Configure Autohealing for Managed Instance Groups

ACE By islever Editorial Last updated

Planning and implementing compute resources.

Short answer

C Create a managed instance group and set the Autohealing health check to healthy (HTTP) to enable automatic VM recreation upon failure detection.

Question

You want to configure autohealing for network load balancing for a group of Compute Engine instances that run in multiple zones, using the fewest possible steps.
You need to configure re-creation of VMs if they are unresponsive after 3 attempts of 10 seconds each. What should you do?
  1. Create an HTTP load balancer with a backend configuration that references an existing instance group. Set the health check to healthy (HTTP)
  2. Create an HTTP load balancer with a backend configuration that references an existing instance group. Define a balancing mode and set the maximum RPS to 10.
  3. Create a managed instance group. Set the Autohealing health check to healthy (HTTP) Correct Answer
  4. Create a managed instance group. Verify that the autoscaling setting is on.

Key Takeaways

The core trap is confusing load balancing health checks (which route traffic away from bad instances) with autohealing health checks (which actually delete and recreate unhealthy instances).

This question tests the distinction between load balancer health checks and managed instance group (MIG) autohealing in Google Cloud. The correct configuration requires setting up an HTTP health check within the MIG to trigger VM recreation, not just traffic routing.

Common mistake: Many learners select Option A because it mentions a health check, but they miss that load balancer health checks do not automatically recreate VMs; they only stop sending traffic to them.

Expert Analysis

Why the Answer Is Correct

Option C is the correct answer because autohealing is a feature of Managed Instance Groups (MIGs), not load balancers. To configure autohealing, you must attach a health check to the MIG. When the health check detects an instance as UNHEALTHY, the MIG controller deletes the instance and creates a new one based on the template. The prompt specifies '3 attempts of 10 seconds each,' which corresponds to configuring the health check's interval and threshold settings within the MIG context.

Why the Other Options Are Wrong

Option A describes configuring a load balancer backend. As noted by community members like DickDastardly and Fidget_, load balancer health checks 'help direct traffic away from non-responsive instances' but do not cause Compute Engine to recreate instances. Options B and D are incorrect because balancing modes (B) and autoscaling (D) do not inherently provide the autohealing mechanism required to replace failed VMs. Autoscaling handles capacity changes, while autohealing handles failure recovery.

Community Comment Notes

The community consensus strongly favors C. ReyBan highlighted that 'Health checks for autohealing detect and recreate failed instances,' distinguishing them from LB health checks. ashrafh confirmed that LB health checks only redirect traffic ('redirect to one which work') rather than recreating VMs. Fidget_ quoted official documentation stating that MIG health checks 'proactively signal to delete and recreate instances.' Tavva_prudhvi provided a step-by-step console guide confirming that autohealing is toggled ON within the instance group settings.

Official Reference

https://cloud.google.com/compute/docs/instance-groups#autohealing

This link opens the official vendor documentation in a new tab.

Exam Strategy

Always distinguish between 'traffic management' (Load Balancing) and 'infrastructure repair' (Autohealing/MIG). If the question asks to 'recreate' or 'repair' instances, look for Managed Instance Group configurations. If it asks to 'distribute' or 'route' traffic, look at Load Balancer backends.

Frequently Asked Questions

Why can't I use a Network Load Balancer health check for autohealing?

Network Load Balancer health checks only determine if an instance receives traffic. They do not trigger the deletion and recreation of the VM instance itself.

Does enabling autoscaling automatically heal broken VMs?

No. Autoscaling adjusts the number of instances based on load metrics. Autohealing is a separate setting within the MIG that monitors health and replaces unhealthy instances.

