Hi Venkat, first i want to tell you that i found and solved the root cause which makes prometheus not to work. But i want to thank you for this video of using Rancher, is very awesome and is working like a charm :). I didn't knew about it. Man you make a real good job with this video series. Thank you for this tutorials. Best Regards.
Hmm. I have seen this issue from many people. Not sure about a proper resolution. In the below link the guy talks about dns name/ip address being wrong when the machine came back up. See if that helps. I haven't tested that myself. Thanks.
Excellent video and better your explanation. I have the following doubts: 1.- What is the function of the following components etcd, control and worker. 2.- In the cluster installation I only have one Master, how could I configure 2 or more masters in the same cluster? Thanks and regards!!
Hi Cesar, thanks for watching. etcd is the data store where the cluster state is stored. Kubernetes master node runs certain components like controller manager, scheduler, api server. All interactions to your cluster are through the api server. Worker nodes run kubelet service which receives the command from the master and schedules the work load on it. For multi-master cluster, I have done few videos. You can check them here. Kubespray: ua-cam.com/play/PL34sAs7_26wOAqYsrIhtDaIviGlSkmfv9.html KinD: ua-cam.com/video/4p4DqdTDqkk/v-deo.html Cheers.
Hi venket, Thanks for your video. I have setup rancher successfully, but it is not working after I reboot the server or restart the docker service. What could be the issue?
great video. I observe normal user( non-root user) get root access in imported Kubernetes cluster version 1.15 in rancher. How a normal user in rancher force to use there u -id and g -id (security context) to avoid root access during pod creation or deployment?
Hi Deepak, thanks for watching. It all depends on the kubeconfig file you use to connect to your cluster. In all my videos, I used cluster admin's kubeconfig file.
Hi Venket When starting the cluster (from your vagrantfile) the scheduler & the controller-manager have the status "unhealthy" .. i've tried to downgrade the server version to an older one (--kubernetes-version=v1.18.5) but i have the same issue (scheduler and controller-manager unhealth ) did you ever cross this problem ?
@@guymasumbuko6119 I have been using my Vagrant provisioned Kubernetes cluster for a long time. I will check this again. Is your host machine Linux, Windows or Mac?
Ah I see what you mean. I too get the same thing. I will explore further. Let me also check if this is happening on other clusters as well. pastebin.com/tRtqtM6F
Hi, I have checked this on my other kubernetes cluster running using Ranchers K3D and I don't see that issue. But it seems to be a common problem. github.com/Azure/AKS/issues/173 This is not affecting the cluster operations in any ways. Component status api is being deprecated it seems.
Just me and Opensource My host is a windows 10 . All the pods are running fine except the scheduler and controller . I'm going to try to install the cluster manually using kubeadm and 3 oracle virtual boxes (without vagrant) to see if it resolve the issue . In any cases thanks very for taking time to reply to me
Great videos, as a matter of fact I can fairly say one of the best for k8s on the net. I have a question regarding this video, for rancher to work i believe it has to have kubernetes provisioned with enable-hostpath-provisioner flag passed to the kube controller manager when bootstraping the cluster. The question is How do you do that when initializng the cluster via kubeadm init?
Hi Ivan, thanks for watching this video and taking time to appreciate. I wasn't aware of that flag to the kube controller. All my videos are based on kubeadm method of cluster installation. And Rancher just works fine. I didn't have to set that flag during kubeadm initialization. Thanks.
Nice video on Cluster Management tool. I notice if I am mounting a local folder /opt/rancher as /var/lib/rancher inside the docker as explained in your video, container is repeatedly restarting. Removing it makes it run consistently but I am unable to save my configuration. Any suggestions? Restating with this statement docker run -d --restart=unless-stopped --name my-rancher -h rancher -p 80:80 -p 443:443 -v /opt/rancher:/var/lib/rancher rancher/rancher:latest Working fine docker run -d --restart=unless-stopped --name my-rancher -h rancher -p 80:80 -p 443:443 rancher/rancher:latest
Hi Water Kingdom, Thanks for watching this video. I think I know your problem. I believe you are running "docker run" command as non-root user without sudo. Thats how I did in this video. But you might not have permission to write files in /opt directory as it is under root filesystem and only root user can write to it. There are various solutions. 1) You can do "sudo docker run" instead of "docker run" command and you won't have permissions problem. or 2) mkdir /opt/rancher and chown it to your user account and then run "docker run" command or 3) instead of /opt/rancher, just use "rancherdata" and a local docker volume will be created. rancherdata:/var/lib/rancher. Try the 3rd option above and let me know if it worked. Thanks.
@@justmeandopensource I really appreciate your prompt response. Sorry I missed out this point that I am running everything as "root" user. When I did it last week it worked without any problems. When I am trying for a demo, it's not working.
Hi.. your video simple and complicated and make me comfortable. thanks you so much for your video`s I hope make a video tutorial for "clustering ubuntu using raspberry with Rancher and Kubernetes"
Hi! Great video! One thing that I would like to understand is how to load balance those workers. I understand that I can create an ingress from rancher, correct? What would be the correct way of doing so?
Hi Max, thanks for watching this video. You can use ingress for load balancing. I have done couple of videos on ingress controllers using Nginx and Traefik. Please search my channel. Might be useful to you. I haven't used rancher for deploying the ingress although you can do it from Rancher. Thanks, Venkat
Hi Gokul, thanks for watching. Can you give me more details about your setup? 1. How is your cluster provisioned? 2. Where are you running your cluster? 3. How are your running Rancher? 4. Which pod crashes? I have tried this video recently and it worked for me. Cheers.
@@justmeandopensource Thank u for reply 1.using kubeadm i have installed in RHEl 7.7 using oracle virtual box 2.for rancher i am running seperate vm and installed docker in same os node agent pods are running fine but cluster agent pod is not runing
@@justmeandopensource sorry i have removed rancher, because i want show dashboard with metrics, if possible please share helm 3 installation steps and prometheus and metrics collector and also grafana steps, i have tried to install tiller helm install command i got error unknown flag for service account
Hi Anant, thanks for watching this video. I was looking at the below documentation, where there was a mention about nginx ingress for Layer 7 load balancing solution. rancher.com/docs/rancher/v2.x/en/k8s-in-rancher/load-balancers-and-ingress/load-balancers/ I have done couple of videos around ingress controllers. Please check these out. Nginx ingress ua-cam.com/video/chwofyGr80c/v-deo.html Traefik Ingress ua-cam.com/video/A_PjjCM1eLA/v-deo.html Thanks
Hi, thanks for watching. I haven't tried importing a cluster in AWS to Rancher. But the process should be the same as long as you have access to the cluster from the machine where you are running Rancher through kubeconfig. Its worth checking the logs of Rancher pods in the cattle-system namespace.
Thank you for the great content. I would like your help a dummy question. I sow your video, but I got confused about where you run rancher command. Why you run runcher command in your local machine and not in your master master? Is it possible to run runcher directly on master machine? I only have one master and two nodes. Sorry if my question is dummy. I also find difficulties on making it available to other machines in the network.
