I used to be a PHP developer 20 years ago. I had no idea I'd be interested in this stuff but I seem to be. It's like lego but cooler. Your videos are very clear, concise, and most of all.... WELL PACED! nice one my man.
Travis, u inspired me 5 months back to go on Devops journey. I am using udemy to prepae for Devops job. I completd : 1.Linux Administrator course : Imran Afzal 2.Bash scripting----> Narendra 3.Git --> Bogdan stashuk 4.Maven --> Bharath Thippireddy As of now, trying to complete AWS [Fundamentals + Solutions Architect Associate] by stephen . Next will b : SonarQube---->Nexus---->Docker--->Kubernetes--->Terraform--->Ansible---->Prometheus & Graffana---->Python
Bro any chance of getting job for you with current knowledge? my issue is I will learn for 4-5 months then i gave up and i have to start all again. I lost hope too. I think we have to learn a lot to get into DevOps junior roles too
@@goutham9831 prepare ur own notes. Keep revising it I am about to finish Stephen mareck solution architect associate. Docker kubernetes terraform Ansible r left now. 4 months.
Docker is a huge prerequisite as well as docker compose and swarm to get you used to microarchitecture, how networks function and communicate as well as services in between containers, and swarm to get you used to orchestration. If you do those first kubernetes will be a lot more familiar to you.
Great video, just wanted to clarify something around the dockerfile, it is a build script that is ONLY used to specify how to build the image as well as any startup commands (CMD). It is not referenced post-build of the image in any way, and running an image post-build does not re-execute the commands in the docker file, these binaries are already installed within the image. At boot time, only the CMD as defined within the dockerfile is executed at runtime
I’m Kubernetes certified CKA & CKAD I’m fully agree with you about the prerequisites and also something really important to emphasize is mastering Linux before going to K8s because this is the only OS Kubernetes deals with
I am studying an intro to containers and I would lije to know in your opinión it is worth certification in containers. Are there many jobs oportunities? Thanks for your answer.
Awesome vid. I'm a sales rep at leading k8s provider and here I am researching containers & k8s from other voices - shows the complexity! Also, regarding networking -- this is hands down where we get held up the most with customer education / onboarding from poc to hardened dev/prod environments . Hands down the most gotchas here. Frustratingly so, but hey opportunities!
Sorry to be picky and I am sure you're aware of this really - but at 3:20 you explain "so when you run the image its going to follow those steps, it adds your package.json, run npm install, copy over your code.." - it has already done this to build the image, when the image is run a container from the image is created and it runs the ENTRYPOINT and/or CMD instructions to run the desired application
@3:18 - "When you run that image, it's going to follow those steps" Umm.. No, I am pretty sure the Dockerfile is run only when creating the image, not when *running* the image (creating a container, from the image, but not using a Dockerfile). If you have an image, it doesn't need to copy package.json (for instance) into it from anywhere .. and you have the image, probably not the source directory that you used to build the image. (so where would you copy it from??)
I came here to say this. What he is stating is simply not correct, The whole purpose of this step is to build the software and package it up, not copy source files and make the container build the artifacts each time - that would be a huge resource waste.
Thanks for confirming. Learning docker myself and this is what I understood too, but when I watched what was said in this video I started to question if I had misunderstood the concept.
Podman and crio are good options. Podman has more useful command line options as a client than docker. Also, you can run as a normal user and not be root.
That's basically the same thing I guess.. It has a lot of enhancements though and the best thing is it automatically starts up when you start the the laptop/computer. No more systemctl start docker 🥲
@Travis I believe there is a mistake in your explanation with regards to Docker @3:20. When one creates a container out of the image, all the necessary dependencies/libraries are already baked into the image, there is no re-running of those instructions when a container is started. If everything runs again and needs to install the libs/dependencies, there is no point of creating the image. One can just use a shell script file as a start up script.
