I hope this clarifies all your questions, if not - leave a comment and I will try to answer it! If it was helpful, please leave a like and subscribe 😊 👍 Follow me on Instagram for behind the scenes content 🎬: bit.ly/2F3LXYJ 🔗 Useful Links: * kubernetes.io/blog/2020/12/02/dockershim-faq/ * kubernetes.io/blog/2020/12/02/dont-panic-kubernetes-and-docker/ ▬▬▬▬▬▬ T I M E S T A M P S ⏰ ▬▬▬▬▬▬ 0:00 - Intro 0:18 - Why has Docker been deprecated? 3:10 - Container Runtime alternatives? 4:17 - What does this change mean for YOU as a Kubernetes User? 5:07 - What does this change mean for YOU as a Kubernetes Administrator? 5:30 - when using Managed Kubernetes Service 6:32 - with self-managed Kubernetes cluster 7:58 - When do you need to take action? 8:45 - Impact on using Kubernetes locally with Minikube and Docker Desktop 9:42 - Still learn Docker? CI/CD Pipeline with Docker? 12:00 - Wrap-Up
I think what's left unanswered for me is, okay so if docker and all of its networking and etc is going away. whats taking over? how is the ipam and host-container nat going to be taken care of?
Nice video: This video just clarified the hoopla and garbage that a lot of people are throwing around. Basically nothing changes for people who are using docker tech correctly, which is proabably 98% of people out there anyways, and the benefit is kubernetes made their product more lightweight and not have to worry about the hell annoying updates of docker that happen all the time, and can break (updates are good, but annoying) Disclaimer: I am not an experienced netop/devops, my main role is developer, but I maintain servers on AWS (8 servers) for my side business, and lead netops/devops (not 100% professional, but probably better than the avg 2 yr experienced fool that has blogs online about bullcrap as I have been working with online tech for 18+ years). As for what others are saying: should I learn docker? I think the easiest answer is this: if you don't know why you are using docker, then maybe you don't use it? If you know why you want to use docker, then use it. Docker is a good choice, all others saying: this is better, that is better - yah probably 1-sided experience. Docker is easy to use, esp for beginners and THIS IS MOST IMPORTANT. If people are saying "10% faster" and stuff like that, this is pure garbage, if you know systems, you aren't concerned about 10% faster for your servers/containers, 10% would be MAYBE important for SQL server, when you are at the cusp of a breaking point of the system, but even then, you gotta change the implementation: split DBs into two servers. So rule: 10% faster means nothing, really, because you just spin up another load balanced server anyways to reduce load - these are just foolish people if they want to really make this into an issue. FASTER BY 10% DOESN'T MATTER REALLY. I will give a case where we want to use docker, and docker is good - docker runs on major OSes, and allows you to build containers (which is basically functions like a VM, without the overhead that most VMs have) - for example, we have a system where we have 4 different servers, we use cloud providers and Centos 7 (screw IBM RH), to develop locally, it's great to use dockers so that we can mimic the interactions of servers right on our computer (local dev env). DOCKER IS GOOD FOR DEVS! So on production, when to use containers? You have a choice now (AWS) - use EC2 instances OR use containers. What are the differences, and benefits for each? A lot goes to having experience - if you running lightweight processes that don't do much, like the majority of nodeJs servers people write, and services which they offer, then containers are a great option on cloud providers especially, because it's cheaper to have 1-2 live and 1 backup containers, then lets say than paying for 3 servers. It's almost always safer to have several servers running than 1 super server (what happens if that hardware dies on the super server, your services are GONE). Keyword: load balancing. Now 1 type of server is our web servers, so obviously you want to load balance two or more in production. CONTAINERS COULD BE CHEAPER THAN SERVERS. Is it easier to maintain containers than lets say traditional servers? I'm going to say if you are spinning up 64 load balanced servers, it's gonna get costly, and containers makes more sense, as they were built for huge numbers. But in all small to medium network sizes of small to medium companies, or start ups, it doesn't really matter tbh. If you actually have 1K concurrent users, then containers are great! If you running even apache php servers, you could go with containers too! it's understanding the resources/performance and costs. CONTAINERS ARE GOOD IF YOU WANT TO LOAD BALANCE 100 "SERVERS" ON LOAD BALANCER. In the end, when using cloud providers, like AWS, you have a choice of deployment now - ec2 load balanced or fargate containers. As far as I am concerned, if you don't have 1k+ concurrent users, you might not necessarily use fargate (or kubernetes) and container images. But you never had to because you could always use the traditional methods of using servers. DEPENDING ON REALISTIC SCALES AND WHAT YOUR SYSTEM IS, CONTAINERS OR NON-CONTAINERS ARE VIABLE IN A LOT OF GENERAL CASES. Anyone pushing anything real hard outside of this are just "experts" in their own fields, because all container tech are about the same, docker is great for the dev process, and it's big, like in this video. But those tools AREN'T NEEDED on your production servers (docker full features) because you shouldn't be building debugging your containers on production and for sure, you shouldn't be using docker network layer AT ALL on production. DOCKER RUNNING IN PRODUCTION IS DANGEROUS (SECURITY), DON'T NEED FULL DOCKER ON PRODUCTION ANYWAYS. In the end, you just have to think of a container as a VM and you can 32/64 "servers" in containers on a computer, think of it that way. It's probably safer and better than running 1 service on a computer (in general), which may crash in an edge case or some other weird issue (which maybe due to degradation over time, bad 3rd party software).