Hi, thanks for watching. Basically you need to run Rancher container somewhere. You can run it anywhere. Why do you think you need to run it inside the master node? It can be run anywhere. You only need to make sure that it can talk to the kubernetes cluster in order to manage it. In my case I ran it on my Linux host.
Hello Sir Do we need to know K8s in detail before working on Rancher or what part of k8s I should know ? Can you please make a list of prerequisite things one should know before starting to work on Rancher (e.g. Docker etc).
Hi Raghunandan, thanks for watching this video. Rancher is only a tool to manage your kubernetes cluster(s). You will need to know all the Kubernetes concepts in order to efficiently deploy and manage the resources. Rancher is a different way of interacting with your Kubernetes cluster. Thats it. Instead of typing "kubectl" commands by hand, you can do that in a GUI. So probably you don't need to learn writing a manifest. But concept wise, you need to know the fundamentals. You can go through the videos in this series and understand various concepts. Thanks.
Hi Mahesh, thanks for watching. I have been trying Rancher installation every now and then and never encountered cert error. Where or at which point exactly you are getting that error? Can you try that on a new k8s cluster?
Hi Venkat, Hope you are doing good. I have installed docker on host machine and i can connect k8s cluster from the same host. used the command to run rancher latest image using docker command, when i check logs it has the following output docker logs -f c23e6740bced ERROR: Rancher must be ran with the --privileged flag when running outside of Kubernetes -- i tried to append --privileged=true while running docker command as well, still the same, did not help google too. Kindly clarifiy, thanks
Hi Khan, Thanks for watching. Try the below command instead. Or try running the docker run command with sudo. docker run -d --privileged --restart=unless-stopped -p 80:80 -p 443:443 rancher/rancher
@@justmeandopensource Hi Venkat, yes i did this. But unfortunately docker container is not running on the host when i run netstat -ntlp , upon checking logs it gave strange errors standard_init_linux.go:178: exec user process caused "exec format error" .. Sorry i tried to fix this myself did not come up with solution, When you get chance please help. thanks
Hi Hussain, thanks for watching. You can use Rancher's RKE to provision HA kubernetes cluster if you like. rancher.com/docs/rancher/v2.x/en/installation/ha/kubernetes-rke/ Thanks.
Hello Venkat, I am facing an error below: Error while creating domain: Error saving the server: Call to virDomainDefineXML failed: invalid argument: could not get preferred machine for /usr/bin/qemu-system-x86_64 type=kvm ➜ vagrant-provisioning git:(master) kvm-ok INFO: Your CPU does not support KVM extensions KVM acceleration can NOT be used Is there any other setup kubernetes? FYI i am using Hetzner cloud.
Hi Venkat.. thanks for video. I tried to install the same but when I tried to import my cluster in to Rancher, it is always in the pending state and when I check the pod logs, it is throwing " Connecting to wss://IP_ADDRESS/v3/connect with token 5szh4dkfwflwxs8bz854d6ntcp9dsgq7g542cz9kqvb4rwb8l6ksdq Connecting to proxy" url="wss://IP_ADDRESS/v3/connect Failed to connect to proxy" error="x509: certificate has expired or is not yet valid Failed to connect to proxy" error="x509: certificate has expired or is not yet valid" " Any help in fixing this is greatly appreciated.
Hi Matheen, thanks for watching this video. What version of Kubernetes cluster you are using? If you are using latest v1.16.0, it won't work. I tried couple days ago. The api versions have changed for few resource types and Rancher haven't updated their manifests. But if you are using a k8s cluster version less than 1.16.0, it should work. When you select the import cluster in Rancher, it will show a page with commands that you need to run in your cluster. Make sure to use the last command. Thanks.
Hello - did something happen with the kmaster. I run the command - scp root@kmaster.example.com:/etc/kubernetes/admin.conf ~/.kube/config I get an error - ssh: Could not resolve hostname kmaster.example.com: Name or service not known
Hi Mani, Rancher is the project/tool that you use to manage multiple Kubernetes clusters. And RKE is Rancher's Kubernetes Engine. Similar to GKE in Google and EKS in AWS, RKE is the Kubernetes offering from Rancher. Thanks.
I installed minikube on ec2 instance. I'm unable to access rancher UI through my public IP after running this command. sudo docker run -d --restart=unless-stopped -p 80:80 -p 443:443 -v /opt/rancher:/var/lib/rancher rancher/rancher I'm able to access this via `curl localhost` but not through public IP. Networking rules sudo netstat -antup | grep 80 tcp6 0 0 :::80 :::* LISTEN 149359/docker-proxy What other changes are required?
Hi Is it possible to deploy rancher for management of k8s cluster which CRI is containerd. Can you update this video with latest version as you know docker is going to deprecated on up-coming version. How we can deploy wit containerd. Thanks.
Hy Venkat, thanks for you tutorial. I want to ask you some case, first we has succesfully create cluster with rancher provisioning vsphere. so how if rancher down? how we can access kubernetes cluster? i find some kubeconfig, but this kubeconfig only can access if rancher power up.
Hi Olivia, thanks for watching. Its not required to have Rancher up in order to access your k8s cluster. K8s cluster is somethine separate and Rancher is a container that you are running separately. You used kubectl commands to deploy certain resources in your cluster in order to start managing the cluster using Rancher. SO you still have access to your cluster using kubectl and kube config file.
Just me and Opensource thanks venkat, but i’m creating my cluster with rancher provisioning with vsphere. so i only have rancher kubeconfig, can’t find cluster kubeconfig.
@@oliviatm2949 just ssh into the master node where the config will be there already. you can run kubectl there. or you can copy the config from .kube/ to your local machine.
Hi Venkat, I have k8s cluster with one master & one node, I have set u in VM environment with CentOS-7, K8s version v1.20.4-0 I followed the same steps, but in my case, it created only one pod. also the pod was not ready for long time, then describe the pod and it is saying that Readiness probe failed: Get "192.168.235.132:8080/health": dial tcp 192.168.235.132:8080: connect: connection refused
Hi Avi, thanks for watching. I used Zsh with oh-my-zsh plugin manager. And for command suggestions I used zsh-autosuggestions plugin. The terminal application was Gnome-terminal but the set up will work on any terminal application. I did a video on my terminal setup a while ago and here it is if you are interested. ua-cam.com/video/soAwUq2cQHQ/v-deo.html
Hi Susheel, as shown in this video, you can run a rancher docker container and import any existing Kubernetes cluster. You can import any number of clusters into Rancher. I don't understand what you mean by "accessing it via HAProxy" Thanks
Just tried and the cluster remained in pending state in Rancher dashboard. This is with Kubernetes cluster with containerd. Can you give me more details about your environment?
