It depends on the configuration and settings you've set up for your MIG. The auto-scaling triggers based on different factors, not just the average CPU utilization but it is mostly based on Average utilization per node so that new nodes can be added/deleted based on the settings
#1. Under Auto Scaling set 1 for "Minimum Number of Instances" & "Maximum Number of Instances" #2. Under Auto Healing->create health check-> Give Name and other details with Protocol as HTTPS #3. While creating MIG-> Under Location-> select Multiple Zones option for Higher Availability
For 3rd, It talks about deployment so, feel free to check this one cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
Thanks for the video. Answers 1. Set min instances to 1 then set max to 1 from console under autoscaling 2. Add health checks 3. Adding proper cool down period. Let me know of this is correct.
Hello Divyendu, Amazing attempt to the questions. First one is absolutely correct, you nailed it! For second, adding health checks yes, please add other steps as well. For third: Questions says : deployment in MIG, that means you are updating the application inside MIG. hint: there is a type of deployment that you can select and adjust a couple of parameters :) Good luck!
@@TechTutorialswithPiyushThank you for taking out time to reply.. I will look again into the MIG. By the way i cleared Gcp associate cloud engineer certification yesterday, and stumbled on your video to understand MIG topic. Thanks again..
1 and 2 are correct, for the third one: feel free to check this one cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
Hello Sree, 1 and 2 are absolutely correct! Kudos to you. For 3rd, It talks about deployment so, feel free to check this one cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
Thank you for this video and It really helps @Piyush If Instances get crashed, what could happen? Select correct 2 options: (open challenge for everyone). I would really appreciate your answer. a). Automatically, the managed Instance group creates an Instance with a random name. b). The manage Instance group creates an instance with a new Instance template. c). Traffic will not be sent to the VM that went down d). Automatically, new Instance gets created with the Instance template that is supplied during the creation of a VM.
Thank you for the amazing feedback! Correct answer is C and D and here's why Why not A: MIG creates the instance but not with the random name. Why not B: MIG doesnt uses a new instance template rather it uses the same instance template Why C: When a VM crashes/stops inside a MIG, then the traffic is stopped on that VM as its health checks fails Why D: MIG uses the same instance template to provision a new VM as soon as it crashes.
Well Q.1 we can provision to set min instance to runnin all time and max also 1, any fail that time auto re-create. Q.2 supposed MIG VMs stop/crash that time traffic will not route. hence we able to check weather is healthy or not instance. Q.3 Suppose the team created a single zone template, and now that zone requires maintenance, we can construct another instance template to provision a new VM.
Thanks again for attempting the knowledge checks. 1) Absolutely correct, min=1 and max=1 2) health checks would be the right way 3) It talks about deployment so, feel free to check this one cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
@@TechTutorialswithPiyush Thank you for your feedback; I have learned a lot from you and my doubts have been cleared up as a result of your question example. I consider you fortunate for me.
Hello Carnaru, Yes, you can do it but why would you wanna do it? Better approach would be to deploy these apps seperately on a kubernetes cluster and you have all the power to seemlessly scale each of them independently. It will reduce your scaling time as well.
🔥 Check out our GCP playlist Namaste Google Cloud - Google Cloud Tutorial for beginners
th-cam.com/play/PLl4APkPHzsUXvkDFARdrH2LMESVqeuH8W.html
In order to add a new node in MIG, the AVG of all the pods should be >60% or if any node goes beyond 60% the auto-scaling will trigger ?
It depends on the configuration and settings you've set up for your MIG. The auto-scaling triggers based on different factors, not just the average CPU utilization but it is mostly based on Average utilization per node so that new nodes can be added/deleted based on the settings
#1. Under Auto Scaling set 1 for "Minimum Number of Instances" & "Maximum Number of Instances"
#2. Under Auto Healing->create health check-> Give Name and other details with Protocol as HTTPS
#3. While creating MIG-> Under Location-> select Multiple Zones option for Higher Availability
For 3rd, It talks about deployment so, feel free to check this one
cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
Thanks for the video. Answers
1. Set min instances to 1 then set max to 1 from console under autoscaling
2. Add health checks
3. Adding proper cool down period.
Let me know of this is correct.