I've been working with k8s as a hobby for over 2 years now. I was looking for (and still working on) a way to run a platform I'd like to commercialize at some point. And, although I have a successfully running k8s cluster with like 30+ apps running, and I'm heading to my own very opinionated platform. I still wouldn't ever call myself an expert. I'm a jack of all trades and master of none. These basics are definitely important, but there is a whole lot more.
That being said, I think a general understanding of what k8s means in terms of not only being an orchestration management system for containerized apps is important, but also understanding the fluidity between development, devOps and turning k8s into a PaaS for your company, is hugely important too. You could theoretically have all development outside of k8s, however, I know this is missing out on taking full advantage of k8s. And, having all dev'ing done outside of k8s is going to be a constant source of uncertainty, when things go wrong. Is the root cause the dev environment or is it the k8s environment? Yes, containers are supposed to help alleviate this concern, but they don't really in the end. The only way to remove that big question is to have the code being developed actually running in a real cluster with all the accessible services too. Once you see this path and can offer it to devs in an automated way, you also increase development velocity manyfold. It's why SRE is being renamed platform engineer slowly but surely. And yeah again, I'm just doing this as a hobby. But, I love it. 😁
Should you start with Ultimate AWS Certified Cloud Practitioner CLF-C02 course or can you start directly with Certified Kubernetes Administrator (CKA) with Practice Tests course on udemy ? I don't intend to be an expert, just learn how to deploy my apps on docker and kubernetes and for future jobs
Because they don’t understand what they do, what they want, and what went wrong in previous projects . Nothing went wrong btw, they just took men who lie in theory CV and girls who look h*t and s**y … instead of good learned men and women
I would include learning Docker Compose as the second step. It's a highly effective method for running containers locally, regardless of whether they're configured to run in a Kubernetes cluster or other environments.
10:41 I very much disagree. Use the actual command as doing so makes copying examples painless. Should you alias kubectl as `k`, you now need to manually edit your example every time you want to share a command in a document, over chat, or when creating a bug report. Alias flags you want to (nearly) always pass. Only shorten commands if you know you won't be sharing examples very often. I alias `n` to nano for example.
I wanted to comment about alias too. You only create expensive code that is not readable thus harder to maintain. For instance, using powershell commands with its splatting feature is way better than bash commands with so many letters. Alias is maybe only useful for writing faster snippets of codes (you can write fast with autocompletion too).
Thank you, i'm a network engineer taking on the devop role at our company. At the moment I am struggling to deploy apps on a eks cluster. I will be checking out the udemy course you recommended.
My company deployed kubernetes on premise. Whats the main difference by running and administraiting an application in the cloud and on premise in a own big data center of the company?
Thank you. This is a great video. Well explained and to the point. There is one topic that i feel is missing in all these container videos, it is an explanation of the cost of containerization and its overhead. I often see containers being used as a mechanism for parallelism. The starting point is often a single threaded language like JavaScript or Python. However, containers are really poor (10 to 100 x) alternatives to multi-threading and software processes. I get that scale out requires less commitment than scale up, but if you need 50x more hardware, then you need a lot of hardware.
We had a VM that was scaling up and ended up costing us a lot of money. It was doing a lot of work in parallel. We did the rearchitecture to containers and succeeded in lowering the cost of operation by literally 10x
Hi Travis. Thankyou for the Video. Even though, I understand the Docker / Kube basics, how would you relate this with Serverless Architecture using Lambda Functions? I started to learn SAM but cannot able differentiate in terms of which one Supersedes when creating server or deploying Apps? Any general guidelines would be helpful. Thankyou.
IMO, Lamdas/Serverless Architecture are a completely different ways to create solutions. A Lambda is an ephemeral process that does not reside on a server, nor in a container. It is a complete, standalone stateless application; the only way to have data persistence is to have the Lambda send whatever data to an external store. If you need to orchestrate Lambdas, Step Functions are what is usually used. Containers, on the other hand, can persist as long as you want, and can be more heavy-weight. Since containers can scale up and down, external stores can be mounted for data persistence.