Hello I have quick question. I have installed rancher in my local mac computer. I have created GKE cluster using Rancher. Only problem when I click the lunch kubectl it is giving error code 1006. I couldnt figure it out the problem. Help is appreciated
Hi, thanks for watching this video. I haven't used Rancher with GKE, so can't comment on that. Searching the internet, I came across the below link where fee people complained the same behaviour. Could be proxy thing butnot sure.
Hi Venkat Hope you are doing good I am getting time out error while running rancher command on AWS cli and i can't add cloud cred on rancher (auth error) could you please help me in this
Hi Syed, thanks for watching. I will add it to my list. I have got videos scheduled for the next two Mondays (21st and 28th). I will record RKE HA video soon and will release it on 5th October. Cheers.
Hi Prashant, thanks for watching this video. Its exactly as described in the video. Create cluster in Rancher which will give you a command to run in your kubernetes cluster. You run the command using kubectl on any machine that has access to your cluster. Thanks
Hi Milind, I have recorded the video for Prometheus already and scheduled to be published later. There are few more videos waiting to be published and then this one will be published. Thanks.
Hi Milind, You might have noticed [ Kube 24 ] Helm video released this morning. I am releasing videos on every Monday. So next Monday Feb 25th, will be a video on Jenkins CI/CD in Kubernetes. And the following Monday Mar 4th, will be the "Kubernetes Monitoring with Prometheus and Grafana" video. Thanks for waiting patiently. Venkat
Hi Venkat thak you for all your kubernetes video, i would like to know it is a possible to do a live migration of ressource between differents kubernetes clusters in rancher ? thanks a
Hi Cheik, thanks for watching this video. The way to migrate resources between clusters is to backup what you want from one cluster and restore what you want into another cluster. You can use Velero to do that. I have done a video on Velero in Kubernetes and waiting to be released. Thanks.
@@cheikamedmaoulida7367 It only backs up the resource definition and state of etcd cluster and any persistent volumes. All the applications will continue to run on the original cluster. You will have to restore the resources into the new cluster and then switch off the old one. There will have to be downtime and I don't know of any solution to do live migration. Thanks.
@@cheikamedmaoulida7367 I did but forgot to mention it to you. Apologies. Here it is. ua-cam.com/video/C9hzrexaIDA/v-deo.html Covered the basic concepts. But if you want your persistent volumes as well to be backed up, it gets little complex that you need to be using a supported storage provider. Cheers.
Hi, could you make a video around multi cluster deployment. I tried with kind clusters and faced x509 certificate issues during fleet deployments for healthy clusters. You could also explore more multi cluster features
hey , i have a query here. what if i using KOPS and i had a cluster and storage will be on s3 then how can i deploy it with rancher. Can you suggest me something.
@@justmeandopensource I have deployed my cluster through KOPS and then tried to install rancher server Inside same cluster but can't coz always some error with deployment file, till now no document is there regarding this in a legit way , so I tired to make a deployment file and gave Rancher latest image to it but I think it doesn't support this way.... May be just have to run a seprate container like you showed and then add this cluster to it.... Coz they say that it requires 4 gb ram and what if I deployed it as a daemon set so every node of cluster shd have minimum 4 gb of ram then.... Or shd i give a dedicated server for rancher. Dnt know what will be the best practice for this.
Hi. Great video. If just installed a k8s on prem and run Rancher. But the pod corresponding to cattle-cluster-agent is restarting over and over again, I notice that that pod can't connect to my Rancher. But when I ping from the k8s cluster It pings normally.
Hi Christian, thanks for watching this video. One of the other viewer commented that if you used Calico overlay network in k8s, you will experience this issue. But I haven't tried it myself. I will be testing it later this week. He also provided a solution in his comment. Thanks.
Hi Prateek, Thanks for your interest in Kubernetes techonology. For setting up High Availability in Kubernetes cluster you can take a look at Official documentation at kubernetes.io/docs/setup/independent/high-availability. Basically you will deploy multiple master nodes behind a load balancer. Or you can have etcd component on a separate group of HA cluster outside of master nodes. Either ways, you can follow the approach in the above documentation. I didn't get your first question right. What you mean by sizing nodes in line cluster? Thanks
@@justmeandopensource and as far as I read in official documents of kubernetes if we allocate a load balancer in front of 3 kube master .. there is still possibility of single point of failure . So how to overcome from that
You are right. Load balancer is again a single point of failure. DigitalOcean has a good article about this. Please check out www.digitalocean.com/community/tutorials/an-introduction-to-haproxy-and-load-balancing-concepts. Regarding system spec, for k8s master nodes, the minimum recommended RAM is 2G at least from K8s recent versions. I used to deploy master node on a Virtual Machine with 1G RAM. But I think from v1.13.0, it has to be 2G otherwise it will fail during cluster initialization. The environment I have is for learning and so entirely running on Virtualization platform using VirtualBox. If in production environment, you will have to identify the workload requirements and design the spec accordingly. Number of CPU cores for inidividual nodes, RAM and disk usage all depend on your workload. So there is no single spec sheet. Thanks
Hi, thanks for watching. Its again the same process. You have kubectl access to your eks cluster. Just run the command given by Rancher and your cluster will be imported.
Thanks for the awesome video. In my case i am trying to create one external etcd for two cluster created using rancher. It will be very helpful if you can give any suggestion.
Hi Sufia, thanks for watching. This is something I haven't tried yet. And also I wouldn't advice having a single etcd cluster for multiple kubernetes cluster. If its for learning purpose, then absolutely fine. But I have no experience on that unfortunately.
Getting this error when i am integrating my AKS cluster "Containerservice.ManagedClustersClient#CreateOrUpdate: Failure sending request: StatusCode=0 -- Original Error: autorest/azure: Service returned an error. Status=400 Code="AgentPoolK8sVersionNotSupported" Message="Version 1.11.5 is not supported in this region. Please use [az aks get-versions] command to get the supported version list in this region. For more information, please check aka.ms/supported-version-list" , My kubernetes and noodpool version is "1.16.13" Please advice me on this.
Sometimes it throw error and cattle-cluster-agent just restarts again and again "ERROR: 192.168.0.106/ping is not accessible (Failed to connect to 192.168.0.106 port 443: Connection timed out)"
I installed rancher from vagrant and i want to copy admin.conf from master Irun the command # scp root@172.16.16.100:/etc/kubernetes/admin.conf .kube/config It ask for root password What it the password to enter?
a tad late but for those who struggled on this: scp root@172.16.16.100:/etc/kubernetes/admin.conf ~/.kube/config password should be "kubeadmin" as per bootstrap.sh steps: github.com/justmeandopensource/kubernetes/blob/4f9a78868f3d27b7473a0d08aea6c00ae219c861/vagrant-provisioning/bootstrap.sh#L74
Can u change cni plugin live, like uninstall flannel switch to calico or weave. And do calico BGP external to physical network like Cisco csr is god save my arse. Thx, dude.
