Compute Engine Error: Not enough resources available to fulfill the request

[Dave Greenly]

I have recently started getting errors when my compute engines are starting.  I have copy and pasted the entire error below, but it is: Not enough resources available to fulfill the request.

This is very problematic for our environment because we have automated processing that happens, and we expect that we are going to be able to start a compute engine when we request one be started. The additional problem is that the compute engine is then put into a state where it is not running, but it shows up on the developers console and it has a grey circle with a white square inside of it as the icon.  Then when we try to start the compute engine again, it does not start because "it is already started".  

As a result of this, we have customers that are not receiving some of the data that we are advertising they will receive automatically and it is a huge strain on us to continually monitor this situation.

Questions:

1) Is there something I can do to guarantee the resources are available for our use when we need them?  I thought this was the nature of the environment, start up what you need when you need it. How does this apply to the SLA agreement?  We are paying money for these services.

2) What is the state of a VM that is grey and has the white square inside it for the icon and is there a way to delete the VM automatically if it does not start correctly with an appropriate error message?

3) We cannot simply "try again" in another zone, because the disks that we are using live in the same zone that we are starting the compute engine in. Is there a way to set things up to automatically "try another zone" using the disk that we currently are trying to boot from?

Thanks

Dave

 "ERROR: (gcloud.compute.instances.create) Some requests did not succeed: - The zone 'projects/sn-auto-data-mgmt/zones/us-central1-a' does not have enough resources available to fulfill the request. Try a different zone, or try again later.

[Alex Martelli]

On Mon, Jun 8, 2015 at 11:06 AM, Dave Greenly <dgree...@gmail.com> wrote:

I have recently started getting errors when my compute engines are starting.  I have copy and pasted the entire error below, but it is: Not enough resources available to fulfill the request.

This is very problematic for our environment because we have automated processing that happens, and we expect that we are going to be able to start a compute engine when we request one be started. The additional problem is that the compute engine is then put into a state where it is not running, but it shows up on the developers console and it has a grey circle with a white square inside of it as the icon.  Then when we try to start the compute engine again, it does not start because "it is already started".  

As a result of this, we have customers that are not receiving some of the data that we are advertising they will receive automatically and it is a huge strain on us to continually monitor this situation.

Questions:

1) Is there something I can do to guarantee the resources are available for our use when we need them?  I thought this was the nature of the environment, start up what you need when you need it. How does this apply to the SLA agreement?  We are paying money for these services.

I am not a lawyer, but what I read at https://cloud.google.com/compute/sla specifically refers to "all running Instances that are hosted across two or more zones" -- consistent with the quote from https://cloud.google.com/compute/docs/zones that "Google recommends deploying applications across multiple zones in a region" (for high availability). Therefore, if you have instead chosen to constrain your application to a single zone, I don't see how the SLA applies; please enlighten me if I'm missing something, of course!

 

2) What is the state of a VM that is grey and has the white square inside it for the icon and is there a way to delete the VM automatically if it does not start correctly with an appropriate error message?

Sorry, I'm not familiar with the UI. To automatically delete instances in an unhealthy state, I personally would focus on a little monitor performing health checks and reacting to their consistent failure (since I don't believe shutdown scripts can run on an instance that was never successfully created and booted up).

 

3) We cannot simply "try again" in another zone, because the disks that we are using live in the same zone that we are starting the compute engine in. Is there a way to set things up to automatically "try another zone" using the disk that we currently are trying to boot from?

Per https://cloud.google.com/compute/docs/disks/persistent-disks#migratingdiskdata, "To migrate your disk data to another zone, use a persistent disk snapshot instead." -- which links to https://cloud.google.com/compute/docs/disks/persistent-disks#snapshots for details about persistent disk snapshots.

Alex

 

Thanks

Dave

 "ERROR: (gcloud.compute.instances.create) Some requests did not succeed: - The zone 'projects/sn-auto-data-mgmt/zones/us-central1-a' does not have enough resources available to fulfill the request. Try a different zone, or try again later.

--
© 2014 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 
Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-dis...@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.
---
You received this message because you are subscribed to the Google Groups "gce-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussio...@googlegroups.com.
To post to this group, send email to gce-dis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/fdce407b-c3b2-4e04-906c-7fcedfa3bddb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Dave Greenly]

Thanks for all of the info Alex!

This is a great discussion, as I am not a lawyer either, :), I just need to understand what I can do to get the highest level of availability that I can.  From some email support and documentation I saw back when I first started on this task, I thought the reliability came from automatic fail-over etc...?  So from this error, and what I am hearing, it is possible that a particular zone will not have enough resources for the amount of users in it? (I just started seeing this error for the first time in the last couple days)

What I am not understanding at this point is how to "host my application across multiple zones". 

I have a single java program that runs when the compute engine is started.

The compute engine is started from a bash shell that uses the "gcloud compute instance create" command. This command requires a zone identified for where it will run.

So from that perspective, I did not see anything that shows how I would enable it across multiple zones.