Holy Smokes... I didn't know it was this involved... I'm halfway home in terms of the requirements, but the stuff about networking will absolutely need to be learned....
I learnt kube from a developer perspective in around 6 months, plus additional 1 months for Jenkins and additional CI/CD stuff. I am now learning something like ansible and planning for AWS. 😅
Hi Good One, But I think for some one to even understand containers, they should not go use docket first, with out docker they should try to containerize a process in their linux using cgroups and namespaces linux kernel concepts which is infac he fundamental concept behind containers and then kubernetes is just a orchestration and management of these containers.. with some intelligence for load placing etc... across nodes... or machines ( physical.. or virtuall.. )
great video but I thought it was going to be learning about "Kubnernetes" not Kubernetes... you fooled me :) seriously though, truly great and appreciated video. Thanks.
Self delete isn't the best option. Policies that crush one party but prop up the other party will always test the sanity of the one being crushed. Can't delete policy.
And yet everyone ive interviewed recently requires it 🙄 Lots of small businesses will just loose money running k8 setups because they think its needed...
If you have less then 100 servers and < 1000 containers don't even bother go into k8s. Most of orgs even some medium-big businesses don't need such complexity to maintain their MONOLITH-microservice architecture)
You haven't mentioned the helm package manager, and where companies will force you to write 100 services for an app if there is no helm for the given target image, and you have to do it in a day :D.
1. Containerization (like Docker)
2. Cloud Basics
3. YAML
4. Networking Basics
5. Linux
Thanks , :)
thanks
Thanks dude
hero
Better than 13min ad I just watched.
I used to be a PHP developer 20 years ago. I had no idea I'd be interested in this stuff but I seem to be. It's like lego but cooler. Your videos are very clear, concise, and most of all.... WELL PACED! nice one my man.
Definitely like legos... that's exactly the way I think about it.
Travis, u inspired me 5 months back to go on Devops journey.
I am using udemy to prepae for Devops job.
I completd :
1.Linux Administrator course : Imran Afzal
2.Bash scripting----> Narendra
3.Git --> Bogdan stashuk
4.Maven --> Bharath Thippireddy
As of now, trying to complete AWS [Fundamentals + Solutions Architect Associate] by stephen .
Next will b : SonarQube---->Nexus---->Docker--->Kubernetes--->Terraform--->Ansible---->Prometheus & Graffana---->Python
Bro any chance of getting job for you with current knowledge? my issue is I will learn for 4-5 months then i gave up and i have to start all again. I lost hope too.
I think we have to learn a lot to get into DevOps junior roles too
@@goutham9831 wat tools have u completed till now ?
@@goutham9831😮
@@goutham9831 prepare ur own notes. Keep revising it I am about to finish Stephen mareck solution architect associate. Docker kubernetes terraform Ansible r left now. 4 months.
🎉
Docker is a huge prerequisite as well as docker compose and swarm to get you used to microarchitecture, how networks function and communicate as well as services in between containers, and swarm to get you used to orchestration. If you do those first kubernetes will be a lot more familiar to you.
Great video, just wanted to clarify something around the dockerfile, it is a build script that is ONLY used to specify how to build the image as well as any startup commands (CMD). It is not referenced post-build of the image in any way, and running an image post-build does not re-execute the commands in the docker file, these binaries are already installed within the image. At boot time, only the CMD as defined within the dockerfile is executed at runtime
One of the best decisions I made lately: subbed to this man's channel. I love the practical approach he takes in sharing the knowledge.
I’m Kubernetes certified CKA & CKAD
I’m fully agree with you about the prerequisites and also something really important to emphasize is mastering Linux before going to K8s because this is the only OS Kubernetes deals with
100%. Are you for that CKS now?!?! I've been putting off the CKA now for like 4 years, myself. 😆
@@TravisMedia I surely will 😊
I am studying an intro to containers and I would lije to know in your opinión it is worth certification in containers. Are there many jobs oportunities? Thanks for your answer.
