I've watched 5 other full courses and none of them explained Ingress and Nodeports as clearly as you did. I finally get it! Thank You! How do I support you or buy your full course?
Happy you found it useful! If you'd like to support the channel, be sure to like and subscribe. If there are any other topics you'd like us to cover, leave us a comment, and we'll see what we can do!
3 года назад
Very nice explanation, very visual and to the point
This is a good short video on Ingress and service but once you need to make one correction on ip addresses. 169 range IP address in not public address. It is reserved range.
thks a lot. I will share your video with my collegues. It was difficult to understand the differences between nodeport and ingress. It's easier with your explanation. One of my collegues said that it would be better to have one ingress per service. What do you think about it ?
Hey there, thanks for watching. Generally, it makes more sense to configure multiple back-end services with a single ingress resource. You can use path-based routing to resolve to different services.
In order for the Ingress resource to work, the cluster must have an ingress controller running. Unlike other types of controllers which run as part of the kube-controller-manager binary, Ingress controllers are not started automatically with a cluster. You may need to deploy an Ingress controller such as ingress-nginx. You can choose from a number of Ingress controllers. Kubernetes as a project supports and maintains AWS, GCE, and nginx ingress controllers.
So if I understand correctly. The ingress lives on the API server? Or is that another device? Where does the public IP live? I'm assuming again on the API controller or API server.
Most production environment use cases are to use a load balancer and Ingress in combination! Is it better to load balancer-> ingress or ingress-> load balancer? I think ibm have more scenes! ibm answer is authoritative! How cloud-native components such as Ingress Istio combine them in a production environment and use them in practice! How to design the specific architecture?So ask you guys
the IP addresses in the beginning of your explanation is belong to the worker nodes not pods. pods generally get their IPs from DHCP server inside the overlay network of the cluster
Nicely explained. Simple and efficient. One question i have about ingress: can you build it so that it routes traffic say like 70% on one service and 30% on another when you've got same pods created in each service? I don't know if i was clear enough (:
Hi Alexander! You can’t do that with Kubernetes Ingress at the moment, but you can do it with a service mesh such as Istio. Check out “Istio Traffic Management”.
Nice video, thank you for share, I have a question, I recently heard at a conference that it was a good practice to use multiple ingress controllers in Kubernetes. Why?
Ok Great, but I still have some questions. Let’s say I have a cluster with 3 Master and 3 worker nodes for HA and an external Load Balancer. Now I need to set up ingress, is the ingress automatically HA and where is it running? On the master or worker node just like any other pod and to wich target I point my load balancer that the traffic gets to the Ingress?
When you are on a cloud provider, the load balancer will be automatically provisioned for you, but if you are on bare-metal, you should do that yourself manually.
I have a question, so essentially an Ingress is a Kubernetes component which is not in the form of actual process running as part of cluster itself, but instead ingress makes sures that it maintains the front end and backend configuration needed for an external load balancer process to pick it up and execute outside the cluster?
Hi there! Ingress is an object type in Kubernetes. The Ingress Controller processes this Ingress object & the configuration/metadata contained within it, to configure the layer 7 proxy contained within the Ingress Controller, in order to expose your applications and services to the outside world. You can read more here 👉 ibm.co/2TGlWpp
Hi , I have a question. My helm chart is deploying 2 pod replicas and they were exposed to DNS host name through ingress host, during the helm upgrade i have to ensure that atleast one pod is available and serving the requests with minimal downtime. How can i introduce this behaviour??
It's simpler than it looks. The key is the image has been flipped horizontally by the editor. You can get a behind the curtain look here: ruclips.net/channel/UCKWaEZ-_VweaEx1j62do_vQcommunity?lb=Ugzf5SL_yh9NglCJzgF4AaABCQ
i have a service with 10 replicas of a pod exposed by a service on k8s cluster. I want to 4 replicas to serve public APIs (lets say they have /public prefix) and the other 6 replicas to server private(internal) APIs (let's say they don't have /public prefix). I don't think ingress with help here. Is there any other way, this can be achieved?
There's 1 thing which is incorrect here (let me know if it's not). Ingress does not have a single External ip address. don't you have to expose ingress using nodeport?
Hi, can anyone teach me how to scale the pods of a single deployment across nodes, or say, group pods from different nodes as a service? I have thought that you can only group pods with labels within a node? I have also thought that you can only scale out pods for a deployment within a node. Is it possible to control that?
I don't understand how can all of the instructor can right backwards? Is it a job requirement for getting into IBM of you guys are simply flipping the video horizontally?
Best 5 minutes video ever for Ingress
Very cool! in 5 minutes you are able to explain that better then a lot of time of anyone else :)
One of the best videos I have seen. Within 5 minutes you explained a very valid points. God bless bro.