Nana, great video. Isn't the question a bit misleading? Is it not the case that K8s are dropping support for docker shim rather than docker? The K8s support for OCI has never been in doubt and the docker obviously is compliant with OCI. K8s are not required to support individual containers but rather focus on the Open Container Initiative, which is still the case. Would you agree to that?
As someone with a self managed and on-prem K8s cluster, I will say that changing the container runtime from docker to containerd was not only fairly straight forward, but I'm also noticing better node resource availability which makes sense given the heaviness of dockershim vs natively supported cri. I will say, the process to change runtimes was not well documented as of yet, but if anyone needs assistance and is also using kubdadm to manage their on-prem clusters, just hit me up!
Bird's eye view of the recent K8's announcement, there were articles that mislead (for various reasons) "Docker" (instead of Dockershime) being dropped.
Hi Nana, just FYI, u mentioned in the video that EKS has adopted containerd as the container runtime, which is not 100% the case, to be precise. EKS for Fargate has adopted containerd, but not for EKS for EC2.
It's the first time I see your channel and I'm really impressed by the quality of your explanations and the way you're presenting them. Pleased continue this great work. Thank you.
There is still going to be an issue. The docker-in-docker pattern will no longer work in Kubernetes, because there is no Docker engine underneath and I don't think that's something criO or containerD can do alone. The Docker images are OCI compliant but the Docker engine isn't, and that's the whole problem Kubernetes are having - the engine, and hence its removal. So, if you're using /var/run/docker.sock in ur cluster, that's going to break.
No matter how difficult or confusing the subject may be, Nana knows the way to break it down to make super duper easy. My 2 year old has no further questions, so do I :) Thanks Nana...
Hey, I hope you are doing good. My doubt is whether can we still use docker containers during development and later during production make the conversion or whatever to deploy. I have very little knowledge in docker and still learning. I haven't even get started with kubernetes too, I just know few commands. If my question feels dump please ignore it :)
Docker is still good for development as it ships with everything you need to build a fixed, repeatable environment. And docker uses containerd, so if you learn about docker, you learn about containerd anyway.
When we watch a video we get so many questions,.how why, and this video answers all those questions the next minute/seconds... Thank you so much. Subscribed!!
Awesome video, huge thanks for the clear explanation. Sadly I can only wow this and give a single like. In my mind I'm sending you many more though. 👍🖖😃
To save you time: What's actually being deprecated by the K8s team: ruclips.net/video/7KUdmFyefSA/видео.html Who's taking over instead: ruclips.net/video/7KUdmFyefSA/видео.html Conclusion: You can either: a.-use Docker runtime ONLY if you do things the K8s way. b.-Install Dockershim separately by referring to Mirantis.
Just brilliant Nana. Keep up the good work !. I have referred a couple of your videos and became your big fan. The way you explain the concept is just amazing. God bless you.
Blimey...that's like Closed Source approach, that means that other microservice platforms are going to either do the same or continue to use the full docker support approach. No more Docker way of doing things then, but k8s way only. Correct me if I'm wrong please. I definitely would need the different networking modes Docker offers for example...Do you know how many people use EC2 and build K8s from scratch?
