Bug 1575883

Summary:	DEFROUTE set to no after update from vdsm 4.17 -> 4.19
Product:	Red Hat Enterprise Virtualization Manager	Reporter:	Nikola Kresic <nkresic>
Component:	vdsm	Assignee:	Edward Haas <edwardh>
Status:	CLOSED INSUFFICIENT_DATA	QA Contact:	Raz Tamir <ratamir>
Severity:	medium	Docs Contact:
Priority:	low
Version:	3.6.0	CC:	cshao, danken, eedri, jbelka, lsurette, mburman, rbarry, srevivo, ycui, ykaul
Target Milestone:	---
Target Release:	---
Hardware:	Unspecified
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2018-06-24 07:21:37 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	Network	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Nikola Kresic 2018-05-08 07:47:18 UTC

Description of problem:
Very similar to the following bugzilla #1506550
Looks like a regression. Customer is updating the nodes, the network config files are old and created long time ago, never changed or edited, defroute=yes overwritten to defroute=no after update so that server hangs at boot until the defroute parameter is fixed manually.

Version-Release number of selected component (if applicable):
vdsm latest

How reproducible:
update from 4.17 to 4.19 tested, no idea if it hits other update paths

Steps to Reproduce:
1. one way
- Updated the host first from RHEVM: result was that avter VDSM Upgrade the host was unreachable.
- Previous SSH session was active 
- Set Host to Maintenance
- yum upgrade for rest of RHEL updates (kernel, etc.)
- Rebooted host
- Host not reachable via SSH
- From HW console: it is determined that there is no default route!!
- Set defroute=yes in /etc/sysconfig/network-scripts/ifcfg-rhevm
- service network restart
- Host is now activateable from RHEVM

2. second way
- Set host to maintenance
- yum upgrade
- reboot
- Host is not reachable via SSH
- On HW console again: no default route
- Edited /etc/sysconfig/network-scripts/ifcfg-rhevm, set defroute=yes
- service network restart
- Activate host in RHEVM


Actual results:
config file changed so that it breaks config

Expected results:
config file untouched during update, or at least not in a way that it breaks config

Additional info:

Comment 1 Dan Kenigsberg 2018-05-15 10:04:45 UTC

Would you please attach supervdsm.log from the time of the upgrade and first boot?

Comment 2 Jiri Belka 2018-05-22 11:37:03 UTC

can't reproduce, 3.6/el7.3 eus/els host to 4.1/el7.5:

# rpm -q vdsm kernel-$( uname -r ) initscripts redhat-release-server
vdsm-4.17.45-1.el7ev.noarch
kernel-3.10.0-693.21.1.el7.x86_64
initscripts-9.49.37-1.el7.x86_64
redhat-release-server-7.3-7.el7.x86_64

both ifcfg-* and vdsm's chech file in /var/run/vdsm have yes for defroute.

putting host into maintenance and doing update:

# rpm -q vdsm kernel-$( uname -r ) initscripts redhat-release-server
vdsm-4.19.51-1.el7ev.x86_64
kernel-3.10.0-862.3.2.el7.x86_64
initscripts-9.49.41-1.el7.x86_64
redhat-release-server-7.5-8.el7.x86_64

reboot

# grep -i 'def.*route' /etc/sysconfig/network-scripts/ifcfg-ovirtmgmt 
DEFROUTE=yes
[root@10-37-140-151 ~]# grep -iIR 'def.*route' /var/run/vdsm/ 2>/dev/null
/var/run/vdsm/netconf/nets/ovirtmgmt:    "defaultRoute": true

Comment 3 Jiri Belka 2018-05-22 11:38:33 UTC

We have already release 4.2, thus vdsm is 4.20.x now, could you try with 4.2?

Comment 4 Jiri Belka 2018-05-23 07:55:34 UTC

I upgraded also from 3.6 to 4.2 and it still works too.

Comment 5 Dan Kenigsberg 2018-06-24 07:21:37 UTC

Please reopen with the required logs. Please also include your /etc/vdsm/vdsm.conf

Comment 6 Franta Kust 2019-05-16 13:03:41 UTC

BZ<2>Jira Resync

Comment 7 Red Hat Bugzilla 2023-09-15 00:08:01 UTC

The needinfo request[s] on this closed bug have been removed as they have been unresolved for 500 days