HI, I haven't tried that. Basically when you initialize your cluster using kubeadm, you pass the pod-cidr-network option to kubeadm init command which varies between the cni plugins you use. And if you want to change the cni plugin, you might need to re-initialize your cluster or tweak the cni plugin for the cidr network. Even if can do this, won't it affect the application you are running in your cluster when pods depends on other pods/services and there will be a disruption while you update the cni plugin. I am not sure about the practice in a production environment, but you can try it on a development environment. Thanks.
Thx for reply. As I. Know flannel don’t have ipam plugin Kubeadm initial CIDR is basically handled by kuber hostport plugin. unlike flannel, calico cilium contiv 3party Cni they have ipam func inside. So I read calico doc, they say, u can change ip pool setting, in other words, they bypass hostport ipam, they don’t care And uninstall flannel switch to others, east west traffic will stop until newer cni plugin up. North south traffic will not going down,if u are using kube Service to expose port, third party ingress controller But in real production, some customer want expose pod network, as I can find. There has ipvlan macvlan cni Just google containernetwork cni github, u will find it. Basically Pod network just looks like BM server. No kube proxy no kube ingress.
And I find some procedures described uninstall flannel blog.51cto.com/liuzhengwei521/2382257?source=dra This is Chinese website. 1.Delete flannel pod 2.Delete nodes flannel interface(vxlan encap logical interface, and funny thing is this type of vxlan dst UDP port is not 4789, is 8472, Linux vxlan driver default port, 4789 is basically Cisco juniper players staff..) 3.rm -rf /var/lib/cni/ 4,rm -f /etc/cni/net.d/* 5.Reload kubelet 6.play new CNI Again thx for doing this series.
Hi Nathan, thanks for watching. That is not a big deal. We can use Lets Encrypt to get an SSL certificate and use it instead of Rancher's self signed certificate. May be for another video.
I can pull from github or some repo of gcp or azure ui for managing cluster or edge.. n customize may b ankitzer 😁😁😄😃😀 well thanks for information.. b ware of replicated.com and charts upload
Hi venket, Thanks for your video. I have setup rancher successfully, but it is not working after I reboot the server or restart the docker service. What could be the issue?
Hi Jino, thanks for watching. How did you run your Rancher container? Did you follow the same docker run command that I used in this video? I used volume binding to bind /opt/rancher to /var/lib/rancher inside the container. So even after a system restart, docker container rancher will have the data. I think you didn't use volume binding. Cheers.
![[ Kube 18 Discussion 1 ] Rancher issues with Calico](http://i.ytimg.com/vi/4_SzQ6sKJwo/mqdefault.jpg)
![[ Kube 18 Discussion 1 ] Rancher issues with Calico](/img/tr.png)
likewise, Venkat always shares an amazing video and I can't believe this was 2 years ago 😄
I still need to learn more 😊
🙂Thanks for your continued interest in my content. It helps me indeed. Cheers.
You showed all cool features of Rancher. Thank you Venkat
Hi Hareesh, thanks for watching this video I'm taking time to comment. you'll find a lot more features as you keep using the tool. Thanks
Thanks for sharing such a wonderful tutorial, I had gone through this video and brought my Rancher connected to my cluster up and running.👍
Good to hear that. Thanks for watching.
Thanks Venket !!. Nice quick summary how to set up a Rancher to manage K8 Cluster(s). I loved it. Kudos to you !!
Hu Mathew, thanks for watching. Cheers.
Happy that I got chance to see this video.. very interesting and understanding for beginners venkat 🙏🙏
Hi Rajkumar, Thanks for watching.
Hi Venkat,
first i want to tell you that i found and solved the root cause which makes prometheus not to work.
But i want to thank you for this video of using Rancher, is very awesome and is working like a charm :).
I didn't knew about it.
Man you make a real good job with this video series.
Thank you for this tutorials.
Best Regards.
Hi Knight, glad to hear that you resolved the issue. Cheers
Thank you very much! All your videos are extremely helpful and very well made.
Hi Richard, Thanks for watching this video and taking time to comment/appreciate. Cheers.
Thank you so much Venket. This is an excellent one to start with.
Hi Namal, thanks for watching. Cheers.
Really good explanation! I was sucked in the video! THANKS!
Hi Francesco, thanks for watching. Cheers.
Very good Explanation. Thankyou!
Hi Devaraj, Thanks for watching.
Thanks !! always used the kubectl command line. I am new to Rancher. nice video to start with. Thanks again!!. Mat.
Thanks Mathew for watching this video.
thanks for video, you should record new video for rancher v2.5+ version, they has got a lot of new features.
Hi, thanks for watching. I will definitely do one with the latest version. Cheers.
When turning off master node and turning it back on, the cluster is stuck in not ready state. How to solve this ?
Hmm. I have seen this issue from many people. Not sure about a proper resolution. In the below link the guy talks about dns name/ip address being wrong when the machine came back up. See if that helps. I haven't tested that myself. Thanks.
Excellent video and better your explanation. I have the following doubts: 1.- What is the function of the following components etcd, control and worker. 2.- In the cluster installation I only have one Master, how could I configure 2 or more masters in the same cluster? Thanks and regards!!
Hi Cesar, thanks for watching.
etcd is the data store where the cluster state is stored. Kubernetes master node runs certain components like controller manager, scheduler, api server. All interactions to your cluster are through the api server. Worker nodes run kubelet service which receives the command from the master and schedules the work load on it.
For multi-master cluster, I have done few videos. You can check them here.
Kubespray: ua-cam.com/play/PL34sAs7_26wOAqYsrIhtDaIviGlSkmfv9.html
KinD: ua-cam.com/video/4p4DqdTDqkk/v-deo.html
Cheers.
Great video. Thanks for sharing :)
Hi Blake, thanks for watching,
Very good introduction to Rancher
Hi Prasanna, thanks for watching.
Hi venket,
Thanks for your video. I have setup rancher successfully, but it is not working after I reboot the server or restart the docker service. What could be the issue?
Already replied in your other comment. Cheers.
thank you for this video, I really appreciate it!
Hi Hazarky, thanks for watching.
great video. I observe normal user( non-root user) get root access in imported Kubernetes cluster version 1.15 in rancher.
How a normal user in rancher force to use there u -id and g -id (security context) to avoid root access during pod creation or deployment?
Hi Deepak, thanks for watching. It all depends on the kubeconfig file you use to connect to your cluster. In all my videos, I used cluster admin's kubeconfig file.
Hi Venket
When starting the cluster (from your vagrantfile) the scheduler & the controller-manager have the status "unhealthy" .. i've tried to downgrade the server version to an older one (--kubernetes-version=v1.18.5) but i have the same issue (scheduler and controller-manager unhealth ) did you ever cross this problem ?
By the way thank you very for all your kubernetes vids , they are awesome
@@guymasumbuko6119 I have been using my Vagrant provisioned Kubernetes cluster for a long time. I will check this again.
Is your host machine Linux, Windows or Mac?
Ah I see what you mean. I too get the same thing. I will explore further. Let me also check if this is happening on other clusters as well.
pastebin.com/tRtqtM6F
Hi, I have checked this on my other kubernetes cluster running using Ranchers K3D and I don't see that issue.
