletion, updation. Even sometimes helm commands are exceeding its timeout and failing. Upon investigation found that the same is happening because of etcd unable to perform extensive read write operation in between the etcd plane nodes as the disk IO speed is very less in the setup around 3 to 4 mb/s. So, Can someone point me to the document explaining the minimum required bandwidth to run operations smoothly for: 1. Internet connectivity. 2. Intranet connectivity. 3. Disk IO Also am I missing something in the above list which can even impact the rke k8 cluster operation and peocesses.

Hello world! Thanks for Rancher Desktop! Unfortunately I can't pull images from the official docker hub, it fails with

Error response from daemon: Head "<https://registry-1.docker.io/v2/library/hello-world/manifests/latest>": EOF

Any ideas how to fix/troubleshoot this? I tried searching on docker, but some issues point to disable hardware acceleration on a FristBox (which I don't have), ot running docker logout.

Hello! I've tried to create a receiver in rancher for my alertmanager. In my receiver I need an oauth2 configuration. But if I add the oauth2 configuration my alertmanager won't load the receiver and there're no error messages in the logs. On my local system (without rancher) my alertmanager loads the receiver without problems. Is there something I do wrong? This is the receiver configuration which I created in rancher:

- name: 'Name'
  webhook_configs:
    - url: 'url'
      http_config:
        oauth2:
          client_id: 'id'
          endpoint_params:
            grant_type: 'string'
          client_secret: 'secret'
          scopes:
            - 'string'
          token_url: 'url'

Rancher Version: 2.5.16 | Prometheus Version: 2.27.1 | Alertmanager Version 0.22.2

Hi, is it possible in Rancher to use server groups with anti-affinity when using Openstack as machine provisioner in a Rancher managed cluster?

Trying to dip a toe into rancher-managed kubernetes and I'm struggling. Some notes: • I'm completely new to kubernetes and rancher. • I've tried to ask GPT4 and the answers it has given me don't appear to work. • Rancher itself is running inside a docker container behind traefik on a single node docker server • In the same ESXi hardware, I ahve 4 more nodes (master 1, 2, 3, and worker 1) that have been set up. Each node has an assigned static IP address from an OPNSense router which handles DHCP I went through the Rancher UI to create a cluster. The cluster has been up for 7 hours now, and currently has the error message

Non-ready bootstrap machine(s) custom-a2f94dac52de and join url to be available on bootstrap node

This is both for K3s and RKE2 Googling around suggests it's a DNS issue, but I'm struggling to figure out how the DNS lookups work and how to correct them. Can anyone offer guidance?

Hi, I setuped RKE on openstack using 3 vms but facing this error in the calico not sure what is wrong, I suspect some issue in the Allowed Ip address section in OpenStack but need some assistance if anyone can help. Thanks

I am new to kubernetes and rancher and am setting up a new cluster. As far as networking providers go is Canal the way to go as it is default?

Depends what you want, I'm currently switching to Cilium

Is there a way for me to change the Pod network CIDR /16? I am assigning the following in my cluster config and I am getting failures deploying the

rke-network-plugin

kube-api:
      service_cluster_ip_range: 10.98.0.0/16
      always_pull_images: false
      pod_security_policy: false
      secrets_encryption_config:
        enabled: true
      service_node_port_range: 30000-32767
kube-controller:
      cluster_cidr: 10.99.0.0/16
      service_cluster_ip_range: 10.98.0.0/16

image.png

There are 10 nodes in this cluster

So when I created my cluster I did not select nginx as an ingest. When I try to install traefik I get "Service does not have load balancer ingress IP address"

Do I need like a node port or something?

This is hte stage each time where I have a disconnect