Ceph Octopus
- Thread starter adamb
- Start date
Why? And first it needs to be released upstream.
https://docs.ceph.com/docs/master/releases/general/
Why? And first it needs to be released upstream.
https://docs.ceph.com/docs/master/releases/general/
Snapshot based mirror is the main reason why.
Looks like it hit the streets
https://ceph.io/releases/v15-2-0-octopus-released/
When do we think we might see it in proxmox?
https://ceph.io/releases/v15-2-0-octopus-released/
When do we think we might see it in proxmox?
Likely sometime after Proxmox VE 6.2.Looks like it hit the streets
https://ceph.io/releases/v15-2-0-octopus-released/
When do we think we might see it in proxmox?
Congratulations on the 6.2 release!
Still super excited about Ceph Octopus.
Let us know if there's any opportunity to test it, if we're willing to be guinea pigs =).
Still super excited about Ceph Octopus.
Let us know if there's any opportunity to test it, if we're willing to be guinea pigs =).
Is it possible to deploy the Ceph Dashboard as part of the Proxmox Ceph setup?
It's seen a lot of improvements and new features in the last few versions of Ceph, and would be great to get it out of the box.
Also - what are your thoughts on the new cephadm feature in Octopus?
Is it worth leveraging to use in Proxmox's Ceph deployment feature or not? It seems to do everything via containers, which could be advantageous in some ways.
It's seen a lot of improvements and new features in the last few versions of Ceph, and would be great to get it out of the box.
Also - what are your thoughts on the new cephadm feature in Octopus?
Is it worth leveraging to use in Proxmox's Ceph deployment feature or not? It seems to do everything via containers, which could be advantageous in some ways.
We build it but do not do in-depth test and do not support it officially.
Which ones do you think off exactly?
It's effectively what pveceph does for years.
I'd rather avoid that.
Well, who noticed?
Code:
pve-manager (6.2-6) pve; urgency=medium
* api: improve node index with missing/broken cert
* sdn: add VLAN-aware option to VNet edit dialog
* sdn: add mtu field to VLAN edit dialog
==> * initial support for upgrading to ceph octopus <==
* show if a ceph pool is using the PG autoscaler
* ui: improve filtering by VMID and search for the backup list view
-- Proxmox Support Team <support@proxmox.com> Wed, 03 Jun 2020 16:37:56 +0200I'm wondering if there is any chance we will be able to connect to external CEPH RDB and FS sources?
Currently connecting to an external Octopus cluster authenticates, but refuses to allow it to be used for actual storage. I have not tried that same configuration with a Nautilus cluster.
Looking back through the Proxmox Docs, it looks like at one time the RDB storage option offered both RDB and External RDB options. It owuld be really useful to be able to connect to native Ceph Octopus and use the native Ceph dashboard for managing a single Ceph cluster shared with both Proxmox and Kubernetes.
Alternately, I suppose I could just figure out how to modify the built in Cepf install to create an admin account that will allow for attaching Kubernetes storage directly to the built in Ceph back-end, but I'd prefer not to do that because I suspect it will create support issues long term.
Currently connecting to an external Octopus cluster authenticates, but refuses to allow it to be used for actual storage. I have not tried that same configuration with a Nautilus cluster.
Looking back through the Proxmox Docs, it looks like at one time the RDB storage option offered both RDB and External RDB options. It owuld be really useful to be able to connect to native Ceph Octopus and use the native Ceph dashboard for managing a single Ceph cluster shared with both Proxmox and Kubernetes.
Alternately, I suppose I could just figure out how to modify the built in Cepf install to create an admin account that will allow for attaching Kubernetes storage directly to the built in Ceph back-end, but I'd prefer not to do that because I suspect it will create support issues long term.
Can be done, needs just copying over a keyring for PVE to use for the added storage: https://pve.proxmox.com/pve-docs/chapter-pvesm.html#_authentication
I wish the instructions had worked. Those were the exact instructions I used when trying to connec to a navitive Ceph Octopus cluster.
Like I said above, I was able to authenticate successfully, but could not actually use any of the storage. Neither RDB or FS worked.
I'll take another shot at setting this up following those instructions again just to make sure I didn't miss something the first time around.
Last edited:
can you post your
pveversion -v output - also available under Node → Summary → Package Versions.Is there any way of testing this (http://download.proxmox.com/debian/ceph-octopus/dists/buster/test/binary-amd64/) yet? =)
@victorhooi you could try the Testing Repos: https://pve.proxmox.com/wiki/Package_Repositories#_proxmox_ve_ceph_testing_repository