But it seems to be a common problem.
github.com/Azure/AKS/issues/173
This is not affecting the cluster operations in any ways. Component status api is being deprecated it seems.
Just me and Opensource My host is a windows 10 . All the pods are running fine except the scheduler and controller . I'm going to try to install the cluster manually using kubeadm and 3 oracle virtual boxes (without vagrant) to see if it resolve the issue . In any cases thanks very for taking time to reply to me
Great videos, as a matter of fact I can fairly say one of the best for k8s on the net. I have a question regarding this video, for rancher to work i believe it has to have kubernetes provisioned with enable-hostpath-provisioner flag passed to the kube controller manager when bootstraping the cluster. The question is How do you do that when initializng the cluster via kubeadm init?
Hi Ivan, thanks for watching this video and taking time to appreciate. I wasn't aware of that flag to the kube controller. All my videos are based on kubeadm method of cluster installation. And Rancher just works fine.
I didn't have to set that flag during kubeadm initialization.
Thanks.
Excellent video start to rancher. Thanks
Hi Sudeesh, thanks for watching.
Nice video on Cluster Management tool. I notice if I am mounting a local folder /opt/rancher as /var/lib/rancher inside the docker as explained in your video, container is repeatedly restarting. Removing it makes it run consistently but I am unable to save my configuration. Any suggestions?
Restating with this statement
docker run -d --restart=unless-stopped --name my-rancher -h rancher -p 80:80 -p 443:443 -v /opt/rancher:/var/lib/rancher rancher/rancher:latest
Working fine
docker run -d --restart=unless-stopped --name my-rancher -h rancher -p 80:80 -p 443:443 rancher/rancher:latest
Hi Water Kingdom, Thanks for watching this video. I think I know your problem.
I believe you are running "docker run" command as non-root user without sudo. Thats how I did in this video. But you might not have permission to write files in /opt directory as it is under root filesystem and only root user can write to it.
There are various solutions.
1) You can do "sudo docker run" instead of "docker run" command and you won't have permissions problem.
or
2) mkdir /opt/rancher and chown it to your user account and then run "docker run" command
or
3) instead of /opt/rancher, just use "rancherdata" and a local docker volume will be created.
rancherdata:/var/lib/rancher.
Try the 3rd option above and let me know if it worked.
Thanks.
@@justmeandopensource I really appreciate your prompt response. Sorry I missed out this point that I am running everything as "root" user. When I did it last week it worked without any problems. When I am trying for a demo, it's not working.
I see. You can just delete /opt/rancher which was created last week when you ran it as root. Thanks.
Hi..
your video simple and complicated and make me comfortable. thanks you so much for your video`s
I hope make a video tutorial for "clustering ubuntu using raspberry with Rancher and Kubernetes"
Hi Torry, thanks for watching.
@@justmeandopensource
hi sir i hope you can create the video tutorial about ubuntu clustering rancher kubernetes using raspberry. 😍
@@torrymonata Hi I wish to do that but I don't have sufficient Raspberry Pis. The couple I am having are for different purposes.
Hi! Great video!
One thing that I would like to understand is how to load balance those workers.
I understand that I can create an ingress from rancher, correct?
What would be the correct way of doing so?
Hi Max, thanks for watching this video. You can use ingress for load balancing. I have done couple of videos on ingress controllers using Nginx and Traefik. Please search my channel. Might be useful to you. I haven't used rancher for deploying the ingress although you can do it from Rancher.
Thanks,
Venkat
hi rancher cluster agent is not working properly, getting crashloopbackoff
Hi Gokul, thanks for watching. Can you give me more details about your setup?
1. How is your cluster provisioned?
2. Where are you running your cluster?
3. How are your running Rancher?
4. Which pod crashes?
I have tried this video recently and it worked for me.
Cheers.
@@justmeandopensource Thank u for reply
1.using kubeadm i have installed in RHEl 7.7 using oracle virtual box
2.for rancher i am running seperate vm and installed docker in same os
node agent pods are running fine but cluster agent pod is not runing
@@gokulakrishna.g I think Rancher deploys all the resources in cattle-system namespace. Can you paste the output of "kubectl -n cattle-system get all"
@@justmeandopensource sorry i have removed rancher, because i want show dashboard with metrics, if possible please share helm 3 installation steps and prometheus and metrics collector and also grafana steps, i have tried to install tiller helm install command i got error unknown flag for service account
@@justmeandopensource i was deployed rancher in sepearate vm may be that will leads to issue can u please suggest
Please can you tell me about the Rancher Loadbalancers. i need to deploy accordingly....Thanks in advance
Hi Anant, thanks for watching this video.
I was looking at the below documentation, where there was a mention about nginx ingress for Layer 7 load balancing solution.
rancher.com/docs/rancher/v2.x/en/k8s-in-rancher/load-balancers-and-ingress/load-balancers/
I have done couple of videos around ingress controllers. Please check these out.
Nginx ingress
ua-cam.com/video/chwofyGr80c/v-deo.html
Traefik Ingress
ua-cam.com/video/A_PjjCM1eLA/v-deo.html
Thanks
In rancher Monitering Api is not ready error displying in rancher dashboard on AWS ec2 env cluster
Hi, thanks for watching. I haven't tried importing a cluster in AWS to Rancher. But the process should be the same as long as you have access to the cluster from the machine where you are running Rancher through kubeconfig. Its worth checking the logs of Rancher pods in the cattle-system namespace.
Thank you for the great content.
I would like your help a dummy question. I sow your video, but I got confused about where you run rancher command. Why you run runcher command in your local machine and not in your master master? Is it possible to run runcher directly on master machine? I only have one master and two nodes. Sorry if my question is dummy.
I also find difficulties on making it available to other machines in the network.
Hi, thanks for watching. Basically you need to run Rancher container somewhere. You can run it anywhere. Why do you think you need to run it inside the master node? It can be run anywhere. You only need to make sure that it can talk to the kubernetes cluster in order to manage it. In my case I ran it on my Linux host.
Hi Venkat Thanks a lot very useful
No worries. Thanks for watching.
U aaah brilliant.
Bare metals series are brilliant.
Thx,,dude.
Many thanks for watching this video and taking time to comment
nice, any clues for automate the number of containers in pods
Hi Alfonso, sorry I didn't get it. What do you exactly mean?
Hello Sir
Do we need to know K8s in detail before working on Rancher or what part of k8s I should know ?
Can you please make a list of prerequisite things one should know before starting to work on Rancher (e.g. Docker etc).
Hi Raghunandan, thanks for watching this video.
Rancher is only a tool to manage your kubernetes cluster(s). You will need to know all the Kubernetes concepts in order to efficiently deploy and manage the resources. Rancher is a different way of interacting with your Kubernetes cluster. Thats it. Instead of typing "kubectl" commands by hand, you can do that in a GUI. So probably you don't need to learn writing a manifest. But concept wise, you need to know the fundamentals. You can go through the videos in this series and understand various concepts.
