[Gluster-users] troubles installing Debian8 as qemu-kvm guest on glusterfs 3.5.4 HA storage volume, need help please!

[Roman]

Ah so many information to share and I forgot one more thing:
if I create the VM on Distributed volume and then convert it to template
and clone it to HA volume, things seem to be working fine also.
So there is something wrong during installation process only.


p.s.
I did iperf tests with all of my proxmox hosts - no losses and they are
connected over 1gbps network. Storage network for gluster uses separate
network with 1 gbps cards also.

2015-06-12 0:38 GMT+03:00 Roman <romeo.r at gmail.com>:

> Hi,
>
> The debian 8.1 is released, but I've got still problems installing it as
> qemu-kvm guest on glusterfs Replicated storage volume on all of my proxmox
> servers. If I chose to raw disk image for virtual HDD, the installation
> just takes ages. If I chose to qcow2 format, it stops the installation on
> random step (most often is on choosing repository for installation. No
> matter which I chose it says its not able to connect to it).
>
>  If I wait long enough for installation to complete using raw file as
> virtual storage and then boot the VM, everything is very-very slow.
>
> If I create a template of this VM and then clone it, using qcow2 storage
> type, the VM works faster, but it seems like every single binary is
> corrupted (not able to install even apache and run some sh commands).
>
> Using latest proxmox
> pve-manager/3.4-6/102d4547
> And just upgraded my glusterfs to 3.5.4 hoping for fixes :(
>
>
> here is my volume info:
>
> Volume Name: HA-1TB-S14A4F-pve
> Type: Replicate
> Volume ID: 8e01b955-f707-4cbf-812f-04d48b7f374f
> Status: Started
> Number of Bricks: 1 x 2 = 2
> Transport-type: tcp
> Bricks:
> Brick1: stor1:/exports/S14A4F/1TB
> Brick2: stor2:/exports/S14A4F/1TB
> Options Reconfigured:
> server.allow-insecure: on
> network.ping-timeout: 3
>
> and
>
> root at stor1:~# cat /etc/glusterfs/glusterd.vol
> volume management
>     type mgmt/glusterd
>     option working-directory /var/lib/glusterd
>     option transport-type socket,rdma
>     option transport.socket.keepalive-time 10
>     option transport.socket.keepalive-interval 2
>     option transport.socket.read-fail-log off
>     option rpc-auth-allow-insecure on
> end-volume
>
> there is nothing usefull in log files.
>
> Please consider installing an instance with proxmox and reproduce this.
> This happens ONLY to debian 8, every other OS installs and runs just fine.
>
> If I place the virtual disk on Distributed glusterfs volume - there are no
> problems with installation nor running.
>
> And this volume configured just the same way
>
> Volume Name: DISTR-4TB-S14A4F-pve
> Type: Distribute
> Volume ID: 21fa0ee6-d70a-424a-b681-008ce5156c91
> Status: Started
> Number of Bricks: 2
> Transport-type: tcp
> Bricks:
> Brick1: stor1:/exports/S14A4F-D/2TB
> Brick2: stor2:/exports/S14A4F-D/2TB
> Options Reconfigured:
> network.ping-timeout: 3
> server.allow-insecure: on
>
> If I install on local storage everything works fine also.
>
>
> --
> Best regards,
> Roman.
>



-- 
Best regards,
Roman.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20150612/3d880c9b/attachment.html>