I love these clips with transparent board. great content
Only 5 mins and so much content... hats off
Thank you so much for the feedback, glad you liked it! 🤠
amazing explanation.... loved it.
Thanks for the nice explanation (and esp. of pointing to the "Kubernetes explained" in the beginning)! Looking forward to other informative videos.
short and very well explain. well done IBM team
Thank you! 🙏
ultimate explanation in short time.. gr8 work
Nicely explained. Can we please get a video on default and custom operators on Kubernetes?
Really well explained, thumbs up !
Very nice format thanks!
The best explanation, I have seen many videos, but this the gem of all
Thank you, Teja, really humbled by your feedback! 🙏 Glad you enjoyed the video, stay tuned for more like this. 💪
Your all videos are great!
amazing work, i like the alert before the interesting parts..
A service is a collection of pods that are loads balanced
Wow, this is the best explanation. Super cool.
Thanks for the quick video and explanation. On point.
Very nice video. Congratulations
Thanks for the great explanation on Ingress. Got a little better understanding on it.
Oh. I would love to watch your personal channel, hard to find videos exactly from you, but they are the best.
Great explanation
This is just what I need. Amazing work!
commenting so that I get recommended more of ibm contents !
always well said, and to the point. 🙏
Excellent explanation!
I've watched 5 other full courses and none of them explained Ingress and Nodeports as clearly as you did.
I finally get it! Thank You!
How do I support you or buy your full course?
Happy you found it useful! If you'd like to support the channel, be sure to like and subscribe. If there are any other topics you'd like us to cover, leave us a comment, and we'll see what we can do!
Very nice explanation, very visual and to the point
This is a good short video on Ingress and service but once you need to make one correction on ip addresses. 169 range IP address in not public address. It is reserved range.
Thank you for the helpful summary
really clean nice explanation
Thank you, Hesham! 🙂
ez to understand... thanks u so muchhh
No better way to explain Ingress. 👍
thks a lot. I will share your video with my collegues. It was difficult to understand the differences between nodeport and ingress. It's easier with your explanation. One of my collegues said that it would be better to have one ingress per service. What do you think about it ?
Hey there, thanks for watching. Generally, it makes more sense to configure multiple back-end services with a single ingress resource. You can use path-based routing to resolve to different services.
Awesome job!!
looking forward to some practical demonstration of ingress
Excellent!
Thank you.
Fantastic!
Great video. Just one confusion I have. Do we need a load-balancer like Ngnix when using Ingress or Ingress itself act as a load balancer
you don't need LB anymore when using ingress.
In order for the Ingress resource to work, the cluster must have an ingress controller running.
Unlike other types of controllers which run as part of the kube-controller-manager binary, Ingress controllers are not started automatically with a cluster. You may need to deploy an Ingress controller such as ingress-nginx. You can choose from a number of Ingress controllers. Kubernetes as a project supports and maintains AWS, GCE, and nginx ingress controllers.
Very Good
Hello Sir, I would like to know if we are using ingress then what type (Cluster IP or Node Port or Load Balancer) should we use in Service Object.
Are nodes and pods the same? It seems like you use them interchangeably.
2:35 why IP is going to change when adding pods?
Good one
Sai Venom .. what a badass name !!!
Very nice!
Damn! This was a good quick lesson!
Absolutely
So if I understand correctly. The ingress lives on the API server? Or is that another device? Where does the public IP live? I'm assuming again on the API controller or API server.
Same question here, but I don't think API server would be the answer.
Well done bro !
well done
Thanks for the video :)
Excellent video about Kubernetes once again!!
Thanks you
Most production environment use cases are to use a load balancer and Ingress in combination! Is it better to load balancer-> ingress or ingress-> load balancer? I think ibm have more scenes! ibm answer is authoritative! How cloud-native components such as Ingress Istio combine them in a production environment and use them in practice! How to design the specific architecture?So ask you guys
2:10 what a perfect angle
and he's writing backward!!
the IP addresses in the beginning of your explanation is belong to the worker nodes not pods. pods generally get their IPs from DHCP server inside the overlay network of the cluster
At 1:27: He said “each one of those NODES has a publicly accessible IP”. He is right.
As you said that we can do routing in ingress; so is that ok to remove the nginx routing logic and implement in ingress ??
Nicely explained. Simple and efficient.
One question i have about ingress: can you build it so that it routes traffic say like 70% on one service and 30% on another when you've got same pods created in each service? I don't know if i was clear enough (:
Hi Alexander! You can’t do that with Kubernetes Ingress at the moment, but you can do it with a service mesh such as Istio. Check out “Istio Traffic Management”.
