Hi, thanks for watching. IPv6 is in my list and I will have to play with it for a while before I can confidently do a video. Will get to it at some point :) Cheers.
I'm really looking forward to the next parts cause I'm trying to get it running but tls refuses to work. Hope that the next parts come out soon! Love your work!
Hi Daniel, thanks for watching. I have Traefik videos scheduled everyday this week until Saturday. Tomorrow is about ingreds routes and the day after is all about tls. Cheers.
Hi sir, juste a question please, with my traefik I have a bad gateway when I try access to my rabbitmq , but when I refresh one or 2 time I get rabbitmq management interface, also I can access it by port-forward I don't understand why please if you can help thanks so mush
Hello! In the first line thank you very much for your videos! I have learned a lot but I still have a silly question: I cannot comprehend why we need MetalLB at all? I have provisioned a single-node k8s cluster with traefik 2.x as ingress and I am ablle to access my application from the Internet. As traefik works as a loadbalancer, what are the use cases of metallb? Thank yo very much for you answer in advance!
Your work is brilliant. Very well polished, consise, perfectly paced, wtih attention to detail par none. The matial is presented in a perfectly logical order, your presentation style is clean and articulate. The _vast_ majority of RUclips-ers would do very well for themselves attempting to emulate your grace and clarity; you really ought to do a demonstration on how to do demos ;) Now the heavy technical question...which terminal app do you use with what plugins? I'm enthrawled...I've never really spent time making my 'terminator' work any better than it does out of the box aside from vim awesome and a spattering of additions. I mostly code in vim or some graphical IDE and likely work harder than I need to with the support of cool tools like you have shown...so I'd love to know what you use, if you don't mind. Thank you for your inspiration and education!
@@RustyWeise Many thanks for watching. I keep switching my setup to find the best comfortable working experience with awesome tools. My terminal setup currently is something like what I have described in this video. ruclips.net/video/PUWnCbr9cN8/видео.html
Great introduction, but I think you'll want to add the setting "ExternalTrafficPolicy: Local" so that the source IP is being retained. You then might also need to scale up the replica so that the VIP can be attached to any worker node.
Hi, thanks for watching. Yeah we could make it a lot better in terms of availability and fault tolerance. I wanted to keep the flow simple as majority of my audience are beginners :)
@@justmeandopensource yeah you're totally right but I was really confused first as to why I just couldn't get my real IP forwarded to my services and explaining that in a further video would be fantastic! :) EDIT: You can easily check if X-Forwarded-For is checked with the whoami container by containous as well.
"/tmp/traefik-values.yaml" E212: Can't open file for writing Many thanks for this interesting tutorial. Unfortunately I cannot edit the yaml file as instructed, I am getting "access denied" when I am trying to save the change. There is anything I am missing or to do. Please help because I am stuck.
Great video, thank you! I'm just entering into this world of kubernetes and I have a question, why is it necessary to use MetalLB instead of the internal load balancer provided by Traefik? I think that probably the approach that you gave is more for an use case to integrate traefik with an external load balancer, so locally you're using MetalLB but in cloud we can replace it by ELB, what I say is correct?
Greetings! Could you explain how would you make this current setup work with a HAProxy external loadbalancer, like you did in the first video back in 2019 ([ Kube 32 ] Set up Traefik Ingress on kubernetes Bare Metal Cluster). Currently, going through my HAproxy does not work. Everything works if I use traefik internal loadbalancer in the kubernetes cluster, as you have depicted.
Hi, thanks for watching. Unfortunately thats the way with recent versions of Traefik implementation. No longer an external load balancer is needed. With cloud native in mind, its easier to use cloud providers load balancer readily. Hence I used MetalLB locally to get load balancing solution.
@@justmeandopensource Thank you for replying. So, is it best practice to have the loadbalancer in the cluster be the single point of entry for client request?
I dont think this works anymore for 1.25+ . Setting pvc, traefik pod gets stuck in pending. Also can't connect to dashboard even after disabling volume persistence
Hi, thanks for reporting this. As with any other tech videos, this one looks like it has gone outdated. I will try to do another one with latest version when I have some time. Cheers.
I just tested the instructions exactly as said in this video on a kubernetes 1.26 cluster and it worked exactly as shown in this video. So the problem is on your setup. May be try starting from scratch on a new cluster.
