[Starlingx-discuss] data-network down

[Peters, Matt]

Hi,
You can examine the logs from the OVS container.

kubectl -n openstack get pods -o wide | grep openvswitch-vswitchd
kubectl -n openstack logs <openvswitch pod name>

From: "von Hoesslin, Volker" <Volker.Hoesslin at swsn.de>
Date: Tuesday, December 10, 2019 at 2:54 AM
To: "Peters, Matt" <Matt.Peters at windriver.com>, "'starlingx-discuss at lists.starlingx.io'" <starlingx-discuss at lists.starlingx.io>
Subject: AW: [Starlingx-discuss] data-network down

Hi,
my OVS is configured like the recommendation from the installations guide: system vswitch_type=none

the lldp discovery shows currently the expected behavior, I will recheck on next error occurs. Btw, where can see OVS error logs, I’m not sure witch one of these billions of logs I can focus on for OVS errors…

Volker.

Von: Peters, Matt [mailto:Matt.Peters at windriver.com]
Gesendet: Montag, 9. Dezember 2019 19:38
An: von Hoesslin, Volker; 'starlingx-discuss at lists.starlingx.io'
Betreff: Re: [Starlingx-discuss] data-network down

Hello,
Are you running with containerized OVS (system vswitch_type=none) or OVS-DPDK on the host (system vswitch_type=ovs-dpdk)?

In either case, I would examine the logs from OVS to see if there are any errors being generated for the host that is exhibiting the failure.

In addition, you can check your infrastructure data network configuration to ensure it is setup as you expected by reviewing the LLDP information for the host (system host-lldp-neighbor-list) or the graphical display on the dashboard.

-Matt

From: "von Hoesslin, Volker" <Volker.Hoesslin at swsn.de>
Date: Monday, December 9, 2019 at 4:34 AM
To: "'starlingx-discuss at lists.starlingx.io'" <starlingx-discuss at lists.starlingx.io>
Subject: [Starlingx-discuss] data-network down

Hi,
my STX 2.0 cluster is up and running (2x controller, 2x worker, 3x storage). but one of my worker have an problem, i can not fix it. after time (i do not know in detail, in most times over night) the data-network from all VMs are broken down, no connection over data-network are available any more. the other worker is running fine without any problems. if i live migrate the VMs from one to another worker and back again, the data-network is comming up and all is fine again. i do not find any logs that will explain this problem, any suggestions or hints for locking for?

greez & thx,
volker.
Senden<https://webmail.sis-schwerin.de/owa/?ae=Item&t=IPM.Note&a=New&to=starlingx-discuss%40lists.starlingx.io&nm=%27starlingx-discuss%40lists.starlingx.io%27&pspid=_1575882727460_952755322>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.starlingx.io/pipermail/starlingx-discuss/attachments/20191212/0ab3df8b/attachment.html>