Follow Jorge for learning such cool stuff on Azure: Linkedin: www.linkedin.com/in/jorgearteiro UA-cam: ua-cam.com/users/azuretar AKS UA-cam Community: youtube.com/@theakscommunity
Thank you Mr.jorge for u r valueble time & efforts , very happy to subscribed & started following u. Abhi bro, u r not only giving technical skils& knledg , I have realized that each one day class is golden package of fund+ concpt + teach skills+ multiple imp other topics touch+big future+inspiration. its not only goodgreat . i pray god for atleast this much coments connectivity with u in future because of your grow spike.🖤💥
Hi Abhi. Thank you so much for bringing the huge useful stuff around the cloud and DevOps. Could you please make a video on recent downtime of meta all Instagram messenger Facebook outage for more than hour. Could you explain it and what are the measures they taking to avoid them in technical aspects.
Does this mean we no longer need Azure DevOps to create pipelines, or will all these features be integrated into Azure DevOps? I wish ACR could connect seamlessly to Azure DevOps, as well as AKS.
Great Video Abhishek One question which is like it took near about 30 sec to scale up the pods, but when it comes to Disney(Hotstar) and there is a match that is going even these 30 seconds would be costlier for us and in that scenario I would like to know your approach (or viewpoint) how would you tackle this?
Good question, you can scale up a new pod using HPA when the pod reach 80-90 percent of its capacity. So meanwhile the new pod and node(if required) are ready. Thats why we keep the nodes in deallocted status and dont delete them. Because new node creation with cloud init scripts execution might take a lot of time.
@@AbhishekVeeramalla yes we can use HPA, but when it comes to node it did take time right(30sec), what can we do in that scenario? basically I am thinking right now about bringing up the node and not thinking about pod.
Node did not take 30 seconds. It took < 30 seconds. Even lets say it took 30 secs, it is a less. When the other node is almost occupied(not completely occupied), we can spin up this new node.
@@AbhishekVeeramalla so you are saying if the node consumption is more than 70% spin up another node to save my time to spin up pod right😀 By the way I can't express my gratitude in words because even that would be less for the service that you provide, but even then THANK YOU 😃
Hi Abhishek & Jorge - I am yet to watch the vdo. Are we scaling pods or nodes ?. If we have huge surge in traffic, I dont think with Azure we can auto scale the nodes. We have to have PDBs for that or pre-scaling done.
Yes, we are scaling the nodes of the cluster. More specifically nodes of the node pool. AKS auto scales the nodes for you like I demonstrated on the video.
in AKS or EKS or GKE the control plane is managed by cloud provider correct? So what Jorge is demonstrating is we now can create a control plane in AKS or it was there from beginning and i was probably in dark till now😮💨
Hi, Abhishek i have a cluster autsclaer in aws eks and i have a 3 nodes. My application is running on these 3 nodes on off suddenly all my 3 nodes are down at same time. Since we have cluster scaling the nodes but when these are down and scaling the nodes scheduling the pods into new nodes it will take time so users will not access the application during the new nodes come up so how to achieve this
Follow Jorge for learning such cool stuff on Azure:
Linkedin: www.linkedin.com/in/jorgearteiro
UA-cam: ua-cam.com/users/azuretar
AKS UA-cam Community: youtube.com/@theakscommunity
Thank you Mr.jorge for u r valueble time & efforts , very happy to subscribed & started following u.
Abhi bro, u r not only giving technical skils& knledg , I have realized that each one day class is golden package of fund+ concpt + teach skills+ multiple imp other topics touch+big future+inspiration.
its not only goodgreat . i pray god for atleast this much coments connectivity with u in future because of your grow spike.🖤💥
amazing jorge, deallocation the node is best practice.
Absolutely!
Great Video !!
Thanks
Hi Abhi. Thank you so much for bringing the huge useful stuff around the cloud and DevOps. Could you please make a video on recent downtime of meta all Instagram messenger Facebook outage for more than hour. Could you explain it and what are the measures they taking to avoid them in technical aspects.
Amazing episode , it is sad thing to AWS doesn't support deallocate state
abshike you really make a devops course as celebration
😎😍
First of all, thanks for the great session. I have one query, can we make these changes (profile and deallocate) to existing AKS cluster?
Does this mean we no longer need Azure DevOps to create pipelines, or will all these features be integrated into Azure DevOps? I wish ACR could connect seamlessly to Azure DevOps, as well as AKS.
Hi Abhi, please make a video about Karpenter.
Sure ✔️
Karpenter is also available on AKS.
@jorgearteiro Thank you, Jorge for your reply.
Great Video Abhishek
One question which is like it took near about 30 sec to scale up the pods, but when it comes to Disney(Hotstar) and there is a match that is going even these 30 seconds would be costlier for us and in that scenario I would like to know your approach (or viewpoint) how would you tackle this?
Good question, you can scale up a new pod using HPA when the pod reach 80-90 percent of its capacity. So meanwhile the new pod and node(if required) are ready.
Thats why we keep the nodes in deallocted status and dont delete them.
Because new node creation with cloud init scripts execution might take a lot of time.
@@AbhishekVeeramalla yes we can use HPA, but when it comes to node it did take time right(30sec), what can we do in that scenario? basically I am thinking right now about bringing up the node and not thinking about pod.
Node did not take 30 seconds. It took < 30 seconds. Even lets say it took 30 secs, it is a less. When the other node is almost occupied(not completely occupied), we can spin up this new node.
@@AbhishekVeeramalla so you are saying if the node consumption is more than 70% spin up another node to save my time to spin up pod right😀
By the way I can't express my gratitude in words because even that would be less for the service that you provide, but even then THANK YOU 😃
Hi Abhishek & Jorge - I am yet to watch the vdo. Are we scaling pods or nodes ?. If we have huge surge in traffic, I dont think with Azure we can auto scale the nodes. We have to have PDBs for that or pre-scaling done.
Yes, we are scaling the nodes of the cluster. More specifically nodes of the node pool. AKS auto scales the nodes for you like I demonstrated on the video.
in AKS or EKS or GKE the control plane is managed by cloud provider correct? So what Jorge is demonstrating is we now can create a control plane in AKS or it was there from beginning and i was probably in dark till now😮💨
Hello can you make a long video zero to hero series combining all the videos
Will it work in production environment in e-commerce..?
Yes
Thanks for the response @@AbhishekVeeramalla
Hi, Abhishek i have a cluster autsclaer in aws eks and i have a 3 nodes. My application is running on these 3 nodes on off suddenly all my 3 nodes are down at same time. Since we have cluster scaling the nodes but when these are down and scaling the nodes scheduling the pods into new nodes it will take time so users will not access the application during the new nodes come up so how to achieve this
Bro, Please create a playlist for Openshift Certification DO280
Hi,Abhishek can you pls teach security context in K8's
Hi Abhishak Akz cluster update from 126.. to 127.. version getting fail is there any video regarding this issue how to solve theis issue?
Need to check Senthil. Not on this channel
Hi Abhishek I'm fan of u, want to talk to you once
Means a lot, thanks !!