Hi Kasturi,
I see. What I was missing is that the engine does not have a nic on
glusternw.
I was hoping that the engine would be able to upload using the hosts.
Anyway, this is not an issue since I can directly scp on export domain as
you mentioned.
Thanx
On Oct 23, 2017 09:46, "Kasturi Narra"
Hi Alex,
Can you check if you have the following on your setup ?
1) gluster volume which will be used as ISO storage domain should have
bricks connected using glusternw.
2) NFS time out is caused due to not having nfs.disable off on the volume.
Can you try to set this and try again.
3) self hos
The only workaround that seems to work for my case is the following:
Enable NFS on ISO gluster volume.
Add ISO volume as NFS using hostname of server on the network that engine
has access.
This is not optimal as HA is not achieve - if I loose that server then I
will need to redefine the ISO domai
If this is the case then I will try to set the gluster network to be at the
network that engine can reach prior to adding the ISO domain. I would
prefere not to add an additional interface to engine as then I will need to
make the storage network a bridge which might incur some overhead and
affect
Hi,
The storage where 'ISO' storage domain resides has to be reachable for the
engine server. Please check network connectivity between the two.
On Wed, Oct 18, 2017 at 7:41 PM, Alex K wrote:
> Hi all,
>
> I have setup a self hosted setup with 3 servers with several volumes in
> replica 3.
> Th
Hi all,
I have setup a self hosted setup with 3 servers with several volumes in
replica 3.
These gluster volumes are set on a separate network (on separate NICs on
same servers), dedicated for gluster traffic. The engine does not have any
NIC in this network.
The ISO storage domain is also on top
6 matches
Mail list logo