Thanks.
Venkat Sir, Thank for your quick response.
@@raghunandanmundada645 You are welcome. Cheers.
I am getting Cert Expired error ? what to do to resolve ?
Hi Mahesh, thanks for watching. I have been trying Rancher installation every now and then and never encountered cert error. Where or at which point exactly you are getting that error? Can you try that on a new k8s cluster?
Great one on Rancher! Thanks.
Thanks for watching.
Thanks so much for this tutorial.
Thanks for watching.
Hi Venkat, Hope you are doing good. I have installed docker on host machine and i can connect k8s cluster from the same host. used the command to run rancher latest image using docker command, when i check logs it has the following output docker logs -f c23e6740bced
ERROR: Rancher must be ran with the --privileged flag when running outside of Kubernetes -- i tried to append --privileged=true while running docker command as well, still the same, did not help google too. Kindly clarifiy, thanks
Hi Khan, Thanks for watching. Try the below command instead. Or try running the docker run command with sudo.
docker run -d --privileged --restart=unless-stopped -p 80:80 -p 443:443 rancher/rancher
@@justmeandopensource Hi Venkat, yes i did this. But unfortunately docker container is not running on the host when i run netstat -ntlp , upon checking logs it gave strange errors standard_init_linux.go:178: exec user process caused "exec format error" .. Sorry i tried to fix this myself did not come up with solution, When you get chance please help. thanks
Great video. Thank you
Hi Ancy, Thanks for watching.
awesome video
Hi Zahurul, thanks for watching.
Thankyou for the video sir.. Can you tell us which screen recorder are you using ??
Hi Himanshu, thanks for watching. I use simplescreenrecorder.
Hi...How can i set up HA k8s cluster using rancher?
Hi Hussain, thanks for watching. You can use Rancher's RKE to provision HA kubernetes cluster if you like.
rancher.com/docs/rancher/v2.x/en/installation/ha/kubernetes-rke/
Thanks.
Hello Venkat, I am facing an error below:
Error while creating domain: Error saving the server: Call to virDomainDefineXML failed: invalid argument: could not get preferred machine for /usr/bin/qemu-system-x86_64 type=kvm
➜ vagrant-provisioning git:(master) kvm-ok
INFO: Your CPU does not support KVM extensions
KVM acceleration can NOT be used
Is there any other setup kubernetes? FYI i am using Hetzner cloud.
Hi Venkat.. thanks for video. I tried to install the same but when I tried to import my cluster in to Rancher, it is always in the pending state and when I check the pod logs, it is throwing "
Connecting to wss://IP_ADDRESS/v3/connect with token 5szh4dkfwflwxs8bz854d6ntcp9dsgq7g542cz9kqvb4rwb8l6ksdq
Connecting to proxy" url="wss://IP_ADDRESS/v3/connect
Failed to connect to proxy" error="x509: certificate has expired or is not yet valid
Failed to connect to proxy" error="x509: certificate has expired or is not yet valid"
"
Any help in fixing this is greatly appreciated.
Hi Matheen, thanks for watching this video. What version of Kubernetes cluster you are using? If you are using latest v1.16.0, it won't work. I tried couple days ago. The api versions have changed for few resource types and Rancher haven't updated their manifests.
But if you are using a k8s cluster version less than 1.16.0, it should work.
When you select the import cluster in Rancher, it will show a page with commands that you need to run in your cluster. Make sure to use the last command.
Thanks.
Hello - did something happen with the kmaster. I run the command -
scp root@kmaster.example.com:/etc/kubernetes/admin.conf ~/.kube/config
I get an error - ssh: Could not resolve hostname kmaster.example.com: Name or service not known
Hu, thanks for watching. Use ip address of kmaster or add an entry to your /etc/hosts file.
what is the difference between rancher and rke??
Hi Mani, Rancher is the project/tool that you use to manage multiple Kubernetes clusters. And RKE is Rancher's Kubernetes Engine. Similar to GKE in Google and EKS in AWS, RKE is the Kubernetes offering from Rancher.
Thanks.
@@justmeandopensource thanks for making me understand so simple
@@manikanthkommoju3176 You are most welcome. Cheers.
I installed minikube on ec2 instance. I'm unable to access rancher UI through my public IP after running this command.
sudo docker run -d --restart=unless-stopped -p 80:80 -p 443:443 -v /opt/rancher:/var/lib/rancher rancher/rancher
I'm able to access this via `curl localhost` but not through public IP.
Networking rules
sudo netstat -antup | grep 80
tcp6 0 0 :::80 :::* LISTEN 149359/docker-proxy
What other changes are required?
Hi Venkat, great video Thanks a lot.
Hi Jagan, thanks for watching.
Hi
Is it possible to deploy rancher for management of k8s cluster which CRI is containerd. Can you update this video with latest version as you know docker is going to deprecated on up-coming version. How we can deploy wit containerd.
Thanks.
can we install rancher using helm inside a cluster ?
Hi Krunal, yes you can.
rancher.com/docs/rancher/v2.x/en/installation/install-rancher-on-k8s/
Hy Venkat, thanks for you tutorial.
I want to ask you some case, first we has succesfully create cluster with rancher provisioning vsphere. so how if rancher down? how we can access kubernetes cluster? i find some kubeconfig, but this kubeconfig only can access if rancher power up.
Hi Olivia, thanks for watching. Its not required to have Rancher up in order to access your k8s cluster. K8s cluster is somethine separate and Rancher is a container that you are running separately. You used kubectl commands to deploy certain resources in your cluster in order to start managing the cluster using Rancher. SO you still have access to your cluster using kubectl and kube config file.
Just me and Opensource thanks venkat, but i’m creating my cluster with rancher provisioning with vsphere. so i only have rancher kubeconfig, can’t find cluster kubeconfig.
@@oliviatm2949 just ssh into the master node where the config will be there already. you can run kubectl there. or you can copy the config from .kube/ to your local machine.
great explanation!! Could you please make one video about.
How to setup Docker to manage your Kubernetes Cluster?
Thanks in Advance..
Hi Suman, thanks for watching. What do you mean by setting up docker to manage k8s cluster? I don't get you.
Hi Venkat, I have k8s cluster with one master & one node, I have set u in VM environment with CentOS-7, K8s version v1.20.4-0
I followed the same steps, but in my case, it created only one pod. also the pod was not ready for long time, then describe the pod and it is saying that
Readiness probe failed: Get "192.168.235.132:8080/health": dial tcp 192.168.235.132:8080: connect: connection refused
is rancher opensource and free of cost ?
Very helpful
Glad to hear that. Thanks for watching.
Does Rancher runs with Podman. Or docker is mandatory?
Excellent guide. A slightly different question - which terminal do you use there?
