I accidentally deleted my ai notebook vm and I hadn't downloaded the notebooks connected to it. I still have the url. Does anybody know if there's a way for me to recover my work?
According to the documentation, there is a life cycle for the instances. Verify the state of your AI Notebook VM to make sure that it is deleted or just turned off.
Unfortunately, if an AI Notebook instance is deleted and there is no snapshot configured, there is no way to restore that instance neither recover the notebooks stored there. There are three ways to prevent this from happening in the future:
Create snapshots to periodically or schedule back up data from your zonal persistent disks (snapshot can be located in multiple zones) or regional persistent disks (You must indicate the region where the disk is located ).
Edit VM instance, go to the deletion protection checkbox to enable it as this option is disabled by default. This setup will avoid that your Notebook instance was deleted by accident.
In the VM instance, go to boot disk, in the drop down list under “When deleting instance” select “Keep Disk” (or you can use gcloud command to disable set-disk-auto-delete)
Related
Hope all are safe and doing well.
I have few running servers on google cloud and for them, snapshots are scheduled on daily basis in an incremental way.
I am trying to create a new instance on a different VPC zone by using the same snapshots but it will be giving me an error.
For reference, I have added an attachment to this question.
Please help me to resolve this issue and thanks in advance.
Assuming that you have created a Snapshot with Application consistency(VSS) enabled:
When you create a VSS snapshot, Windows Server marks the volume in the
snapshot as read-only. Any disks that you create from the VSS
snapshot are also in read-only mode. So, the read-only flag on the new
boot disk prevents the VM instance from booting correctly.
You can follow this documentation to resolve your issue here.
If the disk you created from the VSS snapshot is a boot disk and you want to use it to boot a VM instance, you must temporarily attach the disk to a separate, existing VM instance. Once you complete the following steps, you can detach the disk from that existing VM instance and use it to boot a new VM instance.
When creating an AMI image from an existing EC2 instance, does it require the restart of the existing instance?
I make a copy of the instance, and subsequently the server went down because it turned off a process monitor which led to downtime, but I can't remember if it was because I rebooted the system (I can't remember if I rebooted it), or if it was because I made a copy of the image.
There's option to enable No reboot during create AMI
When creating an AMI image from an existing EC2 instance, does it require the restart of the existing instance
to answer this yes, when ami is being created aws the instance for ensuring data integrity.
doc says this "Amazon EC2 powers down the instance before creating the AMI to ensure that everything on the instance is stopped and in a consistent state during the creation process."
you can override this behaviour by enabling no reboot while creating ami.
No reboot – This option is not selected by default. Amazon EC2 shuts down the instance, takes snapshots of any attached volumes, creates and registers the AMI, and then reboots the instance. Select No reboot to avoid having your instance shut down.
refer 6 point of this https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/creating-an-ami-ebs.html.
also whenever you are unsure about anything just follow the docs of that service, it will be define in one way or the another.
I have a scheduled daily snapshot in GCP for one of my instances. I have several snapshots now. The first one is the full snapshot and the rest of the snapshots only contain changed data.
I want to be able to restore and boot the instance but it fails to boot. Checking the serial console I see reference to a blue screen and then it reboots and shows the same errors again, repeating the reboot cycle.
I have followed the guide in GCP on how to restore an instance from a snapshot by creating a new instance, selecting the snapshot tab and then selecting my snapshot. After saving the instance and trying to boot it I get the blue screen message.
Also, if I create a new instance and use a Windows 2008 R2 Datacenter image the system obviously boots fine but if I try to attach the snapshot disk as a secondary disk (non-boot) then I get the error: Editing VM instance failed. Error: Supplied fingerprint does not match current metadata fingerprint. I'm not sure if this is related to my issue with unable to boot the OS from my snapshot.
I did find a workaround:
1) create an image of running instance (my instance is Win'2008 R2 Datacenter)
2) enable scheduled snapshots of this new instance (with VSS)
3) wait for a scheduled snapshot to get created (hourly so must wait 1 hr)
4) create new instance from the scheduled snapshot
After all this work the instance boots just fine with all my data. Obviously not a very good workaround as now I have two instances with the same data. So I have to schedule the production system for maintenance so that I can bring it down and use the new instance so that future scheduled snapshots work if I try to restore it again. A major paint in the butt.
Anyone have any ideas as to why none of my instances boot from scheduled snapshots without my workaround? Manual snapshots work fine. And new instances also work fine with the same snapshot schedule.
I had this exact same problem. I tried multiple scheduled snapshots with the same result UNTIL I made a change in the VM Instance when attaching the restored snapshot. Maybe it's just Windows but if you named the disk something different then it seems to fail.
My original disk was called disk-1 for example. When I restored the snapshot I did it to disk-1-a and attached it to my instance. It failed the same way yours did. When I attach it and under "Device name" for the boot disk, select Custom and entered my original disk name of disk-1, it booted and RDP worked.
In case any VM instance gets deleted accidentally then is there any way to recover it in Dataproc cluster. In case there is no way to recover a deleted VM instance then can we create a new VM instance and connect to an existing DataProc cluster?
Please suggest.
There is no way to recover data on deleted VMs. Also manually removing workers carries a possibility of bricking the cluster. If your goal is to shrink the cluster we recommend option #1 below.
There's a few ways to add back a deleted VM:
Resize the cluster down and up using gcloud dataproc clusters update --num-workers, you'll want to ensure the deleted VM is in the set of VMs to be removed by downsize operation (this is not always possible)
From Developers Console go to Compute > Instances select any regular worker VM and clone it such that name matches the deleted VM.
Both of these approaches carry identical billing implications.
For some reason my instance group VM's keep getting reset back to the original image. i.e after I've installed and configured software everything gets whiped out. Additionally, in some occasions their IP's also change so I have to go and edit my Cloud SQl instance to allow network connections. Anyone seen this behavior before?
It sounds like you're using Managed Instance Groups, which are designed to work with stateless workloads. MIGs will scale their size up and down, if you have Autoscaler enabled, and scaling down will delete instances. The health checking feature can also destroy and recreate instances.
If you need extra software installed on MIG instances, you need to create a single VM the way you want, and then create a Snapshot of that VM's disk (and then an Image from the Snapshot). The Instance Template creates fresh instances from that Image file every time.
Even if you recreate your image the way you want with all software installed, MIGs will still create and destroy instances assuming there is nothing value on any of them. And yes, their IPs could change too, because new instances are being created.