Resurrecting VMs with local volumes
Hello Bosh Team,
I have a question about an issue we are running into when deploying our application on bosh VMs.
This application needs to use local volumes for performance reasons and all volumes including the “root disk” are mounted on local disks.
We’re able run the application successfully.
However, when a VM does down (or we delete it to check DR), the bosh resurrector tries to bring it back up but this fails, because it is not able to mount back the local drives correctly.
Is VM resurrection is expected to work for VMs with local volumes, on bosh?
Is that supported? Is there a way to get this to work?
Hi Prajakta,toggle quoted messageShow quoted text
Root disk in BOSH semantics are typically ephemeral storage (as well as the extra ephemeral storage that can be mounted on /var/vcap/data), as opposed to the persistent storage that can be mounted on /var/vcap/store.
If your underlying IaaS infrastructure (you didn’t mention which one it is) is supporting persistent storage backed by local volumes, then check your BOSH CPI documentation to verify if this feature is supported.
You can find more information at https://bosh.io/docs/persistent-disks/