Multiple gateway not working after "big update"
Hi, i've just tried to update my sandboxing environmnent after long time without update (i know it's bad...) and i'm facing a problem with multiple gateway. This environment work with 6 hypervisors (centos 8) and one manager on dedicated server, all this baremetal have 2 interface in 2 différent network protected by firewall. All my hypervisords have 2 network for ovirt, one "ovirtmgmt" for management (only SSH and internal admin stuff), the other "ovir_prod" for console, migration, and default gateway. I've 3 hypervisors non updated which i can ping /ssh / ... (according to iptables) on the two interface without problem. But the 3 hypervisors just updated are accessible only on the "ovirt_prod" interface. I'm looking inside ip route/ ip rules it seems to miss something. I always configured my HV by ovirt gui so i does not understand what going wrong. This is ip route/ip rules of my hv, it's like missing ip rules on non working HV The "route-ovirt_prod" file that does not exist on old HV. Working HV (centos 8.3 / vdsm 4.20.26.3-1) : [root@drogo ~]# ip ru 0: from all lookup local 32000: from 10.100.126.0/24 lookup 174358038 32000: from all to 10.100.126.0/24 iif ovirtmgmt lookup 174358038 32000: from 10.5.7.0/24 lookup 168101654 32000: from all to 10.5.7.0/24 iif comte_prod lookup 168101654 32766: from all lookup main 32767: from all lookup default [root@drogo ~]# ip rou default via 10.5.7.254 dev comte_prod proto static metric 425 10.5.7.0/24 dev comte_prod proto kernel scope link src 10.5.7.22 metric 425 10.100.126.0/24 dev ovirtmgmt proto kernel scope link src 10.100.126.22 metric 426 Bugged HV (centos 8.4/ vdsm 4.40.70.6-1) [root@balbo ~]# ip ru 0: from all lookup local 32766: from all lookup main 32767: from all lookup default [root@balbo ~]# ip rou default via 10.5.7.254 dev comte_prod proto static metric 425 10.5.6.0/26 dev enp4s0f1 proto kernel scope link src 10.5.6.6 metric 100 linkdown 10.5.7.0/24 dev comte_prod proto kernel scope link src 10.5.7.25 metric 426 10.100.126.0/24 dev ovirtmgmt proto kernel scope link src 10.100.126.25 metric 425 [root@bingo network-scripts]# cat route-ovirt_prod ADDRESS0=0.0.0.0 NETMASK0=0.0.0.0 GATEWAY0=10.5.7.254 OPTIONS0="table 254" METRIC0=431 Than you for help :) -- Lionel Caignec
On Fri, Jul 9, 2021 at 9:26 AM Lionel Caignec <caignec@cines.fr> wrote:
Hi,
i've just tried to update my sandboxing environmnent after long time without update (i know it's bad...) and i'm facing a problem with multiple gateway. This environment work with 6 hypervisors (centos 8) and one manager on dedicated server, all this baremetal have 2 interface in 2 différent network protected by firewall. All my hypervisords have 2 network for ovirt, one "ovirtmgmt" for management (only SSH and internal admin stuff), the other "ovir_prod" for console, migration, and default gateway.
I've 3 hypervisors non updated which i can ping /ssh / ... (according to iptables) on the two interface without problem. But the 3 hypervisors just updated are accessible only on the "ovirt_prod" interface. I'm looking inside ip route/ ip rules it seems to miss something. I always configured my HV by ovirt gui so i does not understand what going wrong.
This is ip route/ip rules of my hv, it's like missing ip rules on non working HV The "route-ovirt_prod" file that does not exist on old HV.
Working HV (centos 8.3 / vdsm 4.20.26.3-1) : [root@drogo ~]# ip ru 0: from all lookup local 32000: from 10.100.126.0/24 lookup 174358038 32000: from all to 10.100.126.0/24 iif ovirtmgmt lookup 174358038 32000: from 10.5.7.0/24 lookup 168101654 32000: from all to 10.5.7.0/24 iif comte_prod lookup 168101654 32766: from all lookup main 32767: from all lookup default
[root@drogo ~]# ip rou default via 10.5.7.254 dev comte_prod proto static metric 425 10.5.7.0/24 dev comte_prod proto kernel scope link src 10.5.7.22 metric 425 10.100.126.0/24 dev ovirtmgmt proto kernel scope link src 10.100.126.22 metric 426
Bugged HV (centos 8.4/ vdsm 4.40.70.6-1) [root@balbo ~]# ip ru 0: from all lookup local 32766: from all lookup main 32767: from all lookup default
[root@balbo ~]# ip rou default via 10.5.7.254 dev comte_prod proto static metric 425 10.5.6.0/26 dev enp4s0f1 proto kernel scope link src 10.5.6.6 metric 100 linkdown 10.5.7.0/24 dev comte_prod proto kernel scope link src 10.5.7.25 metric 426 10.100.126.0/24 dev ovirtmgmt proto kernel scope link src 10.100.126.25 metric 425
[root@bingo network-scripts]# cat route-ovirt_prod ADDRESS0=0.0.0.0 NETMASK0=0.0.0.0 GATEWAY0=10.5.7.254 OPTIONS0="table 254" METRIC0=431
Than you for help :)
-- Lionel Caignec _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/3NSSFIKMKSMCTP...