Hi Avi, thanks for watching. I used Zsh with oh-my-zsh plugin manager. And for command suggestions I used zsh-autosuggestions plugin. The terminal application was Gnome-terminal but the set up will work on any terminal application. I did a video on my terminal setup a while ago and here it is if you are interested.
ua-cam.com/video/soAwUq2cQHQ/v-deo.html
@@justmeandopensource Thanks for the quick reply. Is this possible under Mac?
@@AviDarks Yes. I use it in Mac as well. Iterm2 terminal. Install zsh, oh-my-zsh and zsh-autosuggestions. The steps should be similar.
Hi Venkat, Can I configure Rancher to import my K8S cluster and access it via HAProxy ?
Hi Susheel, as shown in this video, you can run a rancher docker container and import any existing Kubernetes cluster. You can import any number of clusters into Rancher.
I don't understand what you mean by "accessing it via HAProxy"
Thanks
Its not working for me..can u make rancher with containerd kubernetes cluster
Thanks for watching. I will test and let you know.
Cheers.
Just tried and the cluster remained in pending state in Rancher dashboard. This is with Kubernetes cluster with containerd.
Can you give me more details about your environment?
Hello I have quick question. I have installed rancher in my local mac computer. I have created GKE cluster using Rancher. Only problem when I click the lunch kubectl it is giving error code 1006. I couldnt figure it out the problem. Help is appreciated
Hi, thanks for watching this video. I haven't used Rancher with GKE, so can't comment on that. Searching the internet, I came across the below link where fee people complained the same behaviour. Could be proxy thing butnot sure.
Hi Venkat Hope you are doing good
I am getting time out error while running rancher command on AWS cli
and i can't add cloud cred on rancher (auth error)
could you please help me in this
@venkat, Im new to this environment. Could you please make a video on HA rancher in RKE cluster. thanks :)
Hi Syed, thanks for watching. I will add it to my list. I have got videos scheduled for the next two Mondays (21st and 28th). I will record RKE HA video soon and will release it on 5th October. Cheers.
worth of 33:26 min.. thanks man.
Hi Irfan, thanks for watching.
please share the video for rancher cli setup.
Hi Prabha, thanks for watching. I will give that a try before committing to make a video. Cheers.
Thank you for video.
can you tell me the configuration of workstation please, to have 3 nodes cluster running in it (8 Core and 16G RAM)?
Can I provision my cluster with kubespray and then install Rancher?
Yes, you should be able to do it. It doesn't matter how you provision your kubernetes cluster. All you need is a kubernetes cluster :)
Thank you so much venkat.
Hi Sivaguru, You are welcome and thanks for watching this video.
Hi Venkat, thanks for wonderful video on Rancher Have you ran the command to add cluster in rancher on K8 master ?
Hi Prashant, thanks for watching this video. Its exactly as described in the video. Create cluster in Rancher which will give you a command to run in your kubernetes cluster. You run the command using kubectl on any machine that has access to your cluster.
Thanks
Thank you so much
Thanks for watching.
Hello venkat,
Thanks for the videos can you please create the video onhow to use prometheus in kubernetes.
Hi Milind, I have recorded the video for Prometheus already and scheduled to be published later. There are few more videos waiting to be published and then this one will be published. Thanks.
Thanks venkat hope you will publish it soon.
Hi Milind, You might have noticed [ Kube 24 ] Helm video released this morning. I am releasing videos on every Monday. So next Monday Feb 25th, will be a video on Jenkins CI/CD in Kubernetes. And the following Monday Mar 4th, will be the "Kubernetes Monitoring with Prometheus and Grafana" video.
Thanks for waiting patiently.
Venkat
Hi Venkat thak you for all your kubernetes video,
i would like to know it is a possible to do a live migration of ressource between differents kubernetes clusters in rancher ?
thanks a
Hi Cheik, thanks for watching this video. The way to migrate resources between clusters is to backup what you want from one cluster and restore what you want into another cluster. You can use Velero to do that. I have done a video on Velero in Kubernetes and waiting to be released. Thanks.
@@justmeandopensource
Ok and if we use that backup services will continuse the execution during the transfert infomation?
@@cheikamedmaoulida7367 It only backs up the resource definition and state of etcd cluster and any persistent volumes. All the applications will continue to run on the original cluster. You will have to restore the resources into the new cluster and then switch off the old one. There will have to be downtime and I don't know of any solution to do live migration.
Thanks.
@@justmeandopensource
hi Ventkat, have you ever publi video of velero? i have seen in your git a file of that?
@@cheikamedmaoulida7367 I did but forgot to mention it to you. Apologies.
Here it is.
ua-cam.com/video/C9hzrexaIDA/v-deo.html
Covered the basic concepts. But if you want your persistent volumes as well to be backed up, it gets little complex that you need to be using a supported storage provider.
Cheers.
Hi, could you make a video around multi cluster deployment. I tried with kind clusters and faced x509 certificate issues during fleet deployments for healthy clusters. You could also explore more multi cluster features
hey , i have a query here. what if i using KOPS and i had a cluster and storage will be on s3 then how can i deploy it with rancher. Can you suggest me something.
What have you tried so far in this regard?
@@justmeandopensource I have deployed my cluster through KOPS and then tried to install rancher server Inside same cluster but can't coz always some error with deployment file, till now no document is there regarding this in a legit way , so I tired to make a deployment file and gave Rancher latest image to it but I think it doesn't support this way.... May be just have to run a seprate container like you showed and then add this cluster to it.... Coz they say that it requires 4 gb ram and what if I deployed it as a daemon set so every node of cluster shd have minimum 4 gb of ram then.... Or shd i give a dedicated server for rancher. Dnt know what will be the best practice for this.
Awesome ...Thank you so much ...
Hi Fazle, thanks for watching.
Hi. Great video. If just installed a k8s on prem and run Rancher. But the pod corresponding to cattle-cluster-agent is restarting over and over again, I notice that that pod can't connect to my Rancher. But when I ping from the k8s cluster It pings normally.
Hi Christian, thanks for watching this video. One of the other viewer commented that if you used Calico overlay network in k8s, you will experience this issue. But I haven't tried it myself. I will be testing it later this week. He also provided a solution in his comment. Thanks.
nice
Hi Ragini, thanks for watching.
Hi nice videos for learning kubernetes. Can u share some resources for sizing nodes in line cluster. And how to make to kube HA
Hi Prateek,
Thanks for your interest in Kubernetes techonology.
For setting up High Availability in Kubernetes cluster you can take a look at Official documentation at kubernetes.io/docs/setup/independent/high-availability. Basically you will deploy multiple master nodes behind a load balancer. Or you can have etcd component on a separate group of HA cluster outside of master nodes. Either ways, you can follow the approach in the above documentation.
I didn't get your first question right. What you mean by sizing nodes in line cluster?
Thanks
@@justmeandopensource by sizing a node in cluster I mean how much resources( ram,cpu,disk) should I allocate ?
@@justmeandopensource and as far as I read in official documents of kubernetes if we allocate a load balancer in front of 3 kube master .. there is still possibility of single point of failure . So how to overcome from that
You are right. Load balancer is again a single point of failure. DigitalOcean has a good article about this. Please check out www.digitalocean.com/community/tutorials/an-introduction-to-haproxy-and-load-balancing-concepts.