GIGA CKAD
Have you been able to get a job because of your cka or ckad?
Awesome vid. I'm a sales rep at leading k8s provider and here I am researching containers & k8s from other voices - shows the complexity!
Also, regarding networking -- this is hands down where we get held up the most with customer education / onboarding from poc to hardened dev/prod environments . Hands down the most gotchas here. Frustratingly so, but hey opportunities!
Sorry to be picky and I am sure you're aware of this really - but at 3:20 you explain "so when you run the image its going to follow those steps, it adds your package.json, run npm install, copy over your code.." - it has already done this to build the image, when the image is run a container from the image is created and it runs the ENTRYPOINT and/or CMD instructions to run the desired application
@3:18 - "When you run that image, it's going to follow those steps"
Umm.. No, I am pretty sure the Dockerfile is run only when creating the image, not when *running* the image (creating a container, from the image, but not using a Dockerfile). If you have an image, it doesn't need to copy package.json (for instance) into it from anywhere .. and you have the image, probably not the source directory that you used to build the image. (so where would you copy it from??)
I came here to say this. What he is stating is simply not correct, The whole purpose of this step is to build the software and package it up, not copy source files and make the container build the artifacts each time - that would be a huge resource waste.
Thanks for confirming. Learning docker myself and this is what I understood too, but when I watched what was said in this video I started to question if I had misunderstood the concept.
Podman and crio are good options. Podman has more useful command line options as a client than docker. Also, you can run as a normal user and not be root.
Where I work, we abandoned docker and started using Podman about a year ago.
Why?
That's basically the same thing I guess.. It has a lot of enhancements though and the best thing is it automatically starts up when you start the the laptop/computer. No more systemctl start docker 🥲
Only principal software engineer knows, probably.
podman is essentially same as docker but has rootless advantage and will most likely be the replacement for docker
@@SweetMelon-sd1qu you nailed it, it is more secure than docker
Concept 1 is actually around 2:10
3:10 just a little mistake in the video. Image only runs the command "CMD" steps not the build steps too
@Travis I believe there is a mistake in your explanation with regards to Docker @3:20.
When one creates a container out of the image, all the necessary dependencies/libraries are already baked into the image, there is no re-running of those instructions when a container is started.
If everything runs again and needs to install the libs/dependencies, there is no point of creating the image. One can just use a shell script file as a start up script.
you are quickly being my favourite tech tuber , you make great content travis
I didn’t understand anything but the feeling they were pleasant🤩
google has a paper on Borg that is a good reading to understand some of the design decisions behind kubernetes
can you share the link please?
I've been working with k8s as a hobby for over 2 years now. I was looking for (and still working on) a way to run a platform I'd like to commercialize at some point. And, although I have a successfully running k8s cluster with like 30+ apps running, and I'm heading to my own very opinionated platform. I still wouldn't ever call myself an expert. I'm a jack of all trades and master of none. These basics are definitely important, but there is a whole lot more.
That being said, I think a general understanding of what k8s means in terms of not only being an orchestration management system for containerized apps is important, but also understanding the fluidity between development, devOps and turning k8s into a PaaS for your company, is hugely important too. You could theoretically have all development outside of k8s, however, I know this is missing out on taking full advantage of k8s. And, having all dev'ing done outside of k8s is going to be a constant source of uncertainty, when things go wrong. Is the root cause the dev environment or is it the k8s environment? Yes, containers are supposed to help alleviate this concern, but they don't really in the end. The only way to remove that big question is to have the code being developed actually running in a real cluster with all the accessible services too. Once you see this path and can offer it to devs in an automated way, you also increase development velocity manyfold. It's why SRE is being renamed platform engineer slowly but surely.
And yeah again, I'm just doing this as a hobby. But, I love it. 😁
Outstanding. Thanks for keeping it short and to the point.
concept 1 doesn't start till @2:00
Thanks Travis. I went to that udemy class and doing the absolute beginner class. It’s really good.
