[ovirt-users] iSCSI Multipathing -> host inactive
Elad Ben Aharon
ebenahar at redhat.com
Wed Aug 24 12:39:44 UTC 2016
Network configuration seems OK.
Please provide engine.log and vdsm.log
Thanks
On Wed, Aug 24, 2016 at 3:22 PM, Uwe Laverenz <uwe at laverenz.de> wrote:
> 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>
>>
>>
>> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20160824/8b179898/attachment-0001.html>
More information about the Users
mailing list