😂😂🤣😂 JUST IN I.T. I mean I get it, but it's still funny to see how things change so fast. remember when employers were asking knowledge and experience on both of these is a must? funny sad.
Nana, you are the best to explain the change, I was boiling the ocean to understand the change, but you nailed it in 12 min. Thanks a lot, keep up the good work. always following you.
I don't understand - if Docker can run on any container runtime, because both CRI-O and Docker comply to some OCI standard, why did Kubernetes dropped support for Docker then? How does it make it more efficient in terms of resource usage if you're still going to build Docker image with all that stuff inside and it will still be used in Kubernetes?
Once I did use docker during the development, because the previous it leader really liked it, but then I dropped it and the work sped up by few times. I never really understood why would you even use this tool for a middle sized application that we were developing. Our app only used one virtual server with 8 cores at the time, later expanded to 16. The amount of code was merely about half a million of lines. We didn't use Microservices because this isn't really a useful thing for middle or small apps. We were running docker on windows and that middle sized app was enough to make it run sluggishly ( yea, we did set up everything to maximise speed of the virtual machine). Later I decided to change my machine to the same operating system on which our server worked and the work sped up enormously Guys, don't follow the hypes, think before you will use something. Don't use docker just because it is popular, in 90% of cases you don't need any docker at all and it won't speed up anything.
Kubernetes is dropping Docker support - What does it mean for YOU? If you need to watch this video, then there is huge probability it means nothing to YOU :)
It's not actually dropping docker support? k8 supports the open container initiative, which docker is fully compatible with. So, we can always use docker with k8. They've made the move because they want people to realise that there are other options.
Thanks for the rundown of the cluster changes. Having to fight to find documentation for setting a local private K8 cluster, I was set a back that this was going to be a bad thing. But with your explanation, sound like it will be smooth sailing. I would take this as a advantage to move Docker closer to the top of the stack. Containers are very slow compared to VM's(around half speed. You want VM's over containers for time sensitive services) So to take out another layer of slowness would be a bonus, if you choose the container option. If for some reason you still need the shim , it's an option.
So, if I use minikube as a replacement of Docker Desktop. With this update am not going to be able to do "docker build" or "docker push" from my host machine anymore? 🤔
So what if I want Home assistant server to run in a container and say LG Web OS or Kodi on another. And I want to be able to manage my containers in a grafical web interface. What software should I use? Worth mentioning is that I'll be running it on a raspberry pi 4
misleading video - docker are installed itself as containerd + the rest tools...thus you still can just install docker + k8s in a usual way. Nothing changed.
But after pulling several docker image to Kubernetes node , how to delete those image without docker install, like how we are docker rmi , please suggest solution to delete docker images in my k8's master node , i am using contained now
Thanks again, nice and clear! What are actually the differences/advantages/disadvantages of all these different runtimes and why do we actually have so many?
So we can still use Docker to create containers, even if k8s is not using the Docker Runtime. But is this a sign that we should consider other tools for container creation? Or do we expect Docker to be the best choice for that role for some time to come?
I hope this clarifies all your questions, if not - leave a comment and I will try to answer it!
If it was helpful, please leave a like and subscribe 😊 👍
Follow me on Instagram for behind the scenes content 🎬: bit.ly/2F3LXYJ
🔗 Useful Links:
* kubernetes.io/blog/2020/12/02/dockershim-faq/
* kubernetes.io/blog/2020/12/02/dont-panic-kubernetes-and-docker/
▬▬▬▬▬▬ T I M E S T A M P S ⏰ ▬▬▬▬▬▬
0:00 - Intro
0:18 - Why has Docker been deprecated?
3:10 - Container Runtime alternatives?
4:17 - What does this change mean for YOU as a Kubernetes User?
5:07 - What does this change mean for YOU as a Kubernetes Administrator?
5:30 - when using Managed Kubernetes Service
6:32 - with self-managed Kubernetes cluster
7:58 - When do you need to take action?
8:45 - Impact on using Kubernetes locally with Minikube and Docker Desktop
9:42 - Still learn Docker? CI/CD Pipeline with Docker?
