Extend Disk on Red Hat CoreOS

Latest response

Guys, I have OCP 4.2 deployed but need to extend the disk size on my master and worker nodes. Is there a guide on how to do this and can it be done online?

Responses

Anyone?

Hi Darragh,

We're hoping to be able to test OCP sometime in the future where we work, and we've not used it. Since this is a supported product, I recommend submitting a new case with Red Hat and make it a priority. Please let them know the nature of your storage, I say this because from what I see at https://docs.openshift.com/container-platform/4.2/storage/expanding-persistent-volumes.html might or might not be something that describes your storage. Regardless, your storage is not a simple matter, which you probably certainly already know.

I'd be remiss to attempt to give you advice when I've only used OCP through online providers, and only now are we considering one for where we work. Red-Hatters do occasionally frequent the discussion forum, but since this is a supported product, and you seem to need help sooner than later, please do submit a case

Regards
RJ

Darragh,

I've asked if any others in the Red Hat Accelerator group (we are not Red-Hatters) might know of this as well.

Regards
RJ

What’s is your volume configuration? Everything on root? Are you needing to extend persistent volumes? Have you contacted support? Is this in a cloud provider? VMware? Bare metal?

Once you grow the underlying storage you can use the coreos-growpart script on whatever file system you need to.

/usr/libexec/coreos-growpart /sysroot https://access.redhat.com/solutions/4608041

Darragh,

Jay Ryan above is another member of the Red Hat Accelerator program. While we are not Red Hat employees, the Accelerator group has a lot of experts in our group with various specialties such as Jay. I generally deal with Linux and Satellite and sometimes other things. Please expand your details and see his post just above, even if you post here, and certainly for the priority case you've opened already with Red Hat.

Regards
RJ

All - As far as I know, due to the nature of CoreOS being immutable and encrypted, you can not extend the root volume after the OS has been deployed. That said, if you use AWS you should be able to create/replace nodes with the correct volume sizes during installation or scaleup (i.e., to replace a master or worker with a too small root volume) as is outlined in this document: https://github.com/openshift/installer/blob/master/CHANGELOG.md#fixed-7

NB: I have not tried this myself and would be interested in the results if someone here does attempt this.

We've OCP 4.2 with CoreOS (vsphere) and no problems to growing the FS (I do it). I extend the disk on the hypervisor and later on the node --> (sudo /usr/libexec/coreos-growpart /sysroot )