And I have migrated the disk to the other zones, but, the same question applies, when I migrate the disk, it is migrating it to a specific zone, not across multiple zones.

Thanks

Dave

[Alex Martelli]

I would start with the general background information at https://cloud.google.com/compute/docs/zones . Distributing across zones (whether in the same region or different regions) is recommended to increase availability. Some resources (instances and disks) are zone-specific, others are not (either region-wide or global). 

Then you can continue with https://cloud.google.com/compute/docs/tutorials/robustsystems and will be advised to distribute instances across zones, use load-balancing, and so on.

To create instances across multiple zones (usually in the same region, so that communication among them is fastest) your script needs to use what amounts to a simple equivalent of load balancing -- but it's simple enough that you could e.g just pick a random one among the available zones of interest (and segue with health checks after a while to make sure the instance was correctly built and started, else delete it and try again).

So when you create any specific instance it will have a specific zone (and persistent disks it needs, if not already available there, will need to be snapshotted and migrated there) but over time your load-balanced instances will be evenly spread among the zones of interest and your system's availability will be maximized.

Alex

[John Robinson]

It might be worth checking your quotas. OP was 9 months ago, so you're probably past this, but quotas aren't well publicized or explained. Check the "Quotas" section in the Compute Engine menu. I requested a quota increase today and it was served within a minute, though YMMV. I'm not positive, but I'm pretty sure this issue is down to a quota problem, combined with an expectation on Google's part that you'll be spreading your systems across zones in a region somewhat evenly.

[Dave Greenly]

This was not a quota problem.  We updated the code in our Go App that starts the VMs, to check to be sure that the VM starts correctly.  If it does not, then we try to start it in a different zone.  

This has pretty much resolved the issue now.

Thanks 

Dave

--

© 2014 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 
Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-dis...@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.
---
You received this message because you are subscribed to the Google Groups "gce-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussio...@googlegroups.com.
To post to this group, send email to gce-dis...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/0f61ee0e-66a1-425a-a81d-a518ac5e101d%40googlegroups.com.

[Rohit kumar]

From the last 4 days this instance is not starting and getting error.

Starting VM instance "instance-2" failed. Error: The zone 'projects/seraphic-result-221707/zones/europe-west3-b' does not have enough resources available to fulfill the request. Try a different zone, or try again later.

[adrian.f...@gmail.com]

Same error for few days, and this is  big problem ...

понедельник, 8 июня 2015 г., 21:06:20 UTC+3 пользователь Dave Greenly написал:

[Veera (Google Cloud Support)]

As per my findings the error message “'Try a different location, or try again later: Google Compute Engine does not have enough resources available to fulfill request: europe-west3-b” indicates that resources to fulfill your request is temporarily unavailable at that specific zone. That being said, be assured that our Product team is working hard to resolve it. Our goal is to make sure that there are available resources in all zones. This type of issue is rare. When a situation like this occurs or is about to occur, our Product team is notified immediately and the issue is investigated. 

I would recommend deploying and balancing your workload across multiple zones or regions to reduce the likelihood of an outage. Please review our documentation [1] which outlines how to build resilient and scalable architectures on Google Cloud Platform. 

Again, I understand that it's a difficult situation, but at the moment you could either try a different zone or try again later and rest assured that we are working hard to resolve this and make this an exceptionally rare event. 

[1]: https://cloud.google.com/solutions/scalable-and-resilient-apps

On Tuesday, January 29, 2019 at 5:50:11 PM UTC-5, 

[arthur.k...@tremend.com]

It's hard to believe that after all this time the problem persists. But it does. My region is europe-west3 and both "a" and "b" zones are still unavailable after 1 week. I provisioned all resources with Terraform, always taking into consideration the HA factor. With 2 out of 3 zones unavailable, how can I provide HA for the platform I'm trying to deploy?

[Sunny J (Google Cloud Support)]

Hi, 

Unfortunately, there's still not enough resources in this zone to fulfill your request. 

Please note that, Google continuously monitor zone/region resource utilization and add new resources as per our internal processes and guidelines. However, I'm unable to provide you an ETA. when resources may be added or freed in the affected zone. Hence, as a workaround you can try balancing your workload on a different zone/region. You may also see the following discussion thread for a detailed explanation by (Paul) who's one our Product Manager for Google Compute Engine for a similar incident.

Also could you clarify what you mean by "how can I provide HA for the platform I'm trying to deploy?" 

[kavirajan t]

The zone 'projects/kg-maps-271709/zones/asia-south1-a' does not have enough resources available to fulfill the request. Try a different zone, or try again later.

Hi this is happening while i try to start instance

unfortunately we dont have file backup...aleast we need backup file eventhough the instance cant be started

Please help google team

[Naman Parekh]

Backing up your data with a snapshot schedule is an effective best practice to manage workloads and minimize the effects of data loss. Restoring the PD snapshot and create a VM would be one of the good practice for the any data loss or overcome VM issues.

[1] https://cloud.google.com/compute/docs/disks/restore-and-delete-snapshots