More ACE FAQ →

Discussion 128 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 30 of 128

  1. ReyBan 👍 98
    C, Agreed reference : https://cloud.google.com/compute/docs/tutorials/high-availability-autohealing Pro Tip: Use separate health checks for load balancing and for autohealing. Health checks for load balancing detect unresponsive instances and direct traffic away from them. Health checks for autohealing detect and recreate failed instances, so they should be less aggressive than load balancing health checks. Using the same health check for these services would remove the distinction between unresponsive instances and failed instances, causing unnecessary latency and unavailability for your users.
  2. ashrafh 👍 29
    I also vote for C go to gcp console create a httpa load balancer and in the health check settings take your mouse to question mark it says """Ensures that requests are sent only to instances that are up and running""" so its not recreating, if the vm not working it redirect to one which work. go to gpc console create MIG and check the questions mark of Autohealing health check settings it says """Autohealing allows recreating VM instances when needed. You can use a health check to recreate a VM instance if the health check finds it unresponsive. If you don't select a health check, Compute Engine will recreate VM instances only when they're not running.""" hope this help :)
  3. bryanchew 👍 23
    A. Create an HTTP load balancer with a backend configuration that references an existing instance group. Set the health check to healthy(HTTP) This is a possible answer. This answer assumes that the existing backend is configured correctly. B. Create an HTTP load balancer with a backend configuration that references an existing instance group. Define a balancing mode and set the maximum RPS to 10. This is a possible answer. This answer assumes that the existing backend is configured correctly. This answer adds an additional step over answer A. C. Create a managed instance group. Set the Autohealing health check to healthy(HTTP) This is only a partial solution. The default configuration is auto scaling enabled. You still need to create the HTTP Load Balancer. D. Create a managed instance group. Verify that the auto scaling setting is on. This is only a partial solution. Creating a Managed Instance Group with Auto Scaling is required, but you still need to create the HTTP Load Balancer. Therefore the best answer is A in my opinion.
  4. tavva_prudhvi 👍 8
    https://www.youtube.com/watch?v=dT7xDEtALPQ&list=PLIivdWyY5sqIij_cgINUHZDMnGjVx3rxi&index=36 step-1: go to the instance group step-2: click edit step-3: scroll down you will see auto-healing off by default change to ON step-4: create a health check saying 10 seconds as CHECK INTERVAL and UNHEALTHY THRESHOLD=3
  5. googlelearning 👍 7
    Surely it is C, auto-healing, as that is what it mentions in the question. It does not mention autoscaling.
Show 25 more comments
  1. DickDastardly 👍 6
    It can't be A as a load balancer does not re-create unhealthy instances, as per the requirement. Has to be C https://cloud.google.com/compute/docs/instance-groups
  2. dummyken 👍 5
    I took the exam just now. I can confirm I saw at least 7 questions from this dump. Do keep in mind that I only went up to page 10 so there might be more questions that appeared in both here and the exam.
  3. sadatus212 👍 5
    Hello everyone! Can anyone of you please confirm if the GCP-ACE exam dump questions do infact appear on the real exam. Thank you in advance....
  4. Fidget_ 👍 4
    https://cloud.google.com/compute/docs/instance-groups#autohealing "Load balancing health checks help direct traffic away from non-responsive instances and toward healthy instances; these health checks do not cause Compute Engine to recreate instances. On the other hand, managed instance group health checks proactively signal to delete and recreate instances that become UNHEALTHY"
  5. pYWORLD 👍 4
    I hope you saw that in the questions is stated: "network load balancing"
  6. Conjoinix 👍 3
    Correct Answer is C https://cloud.google.com/compute/docs/instance-groups#health_checking The health checks used to monitor MIGs are similar to the health checks used for load balancing, with some differences in behavior. Load balancing health checks help direct traffic away from non-responsive instances and toward healthy instances; these health checks do not cause Compute Engine to recreate instances. On the other hand, managed instance group health checks proactively signal to delete and recreate instances that become UNHEALTHY.
  7. yas_cloud 👍 3
    C is the right answer. Autohealing allows recreating VM instances when needed. If you don't select a health check, Compute Engine will recreate VM instances only when they're not running. In this case, we need to re-create VMs for a specific health criteria of the HTTP.
  8. Ridhanya 👍 3
    it cannot be option A because as you said, load balancer with the health check is already present and now the problem is simply auto healing. so we need to focus only on recreation which can happen only if option C is correct
  9. Partha117 👍 3 Answered C
    That is the only option with autohealing which will recreate instance, just doing health check using http load balancer will not help to recreate, that can only redirect traffic
  10. ShakthiGCP 👍 3
    I was wrong. A & B are wrong because it should by NetworkLoad Balancer and not HTTP Load balancer. Ans: C is correct .
  11. Raz0r 👍 3 Answered C
    C, Agreed reference : https://cloud.google.com/compute/docs/tutorials/high-availability-autohealing
  12. manu202020 👍 3
    D. First statement says "what you want". Second statement actually tells about the action required.
  13. adedj99 👍 3
    Then answer is A . https://www.youtube.com/watch?v=FtLhVvSFi84
  14. SaiSaiA 👍 2
    C is the only one with the AUTOHEALING option, but it is not really correct. Remember, the GIVEN information are "a NETWORK load balancer and a group of Compute Engine Instances that run in multiple zones" which gives us an idea that the existing configuration is a target pool-based network lb. If we are to use the existing group of VMs, we need to choose UNMANAGED Instance Group, UNMANAGED Instance Group does not have Autohealing, only a health check. Health check only checks if VMs are responsive or not but does not re-create instances as what Autohealing and Autoscaling do. You can also try re-creating the scenario or check this https://cloud.google.com/load-balancing/docs/network/transition-to-backend-services#console So, if a MANAGED INSTANCE group is to be used, then you need to create an instance template and use it for your MIG. Ofc, you cannot use the existing VMs, then you create a new load balancer. Ofc, the existing group of VMs mentioned in the question will no longer be used but rather a new set of VMs based on the instance template will be created. The choices should be updated.
