since numbers can also be in '1.e-10' format, we have to change
how we check for a number
Scalar::Util is already core and we use it in PVE::Tools, so
no new dependecy.
in case of "NaN" or "Infinity" we omit the key/value pair
else we quote like before
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
It makes more sense to have it there, especially since we want to
split out the service parts into a seperate file.
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Most other storage types do not have backup in their default content
selection when creating a new storage, so there it make sense to
initially have maxfiles disabled. But with CephFS we have default
_only_ backup as content type, so let it enabled here - as else one
must change something in the content selector to enable the field.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
While we can allow them via the apparmor profile, they still
won't be usable as the kernel doesn't have FS_USERNS_MOUNT
set on those.
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Reviewed-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Dominik Csapak <d.csapak@proxmox.com>
Pass the ring addresses from the node we use to join a cluster with
the encoded cluster info. Use it to determine if the nodes public
address (i.e., the onei t's node name resolves to) differs from it'S
ring0 address, if that's the case and the cluster was created with
PVE tooling then we know that it's highly probable that the ring0 is
on another network than the public net, so require the ring0_addrss
field in this case. If we have a false positive, it's not to bad
here.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
the package is not big, installed by default in the installer,
makes the pci scan much prettier, and we call lspci in the pvereport
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Allow to specify a separate cluster network when initializing ceph.
Ceph docs[0] imply a possibility for performance increase and
enhanced security in environments where the public network serves not
fully trusted peers, which could else provoke a DOS to the cluster
traffic[0].
Make this optional, but if passed `network` is required too.
[0]: http://docs.ceph.com/docs/luminous/rados/configuration/network-config-ref/
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Introduce the mdsCount again, I know remember again why I had it in
v3 of my CephFS series.. Use this to disable the CephFS create button
if we have no MDS configured, as this is a requirement.
Further change the gettext for 'No XY configure' to a format string
so that it can be reused.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Reviewed-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Dominik Csapak <d.csapak@proxmox.com>
as we depend on ceph-fuse elsewhere (pve-storage) this gets installed
from Debians repositories with the Ceph 10 version.
So ensure that an up to date version, from our current supported Ceph
release, gets installed when doing `pveceph install` else you may
fall into certain issues which would have been already resolved with
a newer version.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
a MDS gets transferred to the active MDS info once it got selected as
active, not once it really _is_ active, it can be there in the
'up:creating' state. So ensure that a MDS with 'u:active' could be
found
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
A MDS gets only active once a FS is there, and we need an MDS active
to be able to add a storage, as the CephFS plugin does an immediate
mount check. As an MDS needs some time to get active we had a
problematic time window where this mounting could fail.
Wait for a MDS to get in active state.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>