[ovirt-users] iSCSI Multipathing -> host inactive
Uwe Laverenz
uwe at laverenz.de
Wed Aug 24 12:22:27 UTC 2016
Hi,
sorry for the delay, I reinstalled everything, configured the networks,
attached the iSCSI storage with 2 interfaces and finally created the
iSCSI-bond:
> [root at ovh01 ~]# route
> Kernel IP Routentabelle
> Ziel Router Genmask Flags Metric Ref Use Iface
> default hp5406-1-srv.mo 0.0.0.0 UG 0 0 0 ovirtmgmt
> 10.0.24.0 0.0.0.0 255.255.255.0 U 0 0 0 ovirtmgmt
> 10.0.131.0 0.0.0.0 255.255.255.0 U 0 0 0 enp9s0f0
> 10.0.132.0 0.0.0.0 255.255.255.0 U 0 0 0 enp9s0f1
> link-local 0.0.0.0 255.255.0.0 U 1005 0 0 enp9s0f0
> link-local 0.0.0.0 255.255.0.0 U 1006 0 0 enp9s0f1
> link-local 0.0.0.0 255.255.0.0 U 1008 0 0 ovirtmgmt
> link-local 0.0.0.0 255.255.0.0 U 1015 0 0 bond0
> link-local 0.0.0.0 255.255.0.0 U 1017 0 0 ADMIN
> link-local 0.0.0.0 255.255.0.0 U 1021 0 0 SRV
and:
> [root at ovh01 ~]# ip a
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> inet 127.0.0.1/8 scope host lo
> valid_lft forever preferred_lft forever
> inet6 ::1/128 scope host
> valid_lft forever preferred_lft forever
> 2: enp13s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master ovirtmgmt state UP qlen 1000
> link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
> 3: enp8s0f0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master bond0 state UP qlen 1000
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 4: enp8s0f1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master bond0 state UP qlen 1000
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 5: enp9s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
> link/ether 90:e2:ba:11:21:d4 brd ff:ff:ff:ff:ff:ff
> inet 10.0.131.181/24 brd 10.0.131.255 scope global enp9s0f0
> valid_lft forever preferred_lft forever
> 6: enp9s0f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
> link/ether 90:e2:ba:11:21:d5 brd ff:ff:ff:ff:ff:ff
> inet 10.0.132.181/24 brd 10.0.132.255 scope global enp9s0f1
> valid_lft forever preferred_lft forever
> 7: ;vdsmdummy;: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN
> link/ether 26:b2:4e:5e:f0:60 brd ff:ff:ff:ff:ff:ff
> 8: ovirtmgmt: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
> link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
> inet 10.0.24.181/24 brd 10.0.24.255 scope global ovirtmgmt
> valid_lft forever preferred_lft forever
> 14: vnet0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master ovirtmgmt state UNKNOWN qlen 500
> link/ether fe:16:3e:79:25:86 brd ff:ff:ff:ff:ff:ff
> inet6 fe80::fc16:3eff:fe79:2586/64 scope link
> valid_lft forever preferred_lft forever
> 15: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 16: bond0.32 at bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master ADMIN state UP
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 17: ADMIN: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 20: bond0.24 at bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master SRV state UP
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
> 21: SRV: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
The host keeps toggling all storage domains on and off as soon as there
is an iSCSI bond configured.
Thank you for your patience.
cu,
Uwe
Am 18.08.2016 um 11:10 schrieb Elad Ben Aharon:
> I don't think it's necessary.
> Please provide the host's routing table and interfaces list ('ip a' or
> ifconfing) while it's configured with the bond.
>
> Thanks
>
> On Tue, Aug 16, 2016 at 4:39 PM, Uwe Laverenz <uwe at laverenz.de
> <mailto:uwe at laverenz.de>> wrote:
>
> Hi Elad,
>
> Am 16.08.2016 um 10:52 schrieb Elad Ben Aharon:
>
> Please be sure that ovirtmgmt is not part of the iSCSI bond.
>
>
> Yes, I made sure it is not part of the bond.
>
> It does seem to have a conflict between default and enp9s0f0/
> enp9s0f1.
> Try to put the host in maintenance and then delete the iscsi
> nodes using
> 'iscsiadm -m node -o delete'. Then activate the host.
>
>
> I tried that, I managed to get the iSCSI interface clean, no
> "default" anymore. But that didn't solve the problem of the host
> becoming "inactive". Not even the NFS domains would come up.
>
> As soon as I remove the iSCSI-bond, the host becomes responsive
> again and I can activate all storage domains. Removing the bond also
> brings the duplicated "Iface Name" back (but this time causes no
> problems).
>
> ...
>
> I wonder if there is a basic misunderstanding on my side: wouldn't
> it be necessary that all targets are reachable from all interfaces
> that are configured into the bond to make it work?
>
> But this would either mean two interfaces in the same network or
> routing between the iSCSI networks.
>
> Thanks,
>
> Uwe
> _______________________________________________
> Users mailing list
> Users at ovirt.org <mailto:Users at ovirt.org>
> http://lists.ovirt.org/mailman/listinfo/users
> <http://lists.ovirt.org/mailman/listinfo/users>
>
>
More information about the Users
mailing list