Should you start with Ultimate AWS Certified Cloud Practitioner CLF-C02 course or can you start directly with Certified Kubernetes Administrator (CKA) with Practice Tests course on udemy ? I don't intend to be an expert, just learn how to deploy my apps on docker and kubernetes and for future jobs
Brilliant video! Thanks!
I'm a software engineer and I have a question: why are companies starting to demand Devops skills from software engineers?
Because they don’t understand what they do, what they want, and what went wrong in previous projects .
Nothing went wrong btw, they just took men who lie in theory CV and girls who look h*t and s**y … instead of good learned men and women
Cause merging dev and ops is the purpose if devops
I would include learning Docker Compose as the second step. It's a highly effective method for running containers locally, regardless of whether they're configured to run in a Kubernetes cluster or other environments.
very clear explanation thank you so much 🥰
What are your recommendations of resources for learning the terminal?
perfect timing! I was just looking for this. Keep up the great work!
10:41 I very much disagree. Use the actual command as doing so makes copying examples painless. Should you alias kubectl as `k`, you now need to manually edit your example every time you want to share a command in a document, over chat, or when creating a bug report.
Alias flags you want to (nearly) always pass. Only shorten commands if you know you won't be sharing examples very often. I alias `n` to nano for example.
I wanted to comment about alias too. You only create expensive code that is not readable thus harder to maintain. For instance, using powershell commands with its splatting feature is way better than bash commands with so many letters. Alias is maybe only useful for writing faster snippets of codes (you can write fast with autocompletion too).
Thank you, i'm a network engineer taking on the devop role at our company. At the moment I am struggling to deploy apps on a eks cluster. I will be checking out the udemy course you recommended.
Hi, I am on that course right now, amazing option, from my point of view the best one I've ever seen about Kubernetes Administration.
Thanks man. Awesome video. Keep it up!
"Not knowing Networking Concepts is nothing to be ashamed of.." - tears in my eyes, Yuge weight off my chest.
Very informative, thanks
Gem of a video. The perfect one for my learning path. 🙏
Same situation here. Familiarity with Helm is also needed.
Yes!
Having a silky smooth voice is the #1
Great video, YAML stands for "Yet Another Markup Language"
Thank you! Even with upper-intermediate English I can understand all of the ideas about DevOps ways. Awesome channel
Thank you for this video, I was looking into learning kubernetes. I like your videos you deserve more subs man :D
My company deployed kubernetes on premise. Whats the main difference by running and administraiting an application in the cloud and on premise in a own big data center of the company?
Thank you. This is a great video. Well explained and to the point.
There is one topic that i feel is missing in all these container videos, it is an explanation of the cost of containerization and its overhead.
I often see containers being used as a mechanism for parallelism. The starting point is often a single threaded language like JavaScript or Python.
However, containers are really poor (10 to 100 x) alternatives to multi-threading and software processes.
I get that scale out requires less commitment than scale up, but if you need 50x more hardware, then you need a lot of hardware.
We had a VM that was scaling up and ended up costing us a lot of money. It was doing a lot of work in parallel. We did the rearchitecture to containers and succeeded in lowering the cost of operation by literally 10x
Thank you so much for the video, it was very helpful!! I'm so happy that I found your channel, your content is amazing!! Keep it up 💗⭐
Great info for starting k8s
Thanks for k8s beginners roadmap
Very clear overview!
So, is there overlap between Kubernetes and docker-compose?
amazing video, thanks!
Thank you sir. you are always on point. u got my sub
Perfect timing. Thank you.
Thank you for sharing this to learn Kubernatares ❤
Hey nice one ! Any prompts on how to go about getting a legacy app containerized..
Thanks Travis for the detail Video. it helps from where we starts.
such great information ❤, well you got a new friend. Thanks @Travis
Hi there.
i am now a CCNA, RHCSA, RHCE and i have knowledge in WinServ
Do you recommend me going through AWS then some basic Docker and then to CKA?
Thank you for this video i wanted to learn all these concept and your video gave me a starting point for that.
