Bug 149737

Summary:	clumanager starts even if no non-slave interface is available
Product:	[Retired] Red Hat Cluster Suite	Reporter:	Lon Hohberger <lhh>
Component:	clumanager	Assignee:	Lon Hohberger <lhh>
Status:	CLOSED WORKSFORME	QA Contact:	Cluster QE <mspqa-list>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	3	CC:	ckloiber, cluster-maint
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2005-09-29 19:47:17 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Lon Hohberger 2005-02-25 21:31:10 UTC

Description of problem:

In an incorrect networking configuration where all physical interfaces
are slaves with no bonding master present, clumanager starts and tries
to run anyway.


Version-Release number of selected component (if applicable): 1.2.22


How reproducible: Unknown


Steps to Reproduce: Unknown

Actual results:

Feb 24 16:36:27 node1 clumembd[6444]: <debug> Multicast address is
225.0.0.11 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> I am member #0 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> Interface IP is 127.0.0.1 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> eth0 is a SLAVE; skipping! 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> eth1 is a SLAVE; skipping! 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> Interface IP is 192.168.0.2 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> Setting up multicast
225.0.0.11 on 192.168.0.2 
Feb 24 16:36:27 node1 clumembd[6444]: <debug> Cluster I/F: eth0
[192.168.0.2] 

Expected results:

Clumanager shouldn't start if no master (and/or only slave) ethernet
devices have the cluster node's IP address bound.

Comment 1 Lon Hohberger 2005-07-11 21:50:30 UTC

Chris, is this related to all the other ARP and bonding routing problems we've
been seeing lately?

Comment 2 Lon Hohberger 2005-07-27 16:44:01 UTC

clumembd checks for IFF_SLAVE during startup -- so does the message service
initialization which is reporting eth0 as the cluster I/F... I can't get this to
happen.

Comment 3 Lon Hohberger 2005-09-29 19:47:17 UTC

This is a network configuration problem - I can't solve them all.