I had the exact same issue and I finally found what it was. If you followed the NFS setup videos and then came back to this video, there's one extra line you have to add in your storageclass "class.yaml": allowVolumeExpansion: true I noticed that setting set to true in this video after Venkat said it generated NFS using helm and mine was set to false. Hope this helps you too.
Love your videos! Learnt alot from them. I would like to ask if you can do some videos about Traefik v2/Nginx Controller with HAProxy, since I wanted to try setting up in some VPS nodes which has only 1 public ip each. I am unable to use MetalLB which allows to reserve private IPs for access.
Excellent videos. easy to follow along. One question : We now installed the latest version of Traefik using helm. What if a new version is released. Can we upgrade to that version using helm ? What would be the syntax of the upgrade command(s)
Hi Hans, thanks for watching. Yes you can upgrade to latest version with helm. Otherwise there is no point in using helm. Command would be something like this helm install --upgrade --version repo/chart release-name
Thanks for the video.. Can you also make a future video on how to use cert-manager to get certificates for traefik instead of using the built in traefik certificate? Because you can't get HA with traefik certificates yet unless you are using traefik enterprise version. I've managed to do it in my cluster but I'm using a rather ugly way to do it.
Thanks for watching. The third video in this series due to be released on Wednesday is all about certificates in Traefik. Hope that will share some light or else we can look into it more specifically. Cheers.
This is a great video! Thanks for taking the time to share. I'm curious, what application are you using @4:02? it seems like a clean way to do documentation.
This is an excellent tutorial, I love the pacing, and your instructions are very clear. That said, I am trying to follow this using K3s, which I installed without servicelb and traefik. I get to the part where I need to setup the Nginx LoadBalancer to test out MetalLB. This is not working, the EXERNAL-IP value is stuck in pending state. I noticed your comes back immediately. Is doing this with K3s a non-starter? If not, can you suggest where I might look to troubleshoot this? I have checked each of the components and they all seem to be happy, I see no warnings, or errors. I have verified that the IP range I set in the configmap is open and available on my network.
Hi, thanks for watching. I don't see any reason why you wouldn't be able to use/run metallb in a k3s cluster. In fact, I just tried it myself on my k3d cluster (k3s cluster in docker containers) and metallb seems to be working fine and I was able to expose an nginx pod as loadbalancer service.
I enjoy watching your videos, but I do not understand why every topic starts with installing Kubernetes. For a pure development environment I would just use k3d or k3s.
Hi Sami, thanks for watching. I am currently focussing on Kubernetes. I have list of other areas to cover and Vagrant is one of them. But unfortunately due to tight schedule, I can't guarantee that I can start it soon. Cheers.
Great video !. I do not understand why dynamic storage is used here and not a nfs pv and pvc as in your video "NFS Persistent Volume in Kubernetes Cluster" I guess that i am missing something here... Regards, Hans
@@justmeandopensource As I understand a new volume will be recreated every time I restart and the previous volume is archived. So any data on the volume is not there anymore. Or am I wrong?
I tested it, the direcory is created on my nfs server when i start the deployment of the nfs-client-provisioner. If i stop the nfs-client-provisioner en start it again a new directory is created that does not have the content of the previous one. What are your settings for onDelete and archiveOnDelete in the class.yaml ? I guess i am missing something. Maybe not enough knowledge .... I do not see why this is persistent storage. If i stop and start the nfs-client-provisioner the existing data has gone.
Aaah, now i got it working. i changed my class.yaml to include a pathPattern : parameters: onDelete: retain pathPattern: "${.PVC.namespace}-${.PVC.name}" I did not use a pathPattern before. If there is no pathPattern, every time the nfs-client-provisioner is started, a new empty directory with a unique name is created. That is why i saw an empty directory after each restart... Thanks for your patience.
@@weitanglau162 If this is your home network, then its not ideal to allow connections from public internet. You will have to go through the pain of setting up port forwarding in you router.
I love that you used lynx to check that the external ip was working! :)
Haha. Thanks for watching.
Appreciated to update old videos with latest one.
Thanks for watching.
Thanks so much,
its like a istio
Hi Pankaj, thanks for watching.