Hello Divyendu, Amazing attempt to the questions. First one is absolutely correct, you nailed it!
For second, adding health checks yes, please add other steps as well.
For third: Questions says : deployment in MIG, that means you are updating the application inside MIG. hint: there is a type of deployment that you can select and adjust a couple of parameters :) Good luck!
@@TechTutorialswithPiyushThank you for taking out time to reply.. I will look again into the MIG. By the way i cleared Gcp associate cloud engineer certification yesterday, and stumbled on your video to understand MIG topic. Thanks again..
@@DivyenduJ That's amazing Divyendu! Give yourself a pat on the back :) Kudos to you
Do I really need a load balancer for my instance group? is it a must?
As MIGs are a group of identical instances which needs a frontend endpoint to balance the load so, yes, load balancer is required
1. min scaling should be 1 and maximum 1
2. create auto-healing
3. setting auto-scaling
Nice keep posting more on GCP love this
Thanks for your support Santosh! Glad you liked it
1.min and max instance 1
2.autohealing
Canary deployment
1 and 2 are correct, for the third one: feel free to check this one
cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
Knowledge Check:
1. min:1 max:1 and autoscaling should be turned ON
2. setup health check
3. make it regional deployment or multi zonal
Hello Sree, 1 and 2 are absolutely correct! Kudos to you.
For 3rd, It talks about deployment so, feel free to check this one
cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
@@TechTutorialswithPiyush so is the answer for 3rd question correct or are you expecting a different response piyush?
@@naskar23 Yes, sorry I missed your comment earlier. You will find the answer in the link I shared.
Tushi great ho Piyush ji
Thank you so much 😁
Thank you for this video and It really helps @Piyush
If Instances get crashed, what could happen? Select correct 2 options: (open challenge for everyone). I would really appreciate your answer.
a). Automatically, the managed Instance group creates an Instance with a random name.
b). The manage Instance group creates an instance with a new Instance template.
c). Traffic will not be sent to the VM that went down
d). Automatically, new Instance gets created with the Instance template that is supplied during the creation of a VM.
Thank you for the amazing feedback! Correct answer is C and D and here's why
Why not A: MIG creates the instance but not with the random name.
Why not B: MIG doesnt uses a new instance template rather it uses the same instance template
Why C: When a VM crashes/stops inside a MIG, then the traffic is stopped on that VM as its health checks fails
Why D: MIG uses the same instance template to provision a new VM as soon as it crashes.
@@TechTutorialswithPiyush Thanks for your valuable response🤝
You're welcome
Waiting 😊
Your wait is over sir!
Well Q.1 we can provision to set min instance to runnin all time and max also 1, any fail that time auto re-create.
Q.2 supposed MIG VMs stop/crash that time traffic will not route. hence we able to check weather is healthy or not instance.
Q.3 Suppose the team created a single zone template, and now that zone requires maintenance, we can construct another instance template to provision a new VM.
Thanks again for attempting the knowledge checks.
1) Absolutely correct, min=1 and max=1
2) health checks would be the right way
3) It talks about deployment so, feel free to check this one
cloud.google.com/compute/docs/instance-groups/rolling-out-updates-to-managed-instance-groups
@@TechTutorialswithPiyush Thank you for your feedback; I have learned a lot from you and my doubts have been cleared up as a result of your question example. I consider you fortunate for me.
@@KARADVAIBHAVPRAKASH Thank you so much ❤️ This feedback means a lot 😊 Keep learning
if I have a vm with nginx, mongo, node, angular this can be a template and do autoscale?
Hello Carnaru, Yes, you can do it but why would you wanna do it? Better approach would be to deploy these apps seperately on a kubernetes cluster and you have all the power to seemlessly scale each of them independently. It will reduce your scaling time as well.
@@TechTutorialswithPiyush Ok! Yes, your answer make sense, but I don't know very well k8s. Thank you for your great work!
Well you can try it out. It will make your life easier 😊
facing issues on Health check ...it is going unhealth ...need more help on firewall
Okay, join our discord community and let's try to fix your issue
p♥r♥o♥m♥o♥s♥m 😣