
Still, now I cannot start none of the 2 machines! I get ID 119 VM proxy2 is down. Exit message: Child quit during startup handshake: Input/output error."" Something similar to bug https://bugzilla.redhat.com/show_bug.cgi?id=1033064, except that in my case selinux is permissive! On Wed, Dec 18, 2013 at 2:10 PM, Gabi C <gabicr@gmail.com> wrote:
in my case $brick_path =/data
getfattr -d /data return NOTHING on both nodes!!!
On Wed, Dec 18, 2013 at 1:46 PM, Fabian Deutsch <fabiand@redhat.com>wrote:
Am Mittwoch, den 18.12.2013, 13:26 +0200 schrieb Gabi C:
Update on Glusterfs issue
I manage to recover lost volume after recretaing the same volume name with same bricks, whisch raised an error message, resolved by, on both nodes:
setfattr -x trusted.glusterfs.volume-id $brick_path setfattr -x trusted.gfid $brick_path
Hey,
good that you could recover them.
Could you please provide $brick_path and getfattr -d $brick_path
The question is if and/or why the fattrs are not stored.
- fabian
On Wed, Dec 18, 2013 at 12:12 PM, Gabi C <gabicr@gmail.com> wrote: node 1:
[root@virtual5 admin]# cat /config/files /etc/fstab /etc/shadow /etc/default/ovirt /etc/ssh/ssh_host_key /etc/ssh/ssh_host_key.pub /etc/ssh/ssh_host_dsa_key /etc/ssh/ssh_host_dsa_key.pub /etc/ssh/ssh_host_rsa_key /etc/ssh/ssh_host_rsa_key.pub /etc/rsyslog.conf /etc/libvirt/libvirtd.conf /etc/libvirt/passwd.db /etc/passwd /etc/sysconfig/network /etc/collectd.conf /etc/libvirt/qemu/networks /etc/ssh/sshd_config /etc/pki /etc/logrotate.d/ovirt-node /var/lib/random-seed /etc/iscsi/initiatorname.iscsi /etc/libvirt/qemu.conf /etc/sysconfig/libvirtd /etc/logrotate.d/libvirtd /etc/multipath.conf /etc/hosts /etc/sysconfig/network-scripts/ifcfg-enp3s0 /etc/sysconfig/network-scripts/ifcfg-lo /etc/ntp.conf /etc/shadow /etc/vdsm-reg/vdsm-reg.conf /etc/shadow /etc/shadow /etc/sysconfig/network-scripts/ifcfg-ovirtmgmt /etc/sysconfig/network-scripts/route-ovirtmgmt /etc/sysconfig/network-scripts/rule-ovirtmgmt /root/.ssh/authorized_keys /etc/vdsm/vdsm.id /etc/udev/rules.d/12-ovirt-iosched.rules /etc/vdsm/vdsm.conf /etc/sysconfig/iptables /etc/resolv.conf /etc/sysconfig/network-scripts/ifcfg-VPO_IPPROXY /etc/sysconfig/network-scripts/ifcfg-enp6s0 /etc/sysconfig/network-scripts/ifcfg-enp6s0.50 /etc/glusterfs/glusterd.vol /etc/selinux/config
node 2:
[root@virtual4 ~]# cat /config/files /etc/fstab /etc/shadow /etc/default/ovirt /etc/ssh/ssh_host_key /etc/ssh/ssh_host_key.pub /etc/ssh/ssh_host_dsa_key /etc/ssh/ssh_host_dsa_key.pub /etc/ssh/ssh_host_rsa_key /etc/ssh/ssh_host_rsa_key.pub /etc/rsyslog.conf /etc/libvirt/libvirtd.conf /etc/libvirt/passwd.db /etc/passwd /etc/sysconfig/network /etc/collectd.conf /etc/libvirt/qemu/networks /etc/ssh/sshd_config /etc/pki /etc/logrotate.d/ovirt-node /var/lib/random-seed /etc/iscsi/initiatorname.iscsi /etc/libvirt/qemu.conf /etc/sysconfig/libvirtd /etc/logrotate.d/libvirtd /etc/multipath.conf /etc/hosts /etc/sysconfig/network-scripts/ifcfg-enp3s0 /etc/sysconfig/network-scripts/ifcfg-lo /etc/shadow /etc/shadow /etc/vdsm-reg/vdsm-reg.conf /etc/sysconfig/network-scripts/ifcfg-ovirtmgmt /etc/sysconfig/network-scripts/route-ovirtmgmt /etc/sysconfig/network-scripts/rule-ovirtmgmt /root/.ssh/authorized_keys /etc/shadow /etc/shadow /etc/vdsm/vdsm.id /etc/udev/rules.d/12-ovirt-iosched.rules /etc/sysconfig/iptables /etc/vdsm/vdsm.conf /etc/shadow /etc/resolv.conf /etc/ntp.conf /etc/sysconfig/network-scripts/ifcfg-VPO_IPPROXY /etc/sysconfig/network-scripts/ifcfg-enp6s0 /etc/sysconfig/network-scripts/ifcfg-enp6s0.50 /etc/glusterfs/glusterd.vol /etc/selinux/config