Hi Travis. Thankyou for the Video. Even though, I understand the Docker / Kube basics, how would you relate this with Serverless Architecture using Lambda Functions? I started to learn SAM but cannot able differentiate in terms of which one Supersedes when creating server or deploying Apps? Any general guidelines would be helpful. Thankyou.
IMO, Lamdas/Serverless Architecture are a completely different ways to create solutions. A Lambda is an ephemeral process that does not reside on a server, nor in a container. It is a complete, standalone stateless application; the only way to have data persistence is to have the Lambda send whatever data to an external store. If you need to orchestrate Lambdas, Step Functions are what is usually used. Containers, on the other hand, can persist as long as you want, and can be more heavy-weight. Since containers can scale up and down, external stores can be mounted for data persistence.
Holy Smokes... I didn't know it was this involved... I'm halfway home in terms of the requirements, but the stuff about networking will absolutely need to be learned....
I learnt kube from a developer perspective in around 6 months, plus additional 1 months for Jenkins and additional CI/CD stuff. I am now learning something like ansible and planning for AWS. 😅
congrats! you are on 1/20 step to understand how everything works
@@ordinaryggso true 😂 IT hole is real
Came at the right time. Thank you
how about kubernetes on on-prem server? should we continue to work on that or move to cloud?
Depends on your requirements. They both have their advantages (and thus disadvantages).
What makes AWS EKS a “managed service?”
Thank you, this is really helpful
Great video, thank you!
Travis, watching your videos gives me good motivation. I also started learning new stuff!! thanks 👍
How about hinglish in the course ?
Amazing info video, completely agree to you Travis. Thanks a lot.
Loved it bro
Thanks for all you do! 💪🏽
You teach from fundamental and I loved it. It gave me a blueprint abt how to understand something in this case kubernetes in the big picture
Content begins at 2:10. You are welcome.
Hi Travis. Amazing content. I wanted to inquire on your video setup. Would you be willing to answer a few questions?
Clear English on the course and is 84% off for the next 10 hours, wish me luck
How did you get a job that you did not have the experience about the job:)
Ctl as control feels a lot more natural to me. Ctl as cuttle seems like raised-by-wolves speak.
please provide the Kubernetes Udemy course link
Travis always has helpful tips..Let's Gooooooooo
Great information
Thank you for this video for us learning kubernetes
Hi Good One, But I think for some one to even understand containers, they should not go use docket first, with out docker they should try to containerize a process in their linux using cgroups and namespaces linux kernel concepts which is infac he fundamental concept behind containers and then kubernetes is just a orchestration and management of these containers.. with some intelligence for load placing etc... across nodes... or machines ( physical.. or virtuall.. )
Thank you so much :)
great video but I thought it was going to be learning about "Kubnernetes" not Kubernetes... you fooled me :)
seriously though, truly great and appreciated video. Thanks.
Self delete isn't the best option. Policies that crush one party but prop up the other party will always test the sanity of the one being crushed. Can't delete policy.
1. Few workloads need kubernetes.
And yet everyone ive interviewed recently requires it 🙄
Lots of small businesses will just loose money running k8 setups because they think its needed...
tank you for the course tips
Great vid Travis! In the same boat as you were haha, need to deploy containers in kubernetes for work😂
Thank you !
I love how I heard him said "kube cuddle"
Great video
Thank you
damn, nicely explained
If you have less then 100 servers and < 1000 containers don't even bother go into k8s. Most of orgs even some medium-big businesses don't need such complexity to maintain their MONOLITH-microservice architecture)
thanks, helpful
Prerequisites Linux an docker ..
I don't think any other thing needed..
YAML aint a superset of JSON ;-)
Travis media?
I was tought that was other channel.
Thank you ❤
You haven't mentioned the helm package manager, and where companies will force you to write 100 services for an app if there is no helm for the given target image, and you have to do it in a day :D.
great info
Thank you.
nice for beginner
Thank you