Hi, from the vdsm version it seems that the upgrade was from 4.2 to 4.4.7. Direct upgrade like that is not supported. There were a lot of changes 4.4 networking among other changes which might result in what you are experiencing. Please take a look at upgrade path from 4.2 to 4.4 in https://www.ovirt.org/documentation/upgrade_guide/index.html Thanks, Ales -- Ales Musil Software Engineer - RHV Network Red Hat EMEA <https://www.redhat.com> amusil@redhat.com IM: amusil <https://red.ht/sig>
Hi thank you for answer, on the ovirt-engine, for what i know i upgraded from 4.4.3 to 4.4.7. so i think i respect upgrade path and i used engine-setup command like in the manual. But on the vdsm side i does not know there is upgrade path, i already tried to remove/readd one buggy host but without succes. so what do you think i can do? Lionel Caignec. ----- Mail original ----- De: "Ales Musil" <amusil@redhat.com> À: "Lionel Caignec" <caignec@cines.fr> Cc: "users" <users@ovirt.org> Envoyé: Vendredi 9 Juillet 2021 09:45:15 Objet: Re: [ovirt-users] Multiple gateway not working after "big update" Hi, from the vdsm version it seems that the upgrade was from 4.2 to 4.4.7. Direct upgrade like that is not supported. There were a lot of changes 4.4 networking among other changes which might result in what you are experiencing. Please take a look at upgrade path from 4.2 to 4.4 in https://www.ovirt.org/documentation/upgrade_guide/index.html Thanks, Ales -- Ales Musil Software Engineer - RHV Network Red Hat EMEA <https://www.redhat.com> amusil@redhat.com IM: amusil <https://red.ht/sig>
For the host there is the https://www.ovirt.org/documentation/upgrade_guide/index.html#Upgrading_hosts... section. In short, migrate all VMs to a different host, put the host into maintenance, remove the host from the engine, reinstall the host OS completely (to clean up any potential leftovers). Then you can proceed as usual. Hopefully this helps. Regards, Ales On Fri, Jul 9, 2021 at 10:12 AM Lionel Caignec <caignec@cines.fr> wrote:
Hi
thank you for answer, on the ovirt-engine, for what i know i upgraded from 4.4.3 to 4.4.7. so i think i respect upgrade path and i used engine-setup command like in the manual.
But on the vdsm side i does not know there is upgrade path, i already tried to remove/readd one buggy host but without succes.
so what do you think i can do?
Lionel Caignec.
----- Mail original ----- De: "Ales Musil" <amusil@redhat.com> À: "Lionel Caignec" <caignec@cines.fr> Cc: "users" <users@ovirt.org> Envoyé: Vendredi 9 Juillet 2021 09:45:15 Objet: Re: [ovirt-users] Multiple gateway not working after "big update"
Hi,
from the vdsm version it seems that the upgrade was from 4.2 to 4.4.7. Direct upgrade like that is not supported. There were a lot of changes 4.4 networking among other changes which might result in what you are experiencing.
Please take a look at upgrade path from 4.2 to 4.4 in https://www.ovirt.org/documentation/upgrade_guide/index.html
Thanks, Ales
--
Ales Musil
Software Engineer - RHV Network
Red Hat EMEA <https://www.redhat.com>
amusil@redhat.com IM: amusil <https://red.ht/sig>
-- Ales Musil Software Engineer - RHV Network Red Hat EMEA <https://www.redhat.com> amusil@redhat.com IM: amusil <https://red.ht/sig>
Ok thank you very much i'll try that. Regards, Lionel. ----- Mail original ----- De: "Ales Musil" <amusil@redhat.com> À: "Lionel Caignec" <caignec@cines.fr> Cc: "users" <users@ovirt.org> Envoyé: Vendredi 9 Juillet 2021 10:22:49 Objet: Re: [ovirt-users] Multiple gateway not working after "big update" For the host there is the https://www.ovirt.org/documentation/upgrade_guide/index.html#Upgrading_hosts... section. In short, migrate all VMs to a different host, put the host into maintenance, remove the host from the engine, reinstall the host OS completely (to clean up any potential leftovers). Then you can proceed as usual. Hopefully this helps. Regards, Ales
That was that i uninstall all ovirt related software purge some /var/lib folder readd client and now all work like charm. Thank you ----- Mail original ----- De: "Lionel Caignec" <caignec@cines.fr> À: "Ales Musil" <amusil@redhat.com> Cc: "users" <users@ovirt.org> Envoyé: Vendredi 9 Juillet 2021 10:27:41 Objet: Re: [ovirt-users] Multiple gateway not working after "big update" Ok thank you very much i'll try that. Regards, Lionel.
Thank you for reporting back. Glad it worked. Regards, Ales On Fri, Jul 9, 2021 at 1:38 PM Lionel Caignec <caignec@cines.fr> wrote:
That was that i uninstall all ovirt related software purge some /var/lib folder readd client and now all work like charm.
Thank you
----- Mail original ----- De: "Lionel Caignec" <caignec@cines.fr> À: "Ales Musil" <amusil@redhat.com> Cc: "users" <users@ovirt.org> Envoyé: Vendredi 9 Juillet 2021 10:27:41 Objet: Re: [ovirt-users] Multiple gateway not working after "big update"
Ok thank you very much i'll try that.
Regards, Lionel.
-- Ales Musil Software Engineer - RHV Network Red Hat EMEA <https://www.redhat.com> amusil@redhat.com IM: amusil <https://red.ht/sig>
participants (2)
-
Ales Musil -
Lionel Caignec