12:00 - Wrap-Up
I think you did not clearly answered to the question: "Should ppl learn/use docker or move forward and use containerd or cri-o?"
I think what's left unanswered for me is, okay so if docker and all of its networking and etc is going away. whats taking over? how is the ipam and host-container nat going to be taken care of?
Nice video: This video just clarified the hoopla and garbage that a lot of people are throwing around. Basically nothing changes for people who are using docker tech correctly, which is proabably 98% of people out there anyways, and the benefit is kubernetes made their product more lightweight and not have to worry about the hell annoying updates of docker that happen all the time, and can break (updates are good, but annoying)
Disclaimer: I am not an experienced netop/devops, my main role is developer, but I maintain servers on AWS (8 servers) for my side business, and lead netops/devops (not 100% professional, but probably better than the avg 2 yr experienced fool that has blogs online about bullcrap as I have been working with online tech for 18+ years).
As for what others are saying: should I learn docker? I think the easiest answer is this: if you don't know why you are using docker, then maybe you don't use it? If you know why you want to use docker, then use it. Docker is a good choice, all others saying: this is better, that is better - yah probably 1-sided experience. Docker is easy to use, esp for beginners and THIS IS MOST IMPORTANT.
If people are saying "10% faster" and stuff like that, this is pure garbage, if you know systems, you aren't concerned about 10% faster for your servers/containers, 10% would be MAYBE important for SQL server, when you are at the cusp of a breaking point of the system, but even then, you gotta change the implementation: split DBs into two servers. So rule: 10% faster means nothing, really, because you just spin up another load balanced server anyways to reduce load - these are just foolish people if they want to really make this into an issue. FASTER BY 10% DOESN'T MATTER REALLY.
I will give a case where we want to use docker, and docker is good - docker runs on major OSes, and allows you to build containers (which is basically functions like a VM, without the overhead that most VMs have) - for example, we have a system where we have 4 different servers, we use cloud providers and Centos 7 (screw IBM RH), to develop locally, it's great to use dockers so that we can mimic the interactions of servers right on our computer (local dev env). DOCKER IS GOOD FOR DEVS!
So on production, when to use containers? You have a choice now (AWS) - use EC2 instances OR use containers. What are the differences, and benefits for each? A lot goes to having experience - if you running lightweight processes that don't do much, like the majority of nodeJs servers people write, and services which they offer, then containers are a great option on cloud providers especially, because it's cheaper to have 1-2 live and 1 backup containers, then lets say than paying for 3 servers. It's almost always safer to have several servers running than 1 super server (what happens if that hardware dies on the super server, your services are GONE). Keyword: load balancing. Now 1 type of server is our web servers, so obviously you want to load balance two or more in production. CONTAINERS COULD BE CHEAPER THAN SERVERS.
Is it easier to maintain containers than lets say traditional servers? I'm going to say if you are spinning up 64 load balanced servers, it's gonna get costly, and containers makes more sense, as they were built for huge numbers. But in all small to medium network sizes of small to medium companies, or start ups, it doesn't really matter tbh. If you actually have 1K concurrent users, then containers are great! If you running even apache php servers, you could go with containers too! it's understanding the resources/performance and costs. CONTAINERS ARE GOOD IF YOU WANT TO LOAD BALANCE 100 "SERVERS" ON LOAD BALANCER.
In the end, when using cloud providers, like AWS, you have a choice of deployment now - ec2 load balanced or fargate containers. As far as I am concerned, if you don't have 1k+ concurrent users, you might not necessarily use fargate (or kubernetes) and container images. But you never had to because you could always use the traditional methods of using servers. DEPENDING ON REALISTIC SCALES AND WHAT YOUR SYSTEM IS, CONTAINERS OR NON-CONTAINERS ARE VIABLE IN A LOT OF GENERAL CASES.
Anyone pushing anything real hard outside of this are just "experts" in their own fields, because all container tech are about the same, docker is great for the dev process, and it's big, like in this video. But those tools AREN'T NEEDED on your production servers (docker full features) because you shouldn't be building debugging your containers on production and for sure, you shouldn't be using docker network layer AT ALL on production. DOCKER RUNNING IN PRODUCTION IS DANGEROUS (SECURITY), DON'T NEED FULL DOCKER ON PRODUCTION ANYWAYS.
