Fix network issues in PHX preventing foreman from installing OS

Description

Working with the local team in PHX to resolve network issues preventing us from installing more servers for Lago and more hypervisors for jenkins slaves.

Activity

Show:
Eyal Edri
June 6, 2016, 7:19 AM

following the thread with Italo, are we OK now with configuration for all hypervisors or there is an action needed to do before reinstalling each hypervisor?

Evgheni Dereveanchin
June 6, 2016, 7:46 AM

We have a proper configuration now, however I'd choose to put hosts into Maintenance and evacuate VMs on each host before changing it just to avoid disruptions. Then we can update settings on host-by-host basis. This only has to be done once, later we will not need to involve the networking team for re-provisioning.

Evgheni Dereveanchin
June 21, 2016, 3:49 PM

So far bonding was fixed on the following hypervisors:

ovirt-srv01
ovirt-srv04
ovirt-srv05
ovirt-srv06
ovirt-srv07
ovirt-srv11
ovirt-srv12
ovirt-srv13
ovirt-srv14
ovirt-srv17
ovirt-srv18

the only two remaining ones are the Production Fedora hosts which are due to be re-provisioned this week:
ovirt-srv02
ovirt-srv03

Eyal Edri
July 4, 2016, 11:17 AM

this was resolved right?

Evgheni Dereveanchin
July 4, 2016, 11:42 AM

just one single hypervisor remains: ovirt-srv02 but we have a known fix to make it work once we need to rebuild. We can close this as the cause was found and fixed on the vast majority of hypervisors.

Assignee

Evgheni Dereveanchin

Reporter

Eyal Edri

Blocked By

None

Components

Priority

Highest
Configure