@@IBMTechnology thank you very much for the answer :D
Nice video, thank you for share, I have a question, I recently heard at a conference that it was a good practice to use multiple ingress controllers in Kubernetes. Why?
You're awesome!
very well done, thx!
Ok Great, but I still have some questions.
Let’s say I have a cluster with 3 Master and 3 worker nodes for HA and an external Load Balancer. Now I need to set up ingress, is the ingress automatically HA and where is it running? On the master or worker node just like any other pod and to wich target I point my load balancer that the traffic gets to the Ingress?
Just a suggestion to not whiteboard at the very bottom of the frame, it ends up overlapping with the subtitles (if enabled)
for load balancer type is it true that we must add 1 ip per service? i thought we can use 1 ip with different port per service..
Sorry for 2 years late, but i have question. Between ingress and pod, is it necessary to have load balancer to handle pod?
Excellent! GReat explanation and I congratule you for writing in a mirror way :D, that's hard!
greatly explain in 5 mini.. to my knowledge/experience NLB is Network Load Balancer right ? correct me if I am wrong !!
When you are on a cloud provider, the load balancer will be automatically provisioned for you, but if you are on bare-metal, you should do that yourself manually.
Basically Ingress in ALB that can do Path based routing.
I have a question, so essentially an Ingress is a Kubernetes component which is not in the form of actual process running as part of cluster itself, but instead ingress makes sures that it maintains the front end and backend configuration needed for an external load balancer process to pick it up and execute outside the cluster?
Hi there!
Ingress is an object type in Kubernetes. The Ingress Controller processes this Ingress object & the configuration/metadata contained within it, to configure the layer 7 proxy contained within the Ingress Controller, in order to expose your applications and services to the outside world.
You can read more here 👉 ibm.co/2TGlWpp
how many ingress can be created for a cluster?
Hi , I have a question. My helm chart is deploying 2 pod replicas and they were exposed to DNS host name through ingress host, during the helm upgrade i have to ensure that atleast one pod is available and serving the requests with minimal downtime. How can i introduce this behaviour??
how are you writing
are you writing mirror image ?
great video, may I ask what applications did you use to create this video. the writing part with marker is amazing
It's simpler than it looks. The key is the image has been flipped horizontally by the editor. You can get a behind the curtain look here: ruclips.net/channel/UCKWaEZ-_VweaEx1j62do_vQcommunity?lb=Ugzf5SL_yh9NglCJzgF4AaABCQ
Thanks
how these pens are working is there a glass or it's some kind of computer trick
Seriously, his writing backward is what is bothering me more than anything.
Did you really write from right to left????
More than likely they write normal and reverse the video
Do you actually write mirrored or is there some magic behind this? :D hats off and thanks for the great explanation!
Hi Fabian! There's a little bit of magic behind it. 😉 We shared some backstage secrets on the Community page, check it out here 👉 ibm.co/3AbglI3
i have a service with 10 replicas of a pod exposed by a service on k8s cluster. I want to 4 replicas to serve public APIs (lets say they have /public prefix) and the other 6 replicas to server private(internal) APIs (let's say they don't have /public prefix). I don't think ingress with help here. Is there any other way, this can be achieved?
There's 1 thing which is incorrect here (let me know if it's not). Ingress does not have a single External ip address. don't you have to expose ingress using nodeport?
how do you know if its incorrect if you are asking other to validate?
Too cool
Hi, can anyone teach me how to scale the pods of a single deployment across nodes, or say, group pods from different nodes as a service? I have thought that you can only group pods with labels within a node? I have also thought that you can only scale out pods for a deployment within a node. Is it possible to control that?
Great content, amazing explanation 👍
It took me 2 and half mins, watched it in 2x :)
if hands on lab please tell me
Ambassador next , plz
I don't understand how can all of the instructor can right backwards?
Is it a job requirement for getting into IBM of you guys are simply flipping the video horizontally?
Did you forget to explain a pair of ingress components, ingress controller and ingress itself?
Basically, Ingress is like Zuul in Spring Cloud
The way you depicted is as if the ingress creates LB outside of node, which is not true and totally confusing.
After the latest update ingress prime always crashes.... So i can Not play.... the funny thing os that says " fixed crashing "
You are writing in mirror or you write normal? Sorry had to ask
Hey there! We shared some backstage "secrets" of our videos on the Community page, check it out here 👉
ibm.co/3pAmVTQ 🙂
Is he drawing on my monitor?
See ibm.biz/write-backwards
cool
bro how can you confidentially write on reverse :D
Premium content 👍
More than the explanation Im always fascinated by how these people write in reverse
I can't write that neat forwards
1 out of 5 minutes you are speaking about anything, but Ingress.
psy venom