In the end, you just have to think of a container as a VM and you can 32/64 "servers" in containers on a computer, think of it that way. It's probably safer and better than running 1 service on a computer (in general), which may crash in an edge case or some other weird issue (which maybe due to degradation over time, bad 3rd party software).
]]i
Nana, great video. Isn't the question a bit misleading? Is it not the case that K8s are dropping support for docker shim rather than docker? The K8s support for OCI has never been in doubt and the docker obviously is compliant with OCI. K8s are not required to support individual containers but rather focus on the Open Container Initiative, which is still the case. Would you agree to that?
Went from "Oh Shit! we're screwed" to "Ah, this is fine" in 12 minutes.
Yep :D
200% clear, no more questions.. Great explanation.. Thank you..
Glad it was helpful for you 😊
I couldn't agree more. It's been a while since I watched a video about tech, with such a clear explanation of the topic - thx!
As someone with a self managed and on-prem K8s cluster, I will say that changing the container runtime from docker to containerd was not only fairly straight forward, but I'm also noticing better node resource availability which makes sense given the heaviness of dockershim vs natively supported cri. I will say, the process to change runtimes was not well documented as of yet, but if anyone needs assistance and is also using kubdadm to manage their on-prem clusters, just hit me up!
My most productive 12.21 minutes of today; Thank you Nana :-)
Glad to hear :)
Yes, the very first minute I saw, I thought kubernetes Twitter page got hacked 🤣🤣🤣
The Queen of DevOps, sharp as always!
Thank you Pedro! 😀
Bird's eye view of the recent K8's announcement, there were articles that mislead (for various reasons) "Docker" (instead of Dockershime) being dropped.
So did the title of this video!
Perfect explanation, just the thing I needed, just as request, can you do a tutorial on crio or containerd.
Thank you Nana... Actually, I never knew about this change and at first, I got worried, as I only worked with docker...
Clear, concise enough, saves people's time to look elsewhere. Nana, I support you.
Thanks for your feedback Allen, glad it's helpful :)
i wish i had discovered this channel earlier ... you got yourself a new subscriber
Thanks for sharing new valuable insights. Love your approach. All content creators should create value as maximum as possible.
Thanks so much! :)
Hi Nana, just FYI, u mentioned in the video that EKS has adopted containerd as the container runtime, which is not 100% the case, to be precise. EKS for Fargate has adopted containerd, but not for EKS for EC2.
It's the first time I see your channel and I'm really impressed by the quality of your explanations and the way you're presenting them. Pleased continue this great work. Thank you.
There is still going to be an issue. The docker-in-docker pattern will no longer work in Kubernetes, because there is no Docker engine underneath and I don't think that's something criO or containerD can do alone. The Docker images are OCI compliant but the Docker engine isn't, and that's the whole problem Kubernetes are having - the engine, and hence its removal. So, if you're using /var/run/docker.sock in ur cluster, that's going to break.
i am a simple man, i see Nana posts something i watch immediately :) !
😀
No matter how difficult or confusing the subject may be, Nana knows the way to break it down to make super duper easy. My 2 year old has no further questions, so do I :) Thanks Nana...
One of the best videos about this subject... but I just don't know why are you not able to say Azure, when you are listing clouds providers lol
i love how she explained.
She awesome!!!
Hey, I hope you are doing good. My doubt is whether can we still use docker containers during development and later during production make the conversion or whatever to deploy. I have very little knowledge in docker and still learning. I haven't even get started with kubernetes too, I just know few commands. If my question feels dump please ignore it :)
I am in love with this women. The way she explains the things is "Next Level".
Thanks Shahvez! :)
So now you can go back to the dictionary and learn about the difference between "woman" and "women".
@@goqsane "woman" is singular and "women" is plural.
So I'm I, I am now trying not to watch 😊🙆
Simp
Thank you Nana! Such a simple explanation🙏 Subscribed!
Thanks for sharing this update! Joining the fan club. :D
I was clapping👏when you clearly explaining the CI/CI part needs docker or not. thank's nana.
😊🙏
Nana, thank you for this!
Not dropping the support but just moving ! It is wrong interpretation Nana!
Yemi 4elli guelti wade7 n7n n5la3ne 3an projehne lahi ye5ser.