I love that you use lynx text browser. I haven't seen that in years. Great. 🙏😍
Haha. Did that show my age or my ops background? Thanks for watching. Cheers.
Sir, your videos are amazing. Your level of knowledge is so deep! Please do a video on IPv6 enabled k8s clusters!
Hi, thanks for watching. IPv6 is in my list and I will have to play with it for a while before I can confidently do a video. Will get to it at some point :) Cheers.
Really really waiting for this. Thank you!
No worries. Thanks for watching. Cheers.
Awesome stuff. Will def be checking out your other videos.
Hi Martin, Thanks for watching.
I'm so glad I found your channel. Such a great resource pool for me next month ❤
Hi Ali, thanks for watching. Cheers.
An excellent job.
This new presentation model was too top.
Hi Juliano, thanks for the feedback. Cheers.
Amazing! Thank you for this great job!!
Hi Marcos, thanks for watching.
I'm really looking forward to the next parts cause I'm trying to get it running but tls refuses to work. Hope that the next parts come out soon! Love your work!
Hi Daniel, thanks for watching. I have Traefik videos scheduled everyday this week until Saturday. Tomorrow is about ingreds routes and the day after is all about tls. Cheers.
thank you. another helpful series
Hi Murat, thanks for watching. Cheers 😊
Nice video sir.. It helps to improve the understanding
Hi Amit, Thanks for watching.
Awesome, really excited!
Thanks.
@@justmeandopensource Hi, can u make videos on openEBS ?
Yeah its in my list.
@@justmeandopensource Thanks
Great video, Venkat.
Hi Laurynas, thanks for watching.
Your videos is amazing! Thank you a lot!
Many thanks for watching.
Great tut, keep up the good work !
Hi Paul, thanks for watching.
Very good tutorial. This helped me a lot. Thanks
But instead of NFS, I used Longhorn
Cool.
Very good tutorial
Thanks for watching.
Hey, What do you do when the first IP works and the others don't? I've run into that exact issue!
Do you mean on the MetalLB? Thats an infrastructure issue.
Very good and detailed tutorial, if tcp-based weighted round robin can be de demonstrated that would be best! Many appreciates。
Hi Sheng, thanks for watching. I will see if I can add it. Cheers.
Hi sir, juste a question please, with my traefik I have a bad gateway when I try access to my rabbitmq , but when I refresh one or 2 time I get rabbitmq management interface, also I can access it by port-forward I don't understand why please if you can help thanks so mush
Hello! In the first line thank you very much for your videos! I have learned a lot but I still have a silly question: I cannot comprehend why we need MetalLB at all? I have provisioned a single-node k8s cluster with traefik 2.x as ingress and I am ablle to access my application from the Internet. As traefik works as a loadbalancer, what are the use cases of metallb? Thank yo very much for you answer in advance!
Your work is brilliant. Very well polished, consise, perfectly paced, wtih attention to detail par none. The matial is presented in a perfectly logical order, your presentation style is clean and articulate. The _vast_ majority of RUclips-ers would do very well for themselves attempting to emulate your grace and clarity; you really ought to do a demonstration on how to do demos ;) Now the heavy technical question...which terminal app do you use with what plugins? I'm enthrawled...I've never really spent time making my 'terminator' work any better than it does out of the box aside from vim awesome and a spattering of additions. I mostly code in vim or some graphical IDE and likely work harder than I need to with the support of cool tools like you have shown...so I'd love to know what you use, if you don't mind.
Thank you for your inspiration and education!
Ahhh...rewatching, you answered my question already ruclips.net/video/dEAtD9PVr_Q/видео.html
@@RustyWeise Many thanks for watching. I keep switching my setup to find the best comfortable working experience with awesome tools.
My terminal setup currently is something like what I have described in this video.
ruclips.net/video/PUWnCbr9cN8/видео.html
Hi,
I like your arch setup, can you do complete video on it including installation & customizations , we all driver installation,
If I install traefik in my cluster, what IP should I forward my external load balancer to ? Will it be the IP of the master node?
How to find the compatible traefik version for various kubernetes versions and releases
A great intro video. I can't wait for more. #Venkatrules
You won't be disappointed. I have all videos scheduled until Saturday. Thanks for watching. Cheers.
