The instance group viewing page in the Google Cloud console should show the health check of the VMs. At the moment, there's no way to view the health status of VMs in an instance group from the console.
It should show the health status on this page:11 votes
Connections to the VM with OpenSSH only succeed if the correct username is given. Otherwise it returns "Permission denied (publickey)".
Some external tools, e.g. Ansible, use the SSH config file resp. the SSH agent.
Adding the "User" to each "Host" entry would be help here. Then it's also not required to add it somewhere in the tool config (if possible at all) and therefore these configs can easier be shared without user state in it.11 votes
Thanks, we’ll take a look.
it's such a hustle attaching a GPU using that command interface, spent over 8 hours on it today, only yo reach the end and the interface said not available in the trial, please help us get them in the browser and offer them on the free trial.11 votes
Thanks, these options will be available at a future date, these are just currently not available in the beta program.
I'd like to spin up a larger-than-micro instance for a cloud shell (micro appears to be too small to run the HBase shell). Would save me the trouble of installing maven, sdks, etc when I provision the instance.10 votes
Hi, thanks for the idea. Cloud Shell is meant to be a fast and easy way to quickly interact with a VM shell, and the free nature is connected to the fact that it’s a small VM you’re getting. For more elaborate work, you may want to consider configuring an Image in GCE just the way you want it, and then spinning up instances of that Image as necessary. With our 10-minute minimum billing duration for a VM, it would likely still be quite affordable.
My wish list is that you can also secure your instance's egress traffic by having the capability to have rules for specific target and ports.
If I want to protect my instance to disable internet access, I will configure (egress) to block port 80 on any target. Then I can also filter ports & IP address or IP Block for the egress traffic.10 votes
Thanks, we’re considering this in the future.
There is no way to install additional applications for a container-vm-image. So please provide docker-compose.10 votes
Thanks, we’ll share your idea with the team.
A method to forward e-mails coming to administrative users (like root) of a VM instance to project users
Since all outbound TCP connections to ports 25, 465, and 587 from a GCE instance are not allowed, there should be an easy method to forward e-mails coming to administrative users of a VM instance (like root, ubuntu, etc. in a Linux instance) to project users. Preventing sending out e-mails from an instance also makes life difficult for many other tools (monitoring etc.) that rely on e-mail communication to send alerts.10 votes
Thanks for the suggestion, we’re currently looking into enhancements in this area. We’ll update this item as information becomes available.
Use the same names for regions on the pricing page (https://cloud.google.com/compute/pricing) as you do when looking at instances.
Display the actual hourly pricing, not including discounts when you select or edit a product.10 votes
Thanks for your suggestion, we’ll look into this.
When transferring k8s cluster to another gcp project due to billing issues, we had to release and reserve the new static ip addresses. But the old ones we had whitelisted on several endpoints, updating which takes quite a while. It would be great if we could just transfer those ip addresses to another gcp project we own.10 votes
Thanks, we’ve shared this suggestion with the feature team so they may consider it for their backlog.
Currently there is no easy way to set username and password for Windows machine instances in the Console. Creation wizard only allows to set SSH keys for linux instances.
Password could not be easily set even programmatically https://github.com/GoogleCloudPlatform/compute-image-windows/issues/10310 votes
The cross-posted issue on our agent github project is updated to note that this is planned for a future release.
In [VM Instances] etc...
The quick access buttons [Reset] & [Delete] should be moved to the far right on the bar and/or coloured red rather than blue.
Reset is also a confusing name, should be Hard Reset or Image Reset...9 votes
Return labels for a VM instance when accessing it's metadata from inside the VM using the url http://metadata.google.internal/computeMetadata/v1/instance/labels similar to http://metadata.google.internal/computeMetadata/v1/instance/tags9 votes
Some VMs go up with "external-nat", others with "External NAT", making scripting a living hell and causing problems with VM cicling in a single static IP.9 votes
Would be helpful to allow me to give you an interval and you tell me how much network bandwidth is consumed during that time interval for a vm. My bill is charged by GB, not B/sec.9 votes
i have a special need with an instance running a server and it cant have any downtime, previously i had to increase its disk size for a eventuality but now isnt necesary that space.
I think that maybe its posible add this feature with a workload using a shrink disk and add the feature of delete the useless volume generated, or something that make possible small its size, like we can grow it in the console actually9 votes
Thanks for the request. We currently don’t support shrinking disk volumes currently, but it’s something we’re considering for the future.
This would be similar to the CPU utilization graphs that are associated with each vm instance.9 votes
Thanks, we’ll consider this in the future. Because our GPUs are a “pass through” operational mode, it’s not clear that GCE has access to the metrics that would be necessary to implement this.
Let me grant permission/authority over a single instance so I can provide these to my contractors without concerns about them touching other
Let me grant permission/authority over a single instance so I can provide these to my contractors without concerns about them touching other9 votes
We expect to allow finer grained controls in the future.
Today we can't see a historic data of the number of instances we had in a specific instance group.
I think a graph with the number of instances can be very useful9 votes
Autoscaler charts are now available in Beta. Please refer to:
Changing RAM/Core count requires to shutdown the machine, editing it and starting it again.
It would be nice if one could schedule a hardware upgrade so that all of this happens automatically at a chosen time and date. This would result in minimal downtime for non-redundant system.8 votes
- Don't see your idea?