Thank you so much.
Whew, thank you for this video! As a newbie using docker for development this took a load off my mind.
Спасибо, Нана. Очень хорошее ревю
Exceptional knowledge!
So the question u haven't answered clearly. Should newcomers keep learning docker or containerd instead? Thanks for the great video !
Docker is still good for development as it ships with everything you need to build a fixed, repeatable environment. And docker uses containerd, so if you learn about docker, you learn about containerd anyway.
I taught containerd was only for running containers. And docker, among others, also for building them
You still should learn docker, since you build your images with it. containerd is just the component that runs the containers.
Simple, sweet, and concise. 👍
@Nana - Thanks for the clear explanation. I am a new subscriber to your YT channel, interested to learn Ansible, Docker & K8s tools.
good job
Tqs madam for giving good information
Alarmist title to attract views. Lame!
Now I am cristal clear about changes! Thanks, Tina!
Thanks for the explanation, Nana! Great as always :D
Thank you!! Waiting for an Istio and service mesh
Hi Dmitriy, that video will come next! 👍 I had to postpone it because of this announcement 🙈
Your videos really create impact on freshers like me... Thanks Nana form bottom of my heart.....
Happy to hear, thank you 🙂
When we watch a video we get so many questions,.how why, and this video answers all those questions the next minute/seconds... Thank you so much. Subscribed!!
Welcome to the channel 😊 Thank you so much for your positive comment!
Wow, very informative and such a high-quality content!
Thank you Daniel! Appreciate your comment!
Awesome video, huge thanks for the clear explanation. Sadly I can only wow this and give a single like. In my mind I'm sending you many more though. 👍🖖😃
To save you time:
What's actually being deprecated by the K8s team:
ruclips.net/video/7KUdmFyefSA/видео.html
Who's taking over instead:
ruclips.net/video/7KUdmFyefSA/видео.html
Conclusion: You can either:
a.-use Docker runtime ONLY if you do things the K8s way.
b.-Install Dockershim separately by referring to Mirantis.
Just brilliant Nana. Keep up the good work !. I have referred a couple of your videos and became your big fan. The way you explain the concept is just amazing. God bless you.
Thank you..Please make videos on AWS services
Thanks
I guess docker commands will no longer works in Kubernetes cluster (for troubleshooting purposes only). Is my understanding correct?
Blimey...that's like Closed Source approach, that means that other microservice platforms are going to either do the same or continue to use the full docker support approach. No more Docker way of doing things then, but k8s way only. Correct me if I'm wrong please. I definitely would need the different networking modes Docker offers for example...Do you know how many people use EC2 and build K8s from scratch?
😂😂🤣😂 JUST IN I.T. I mean I get it, but it's still funny to see how things change so fast. remember when employers were asking knowledge and experience on both of these is a must? funny sad.
Nana, you are the best to explain the change, I was boiling the ocean to understand the change, but you nailed it in 12 min. Thanks a lot, keep up the good work. always following you.
omg! you have structured your content perfectly.. in the middle I started think so in that case... and next second you have answer to that as well :-)
Good to hear 😊 Thanks for your great feedback!
I don't understand - if Docker can run on any container runtime, because both CRI-O and Docker comply to some OCI standard, why did Kubernetes dropped support for Docker then? How does it make it more efficient in terms of resource usage if you're still going to build Docker image with all that stuff inside and it will still be used in Kubernetes?
Once I did use docker during the development, because the previous it leader really liked it, but then I dropped it and the work sped up by few times. I never really understood why would you even use this tool for a middle sized application that we were developing. Our app only used one virtual server with 8 cores at the time, later expanded to 16. The amount of code was merely about half a million of lines. We didn't use Microservices because this isn't really a useful thing for middle or small apps.
We were running docker on windows and that middle sized app was enough to make it run sluggishly ( yea, we did set up everything to maximise speed of the virtual machine).
Later I decided to change my machine to the same operating system on which our server worked and the work sped up enormously
Guys, don't follow the hypes, think before you will use something. Don't use docker just because it is popular, in 90% of cases you don't need any docker at all and it won't speed up anything.