Great introduction, but I think you'll want to add the setting "ExternalTrafficPolicy: Local" so that the source IP is being retained. You then might also need to scale up the replica so that the VIP can be attached to any worker node.
Hi, thanks for watching. Yeah we could make it a lot better in terms of availability and fault tolerance. I wanted to keep the flow simple as majority of my audience are beginners :)
@@justmeandopensource yeah you're totally right but I was really confused first as to why I just couldn't get my real IP forwarded to my services and explaining that in a further video would be fantastic! :)
EDIT: You can easily check if X-Forwarded-For is checked with the whoami container by containous as well.
"/tmp/traefik-values.yaml" E212: Can't open file for writing
Many thanks for this interesting tutorial. Unfortunately I cannot edit the yaml file as instructed, I am getting "access denied" when I am trying to save the change. There is anything I am missing or to do. Please help because I am stuck.
Great video, thank you! I'm just entering into this world of kubernetes and I have a question, why is it necessary to use MetalLB instead of the internal load balancer provided by Traefik? I think that probably the approach that you gave is more for an use case to integrate traefik with an external load balancer, so locally you're using MetalLB but in cloud we can replace it by ELB, what I say is correct?
Hello and thank you for your great videos!! i will be happy if you make more viedeo about Helm3 . charts from scratch !! thank you
Thanks for watching. I can certainly look into that area. Cheers.
Greetings! Could you explain how would you make this current setup work with a HAProxy external loadbalancer, like you did in the first video back in 2019 ([ Kube 32 ] Set up Traefik Ingress on kubernetes Bare Metal Cluster).
Currently, going through my HAproxy does not work. Everything works if I use traefik internal loadbalancer in the kubernetes cluster, as you have depicted.
Hi, thanks for watching. Unfortunately thats the way with recent versions of Traefik implementation. No longer an external load balancer is needed. With cloud native in mind, its easier to use cloud providers load balancer readily. Hence I used MetalLB locally to get load balancing solution.
@@justmeandopensource Thank you for replying. So, is it best practice to have the loadbalancer in the cluster be the single point of entry for client request?
Just in the time when I needed it. Thanks!. Also when other videos in the same series will be uploaded?
Hi Varun, thanks for watching. Everyday this week until Saturday. Cheers.
@@justmeandopensource Great news! Thanks!
@@varundeolase6802 You are welcome. Cheers.
if using cloud AWS, should we still use a metal bare?
If you mean metallb, it is not supported in most of the public clouds.
@@justmeandopensource maybe I could use the Alb for this case?
I dont think this works anymore for 1.25+ . Setting pvc, traefik pod gets stuck in pending. Also can't connect to dashboard even after disabling volume persistence
Hi, thanks for reporting this. As with any other tech videos, this one looks like it has gone outdated. I will try to do another one with latest version when I have some time. Cheers.
I just tested the instructions exactly as said in this video on a kubernetes 1.26 cluster and it worked exactly as shown in this video. So the problem is on your setup. May be try starting from scratch on a new cluster.
@@justmeandopensource Okay will retry. What is the browser you are using?
@@vivahernando1 Qutebrowser
I had the exact same issue and I finally found what it was. If you followed the NFS setup videos and then came back to this video, there's one extra line you have to add in your storageclass "class.yaml": allowVolumeExpansion: true
I noticed that setting set to true in this video after Venkat said it generated NFS using helm and mine was set to false.
Hope this helps you too.
Thank you - BTW which CNI you used for your test kubernetes cluster ? (calico, weavenet, etc) ?
Calico. Thanks for watching.
Love your videos! Learnt alot from them.
I would like to ask if you can do some videos about Traefik v2/Nginx Controller with HAProxy, since I wanted to try setting up in some VPS nodes which has only 1 public ip each.
I am unable to use MetalLB which allows to reserve private IPs for access.
Hi, thanks for watching. I can try but can't guarantee due to busy schedule. Cheers.
Excellent videos. easy to follow along.
One question :
We now installed the latest version of Traefik using helm.
What if a new version is released. Can we upgrade to that version using helm ?
What would be the syntax of the upgrade command(s)
Hi Hans, thanks for watching. Yes you can upgrade to latest version with helm. Otherwise there is no point in using helm.
Command would be something like this
helm install --upgrade --version repo/chart release-name
what is the navigator you ues?it amazing!!