  15. gcper 👍 2
    I'd like to add that this question is about Network Load balancing Autohealing. Network load balancers require legacy health checks. Quote: "Network Load Balancing requires legacy health checks that use the HTTP protocol." source: https://cloud.google.com/load-balancing/docs/health-check-concepts Second, the answers that want you to "set a check as healthy" don't make sense, considering that there is no such option! If you look at the gcloud or console, you will not find such parameters. The closest thing to such a parameter is the HEALTHY_THRESHOLD. source: https://cloud.google.com/load-balancing/docs/health-checks#gcloud_5 C and D don't create a load balancer and the question is about network load balancer health checks so those are ruled out. A and B create HTTP load balancers but network load balancers only support TCP/UDP traffic, so neither of them would work either. Layer 7 vs layer 4. source: https://cloud.google.com/load-balancing/docs/network Is this a trick question?
  16. DickDastardly 👍 2
    From https://cloud.google.com/compute/docs/instance-groups Health checking The health checks used to monitor MIGs are similar to the health checks used for load balancing, with some differences in behavior. Load balancing health checks help direct traffic away from non-responsive instances and toward healthy instances; these health checks do not cause Compute Engine to recreate instances. On the other hand, managed instance group health checks proactively signal to delete and recreate instances that become UNHEALTHY Health checks for load balancers do NOT re-create the instance, as per the requirement, they merely direct traffic away from the sick instance. A does not solve this requirement therefore it has to be C
  17. JohnnyS20 👍 2
    Feels like most people are trapped by option C because it has the keyword 'autoheal' But I have to side to the few who chose A. You really have to read the question slowly and analyze what the requirement is and what are the given. After going over the many points raised in this discussion, I'm inclined to eliminate C and D. This is because while C and D shows minimal steps, they are still missing the essential part which is the load balancer. The question also states that there's already a set of instances that are running on different zones. From there it clearly explains that there's no need to create new managed instance groups.
  18. Ramyak 👍 2
    Correct Answer is C. Autohealing should be at instance group level not at LB level. Check the explanation below. Use separate health checks for load balancing and for autohealing. Health checks for load balancing can and should be more aggressive because these health checks determine whether a VM receives user traffic. You want to catch non-responsive VMs quickly, so you can redirect traffic if necessary. In contrast, health checking for autohealing causes Compute Engine to proactively replace failing VMs, so this health check should be more conservative than a load balancing health check.
  19. Buruguduystunstugudunstuy 👍 2 Answered C
    Option C is correct because creating a managed instance group allows you to use autohealing to automatically recreate VMs that are unresponsive after 3 attempts of 10 seconds each. You can set the Autohealing health check to healthy (HTTP) to specify the health check that determines whether the instances are considered healthy or not. If an instance becomes unresponsive, Autohealing will recreate the instance and attach it to the managed instance group. https://cloud.google.com/compute/docs/instance-groups/autohealing-instance-groups
  20. JohnnyS20 👍 2
    Now, between A and B, A has the fewest steps. However, it should still meet the requirements. The only questionable part about the health check is that it is much accurate if there is a separate health check for http and a separate health check for autoheal. Setting aside, A still meets the requirement. Remember that the question is if the option meets the requirement, regardless if there are better ways to solve it. We must stick with the options that are given.
  21. evangelist 👍 2
    The answer is D. this is because by being ambiguous enough. C implies you just set the health check protocol being HTTP and nothing else as an action taken. but D "verify that the autoscaling setting is on" implies a multi-steps actions including setting the health check protocol to HTTP, set the interval to be 10 and attempts to be 3 and so on. I am shocked GCP test is being such a play of word games, but this is it.
  22. tavva_prudhvi 👍 2
    https://www.youtube.com/watch?v=dT7xDEtALPQ&list=PLIivdWyY5sqIij_cgINUHZDMnGjVx3rxi&index=36 I request everyone to watch this video, as in the question we have a few conditions like, the re-creation of VMs if they are unresponsive after 3 attempts of 10 seconds each. we can configure in the instanced group by setting the health check on!!!! I'm sorry if I misled anyone the correct answer is C.
  23. GCP_Student1 👍 2
    You want to configure autohealing for (network load balancing) for a (group of Compute Engine instances) that run in multiple zones, using the fewest possible steps. You need to (configure re-creation) of VMs if they are unresponsive after 3 attempts of 10 seconds each. What should you do? Pay attention to the phrases in parenthesis in the question, it will make it easy to choose the answer.
  24. rambest 👍 2
    It is C for sure, if we see the question carefully, there are 1 trigger point is there "You need to configure re-creation of VMs if they are unresponsive after 3 attempts of 10 seconds each." This can be done only at create health check section of auto healing.
  25. jabrrJ68w02ond1 👍 2
    It must be C. Answer A indeed does a health check, but a failing LB health check only results in the unhealthy machine not getting any traffic. We want it to be auto-healed (restarted), so we must use a MIG + health check --> C

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

Single Instance Managed Instance Group Configuration on GCP

S-grade

Authenticating On-Premises Apps to GCP AutoML APIs

S-grade

Consolidating Multi-Project Monitoring in One Stackdriver Dashboard

S-grade

How to Enable Cloud Pub/Sub API for App Engine

S-grade

Practice All ACE Questions

Access 84 questions with complete answers and detailed explanations.

View Full ACE Practice Test →

← Back to ACE Study Guide