Thank you so much) you save me more time
Glad to hear Stanislav :)
Very much clear 👍🙂. as always great explanation Nana...👍
Thanks a lot :)
Kubernetes is dropping Docker support - What does it mean for YOU? If you need to watch this video, then there is huge probability it means nothing to YOU :)
It's not actually dropping docker support? k8 supports the open container initiative, which docker is fully compatible with. So, we can always use docker with k8. They've made the move because they want people to realise that there are other options.
Thanks for the rundown of the cluster changes. Having to fight to find documentation for setting a local private K8 cluster, I was set a back that this was going to be a bad thing. But with your explanation, sound like it will be smooth sailing. I would take this as a advantage to move Docker closer to the top of the stack. Containers are very slow compared to VM's(around half speed. You want VM's over containers for time sensitive services) So to take out another layer of slowness would be a bonus, if you choose the container option. If for some reason you still need the shim , it's an option.
Excellent as always!! How do you anticipate all the questions that can arise!! Just awesome..
😀 thank you so much! I read a lot of your comments and questions and try to answer them in the videos 😊
Very clearly explained. Thanks for the Video!!!
So, if I use minikube as a replacement of Docker Desktop. With this update am not going to be able to do "docker build" or "docker push" from my host machine anymore? 🤔
What if you have created a cluster using KOPS (github.com/kubernetes/kops)? Do you need any extra configuration ?
Wow!! Thanks for very clear explanation...
You're welcome, happy it was useful for you :)
Excellent explanation. Thank you
Perfect video, very clarify answers my questions about this news! thanks for your video
super fantastic... you are mind blowing..
very much hard topic now i understand in few minutes..
credit goes to ur video..
thanks
If I am on 19.x and upgrade to 22.x, will the run time get migrated to containerd automatically without me changing anything? Any impact>
So what if I want Home assistant server to run in a container and say LG Web OS or Kodi on another. And I want to be able to manage my containers in a grafical web interface. What software should I use?
Worth mentioning is that I'll be running it on a raspberry pi 4
Very good explanation, i subscribed and going to watch all your videos Nana. please put more videos for learning :)
Awesome clear explanation. Thanks a lot for publishing this video.
Subscribed!
I wish that I can give you 1000 likes
😊🙏
I love how well researched and articulated this is!!! 👏 thank you! I learnt a lot about “the in between bits” too.
Thank you Josh, appreciate your comment :)
misleading video - docker are installed itself as containerd + the rest tools...thus you still can just install docker + k8s in a usual way. Nothing changed.
wow! that's clear and precise. No bullshit. From where did you learn all these devops concepts? can you share your experiences
Can't u just build Open Container Images with docker and run your stuff anywhere?
Excellent explanation, Thanks Nana
Happy to hear, thank you :)
But after pulling several docker image to Kubernetes node , how to delete those image without docker install, like how we are docker rmi , please suggest solution to delete docker images in my k8's master node , i am using contained now
Thanks again, nice and clear! What are actually the differences/advantages/disadvantages of all these different runtimes and why do we actually have so many?
Hi I found multiple videos on docker and kubernet in playlist. Which one I follow?.
Thanks for the explanation
Use UBiqube MSA for Container and VNF orchestration
Thanks Nana, my wait and patience is over, now I can have my coffee.😊 👍😊
Perfect... More such videos are welcome whenever we hear new announcements or deprecations
after 4 month. kubernates running very well and faster than normal kvm / other virtualization. omg
Finally someone that explains clearly! Great Work!
Thanks a ton Sergio! 😊
Thanks for sharing it's very useful.
Thanks for your feedback :)
What DevOps is relevant in 2021. Please advise wrt career opportunities 🙏
Best explanation ever, thank you very much, Greetings! 😊
If I push a commit to a github repo and this triggers a build on my CI, should the CI test my code b4 or after building the image?
Am kind happy they dropped Docker it was really heavy in resources.
Thanks for the informative, well organized video.
is it because of things like Azure? Is docker simply being outclassed by the big dogs?
So we can still use Docker to create containers, even if k8s is not using the Docker Runtime.
But is this a sign that we should consider other tools for container creation? Or do we expect Docker to be the best choice for that role for some time to come?
Thanks for sharing it, really appreciate it.
Glad to hear, thank you Rashid :)
Thanks
This video is the best place to acknowledge that the author amongst other things is really beautiful girl.
Clear all the doubts.Great explanation