This PR https://github.com/ceph/ceph/pull/40305 should fix the issue.
Is this section in the downstream docs at all? I'm wondering if we should close this as UPSTREAM if it's not applicable to the current downstream docs.
(In reply to Ken Dreyer (Red Hat) from comment #3) > Is this section in the downstream docs at all? I don't think so. > I'm wondering if we should > close this as UPSTREAM if it's not applicable to the current downstream docs. I think there are no downstream docs yet. Should we convert this into a downstream docs bz?
I will change the summary of this BZ and say " Restrict creation of multiple nfs clusters on a single node". As it allows to create multiple clusters. Please restrict creation of multiple nfs cluster on a single port.
Sure, we can move this to 5.1 I have made sure the doc has a note saying "Only one NFS cluster can be deployed on a node."
Hemanth, I'm confused. Do you want to b) deploy multiple NFS daemons belonging to the same NFS cluster on a single host a) deploy multiple NFS clusters on the same set of hosts Why do you want to do a)? For HA you would want deploy NFS daemons on different hosts. In the bug description you're trying to deploy 4 NFS daemons on the same 2 hosts. Why? b) should be possible right now if NFS daemons of different clusters listen on different ports. The commit to make NFS daemons listen on non-standard port is available in v16.2.7. Is anyone requesting the feature of running multiple NFS clusters to deployed on the same set of hosts? Note that we don't have native HA support for a single NFS cluster if one of the NFS daemon hosts goes down, and cephadm is the orchestrator used. See, https://tracker.ceph.com/issues/47038
(In reply to Ram Raja from comment #15) > Hemanth, I'm confused. Do you want to > > b) deploy multiple NFS daemons belonging to the same NFS cluster on a single > host > a) deploy multiple NFS clusters on the same set of hosts I mean to write, do you want to a) deploy multiple NFS daemons belonging to the same NFS cluster on a single host (or) b) deploy multiple NFS clusters on the same set of hosts > > Why do you want to do a)? For HA you would want deploy NFS daemons on > different hosts. > In the bug description you're trying to deploy 4 NFS daemons on the same 2 > hosts. Why? > > b) should be possible right now if NFS daemons of different clusters listen > on different ports. The commit to make NFS daemons listen on non-standard > port is available in v16.2.7. > You can try setting the --port argument with `ceph nfs cluster create` command. https://docs.ceph.com/en/pacific/mgr/nfs/#create-nfs-ganesha-cluster > > Is anyone requesting the feature of running multiple NFS clusters to > deployed on the same set of hosts? Note that we don't have native HA support > for a single NFS cluster if one of the NFS daemon hosts goes down, and > cephadm is the orchestrator used. See, https://tracker.ceph.com/issues/47038
This isn't a blocker on a technical level, but based on the risk of the change and the reduction in user confusion (as evidenced by the discussion here with Hemanth) I am marking it as such so so we can merge for 5.2.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: Red Hat Ceph Storage Security, Bug Fix, and Enhancement Update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2022:5997