On Wed, Dec 18, 2013 at 12:07 PM, Fabian Deutsch <fabiand@redhat.com> wrote: Am Mittwoch, den 18.12.2013, 12:03 +0200 schrieb Gabi C: > So here it is: > > > in tab volumes add new volume - Replicated, then added storage - > data/glusterfs. Then I impoerted Vm, ran them and at some point, > needing some space for a Redhat Satellite instance I decided to put > both node in maintenace stop them add new disk devices and restart, > but after restart the gluster volume defined under Volumes Tab > vanished!
Antoni,
can you tell what log files to look at to find out why that storage domain vanished - from a Engine side?
And do you know what files related to gluster are changed on the Node side?
Gabi,
could you please provide the contents of /config/files on the Node.
> Glusterfs data goes under /data directory which was automatically > configured when I installed the node.
Yep, /data is on the Data LV - that should be good.
- fabian
> > > On Wed, Dec 18, 2013 at 11:45 AM, Fabian Deutsch <fabiand@redhat.com> > wrote: > Am Mittwoch, den 18.12.2013, 11:42 +0200 schrieb Gabi C: > > Yes, it is the VM part..I just run into an issue. My setup > consist in > > 2 nodes with glusterfs and after adding supplemental hard > disk, after > > reboot I've lost glusterfs volumes! > > > Could you exactly explain what you configured? > > > > > How can I persist any configuration on node and I refer here > to > > ''setenforce 0'' - for ssh login to work- and further > > > How changes can be persisted on Node can be found here: >
http://www.ovirt.org/Node_Troubleshooting#Making_changes_on_the_host
> > Do you know into what path the glusterfs data goes? Or is it > written > directly onto a disk/LV? > > - fabian > > > "" http://www.ovirt.org/Features/GlusterFS_Storage_Domain > > * option rpc-auth-allow-insecure on ==> in > glusterd.vol (ensure > > u restart glusterd service... for this to take > effect) > > > * volume set <volname> server.allow-insecure on ==> > (ensure u > > stop and start the volume.. for this to take > effect)'' > > > > > > Thanks! > > > > > > > > > > > > On Wed, Dec 18, 2013 at 11:35 AM, Fabian Deutsch > <fabiand@redhat.com> > > wrote: > > Am Mittwoch, den 18.12.2013, 08:34 +0200 schrieb > Gabi C: > > > Hello! > > > > > > > > > In order to increase disk space I want to add a > new disk > > drive to > > > ovirt node. After adding this should I proceed as > "normal" - > > pvcreate, > > > vgcreate, lvcreate and so on - or these > configuration will > > not > > > persist? > > > > > > Hey Gabi, > > > > basically plain LVM is used in Node - so yes > pvcreate and > > lvextend can > > be used. > > What storage part do you want to extend? The part > where the > > VMs reside? > > You will also need to take care to extend the > filesystem. > > > > - fabian > > > > > > > > > >