Regarding system spec, for k8s master nodes, the minimum recommended RAM is 2G at least from K8s recent versions. I used to deploy master node on a Virtual Machine with 1G RAM. But I think from v1.13.0, it has to be 2G otherwise it will fail during cluster initialization. The environment I have is for learning and so entirely running on Virtualization platform using VirtualBox. If in production environment, you will have to identify the workload requirements and design the spec accordingly. Number of CPU cores for inidividual nodes, RAM and disk usage all depend on your workload. So there is no single spec sheet.
Thanks
great video, can you explain how to import existing aws eks cluster
Hi, thanks for watching. Its again the same process. You have kubectl access to your eks cluster. Just run the command given by Rancher and your cluster will be imported.
great
Hi Mohammad, thanks for watching.
Very great video. Keep it up!
Hi Miquel, thanks for watching this video.
Thanks for the awesome video.
In my case i am trying to create one external etcd for two cluster created using rancher. It will be very helpful if you can give any suggestion.
Hi Sufia, thanks for watching. This is something I haven't tried yet. And also I wouldn't advice having a single etcd cluster for multiple kubernetes cluster. If its for learning purpose, then absolutely fine. But I have no experience on that unfortunately.
Just me and Opensource Thanks for the reply. For learning purpose only i wanted to try.
@@sufiaalmas5354 Okay. I haven't tried using an external etcd cluster for k8s.
can you give a pointer to your bash customization ?
Hi, thanks for watching. I have done a video on my terminal customization.
ua-cam.com/video/PUWnCbr9cN8/v-deo.html
Have anyone use rancher in production? How experienced on it?
I have used it very briefly in production.
hi, i cant add kub on rancher , receive level=error msg="Remotedialer proxy error" error="websocket: bad handshake"
can you ping worker to master?
@@PePTo-dx2yj are you using LXC?
@@BYAZIT docker, use cattle, cgroup-driver=cgroupfs, may be you have some links
in Internet with good instruction about install kuber ?
@@BYAZIT yes i can
@@PePTo-dx2yj ua-cam.com/video/oILc0ywDVTk/v-deo.html
Getting this error when i am integrating my AKS cluster "Containerservice.ManagedClustersClient#CreateOrUpdate: Failure sending request: StatusCode=0 -- Original Error: autorest/azure: Service returned an error. Status=400 Code="AgentPoolK8sVersionNotSupported" Message="Version 1.11.5 is not supported in this region. Please use [az aks get-versions] command to get the supported version list in this region. For more information, please check aka.ms/supported-version-list" , My kubernetes and noodpool version is "1.16.13"
Please advice me on this.
wow
Thanks for watching.
really nice
Thanks Gordan.
Sometimes it throw error and cattle-cluster-agent just restarts again and again "ERROR: 192.168.0.106/ping is not accessible (Failed to connect to 192.168.0.106 port 443: Connection timed out)"
because centos 8 - Rancher 2x only support centos 7.
I installed rancher from vagrant and i want to copy admin.conf from master
Irun the command # scp root@172.16.16.100:/etc/kubernetes/admin.conf .kube/config
It ask for root password
What it the password to enter?
a tad late but for those who struggled on this:
scp root@172.16.16.100:/etc/kubernetes/admin.conf ~/.kube/config
password should be "kubeadmin"
as per bootstrap.sh steps: github.com/justmeandopensource/kubernetes/blob/4f9a78868f3d27b7473a0d08aea6c00ae219c861/vagrant-provisioning/bootstrap.sh#L74
Hi Mourad, thanks for responding to this question. Much appreciated. Cheers.
Can u change cni plugin live, like uninstall flannel switch to calico or weave.
And do calico BGP external to physical network like Cisco csr is god save my arse.
Thx, dude.
HI, I haven't tried that. Basically when you initialize your cluster using kubeadm, you pass the pod-cidr-network option to kubeadm init command which varies between the cni plugins you use. And if you want to change the cni plugin, you might need to re-initialize your cluster or tweak the cni plugin for the cidr network. Even if can do this, won't it affect the application you are running in your cluster when pods depends on other pods/services and there will be a disruption while you update the cni plugin.
I am not sure about the practice in a production environment, but you can try it on a development environment.
Thanks.
Thx for reply.
As I. Know flannel don’t have ipam plugin
Kubeadm initial CIDR is basically handled by kuber hostport plugin.
unlike flannel, calico cilium contiv 3party Cni they have ipam func inside.
So I read calico doc, they say, u can change ip pool setting, in other words, they bypass hostport ipam, they don’t care
And uninstall flannel switch to others, east west traffic will stop until newer cni plugin up. North south traffic will not going down,if u are using kube Service to expose port, third party ingress controller
But in real production, some customer want expose pod network, as I can find.
There has ipvlan macvlan cni
Just google containernetwork cni github, u will find it.
Basically
Pod network just looks like BM server.
No kube proxy no kube ingress.
And
I find some procedures described uninstall flannel
blog.51cto.com/liuzhengwei521/2382257?source=dra
This is Chinese website.
1.Delete flannel pod
2.Delete nodes flannel interface(vxlan encap logical interface, and funny thing is this type of vxlan dst UDP port is not 4789, is 8472, Linux vxlan driver default port, 4789 is basically Cisco juniper players staff..)
3.rm -rf /var/lib/cni/
4,rm -f /etc/cni/net.d/*
5.Reload kubelet
6.play new CNI
Again thx for doing this series.
@@johnclarkson6120 Cool. I will have a look. Thanks for sharing this.
U r welcome. U did great job, thx again.
Typical of most installations, never any mention on off SSL Certificates. Your using Rancher unsecured.
Hi Nathan, thanks for watching. That is not a big deal. We can use Lets Encrypt to get an SSL certificate and use it instead of Rancher's self signed certificate. May be for another video.
Yes but I need a domain name for that.
Yes you need one.
@@nathanstott1909 if you have an opnsense or pfsense box, it's pretty easy to setup a CA to get free local trusted certificates
Hope you can have a video running Rancher on Windows Server. Thank you.
I can pull from github or some repo of gcp or azure ui for managing cluster or edge.. n customize may b ankitzer 😁😁😄😃😀 well thanks for information.. b ware of replicated.com and charts upload
set playback speed to 1.5 is the best option
Hi, I didn't get you. Did you mean I was going too slow?
Mark
Hi venket,
Thanks for your video. I have setup rancher successfully, but it is not working after I reboot the server or restart the docker service. What could be the issue?
Hi Jino, thanks for watching. How did you run your Rancher container? Did you follow the same docker run command that I used in this video? I used volume binding to bind /opt/rancher to /var/lib/rancher inside the container. So even after a system restart, docker container rancher will have the data. I think you didn't use volume binding. Cheers.
Very helpful
Hi Abhijit, thanks for watching.