Hi, thanks for watching. I guess you are asking about my web browswer. It is Qutebrowser which is python based keyboard driven/centric browser.
What laptop are you using? I am in the market for a new one.
www.currys.co.uk/gbuk/computing/laptops/laptops/dell-xps-13-9310-13-4-laptop-intel-core-i7-512-gb-ssd-silver-10217072-pdt.html
Thanks for the video.. Can you also make a future video on how to use cert-manager to get certificates for traefik instead of using the built in traefik certificate? Because you can't get HA with traefik certificates yet unless you are using traefik enterprise version. I've managed to do it in my cluster but I'm using a rather ugly way to do it.
Thanks for watching. The third video in this series due to be released on Wednesday is all about certificates in Traefik. Hope that will share some light or else we can look into it more specifically. Cheers.
This is a great video! Thanks for taking the time to share. I'm curious, what application are you using @4:02? it seems like a clean way to do documentation.
Hi Jack, thanks for watching. Thats a website where you can draw pictures/illustrations etc. So handy for me since I found it.
Excalidraw!!
This is an excellent tutorial, I love the pacing, and your instructions are very clear. That said, I am trying to follow this using K3s, which I installed without servicelb and traefik. I get to the part where I need to setup the Nginx LoadBalancer to test out MetalLB. This is not working, the EXERNAL-IP value is stuck in pending state. I noticed your comes back immediately. Is doing this with K3s a non-starter? If not, can you suggest where I might look to troubleshoot this? I have checked each of the components and they all seem to be happy, I see no warnings, or errors. I have verified that the IP range I set in the configmap is open and available on my network.
Hi, thanks for watching.
I don't see any reason why you wouldn't be able to use/run metallb in a k3s cluster. In fact, I just tried it myself on my k3d cluster (k3s cluster in docker containers) and metallb seems to be working fine and I was able to expose an nginx pod as loadbalancer service.
@@justmeandopensource Thanks for the speedy reply.
@@RhythmicDevil No worries. Cheers.
I enjoy watching your videos, but I do not understand why every topic starts with installing Kubernetes. For a pure development environment I would just use k3d or k3s.
Hi Dennis, thanks for watching. I personally use k3d. My videos are mostly administrator focused hence less of a developer content.
could you please present some video for vgrant
Hi Sami, thanks for watching. I am currently focussing on Kubernetes. I have list of other areas to cover and Vagrant is one of them. But unfortunately due to tight schedule, I can't guarantee that I can start it soon. Cheers.
Great video !.
I do not understand why dynamic storage is used here and not a nfs pv and pvc as in your video "NFS Persistent Volume in Kubernetes Cluster"
I guess that i am missing something here...
Regards,
Hans
Hi, why do you want to manually create PVs while you can set up dynamic volume provisioning?
@@justmeandopensource
As I understand a new volume will be recreated every time I restart and the previous volume is archived.
So any data on the volume is not there anymore.
Or am I wrong?
@@justmeandopensource
OK thanks I will test that.
I tested it, the direcory is created on my nfs server when i start the deployment of the nfs-client-provisioner.
If i stop the nfs-client-provisioner en start it again a new directory is created that does not have the content of the previous one. What are your settings for onDelete and archiveOnDelete in the class.yaml ?
I guess i am missing something. Maybe not enough knowledge ....
I do not see why this is persistent storage. If i stop and start the nfs-client-provisioner the existing data has gone.
Aaah, now i got it working.
i changed my class.yaml to include a pathPattern :
parameters:
onDelete: retain
pathPattern: "${.PVC.namespace}-${.PVC.name}"
I did not use a pathPattern before.
If there is no pathPattern, every time the nfs-client-provisioner is started, a new empty directory with a unique name is created.
That is why i saw an empty directory after each restart...
Thanks for your patience.
Great video!
Just wanna know how do we route public domain to the private Load Balancer IP?
Hi Lau, thanks for watching. What do you mean by private load balancer IP?
@@justmeandopensource thanks for replying :)
Basically the IP address that metallb gives to traefik service
@@weitanglau162 If this is your home network, then its not ideal to allow connections from public internet. You will have to go through the pain of setting up port forwarding in you router.
@@justmeandopensource I see. This is because I wanna expose my services to the public haha. Thanks for the advise!