Bug 161513

Summary:	arping logic in /etc/init.d/netdump doesn't scale
Product:	Red Hat Enterprise Linux 4	Reporter:	Joshua Jensen <joshua>
Component:	netdump	Assignee:	Jeff Moyer <jmoyer>
Status:	CLOSED ERRATA	QA Contact:
Severity:	medium	Docs Contact:
Priority:	medium
Version:	4.0	CC:	rpb, tao
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:	RHBA-2006-0038	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2006-03-07 18:24:34 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:
Bug Depends On:
Bug Blocks:	168429

Description Joshua Jensen 2005-06-23 20:40:41 UTC

Description of problem:

For security reasons, my global organization may have to turn of proxy arp. 
This means that when arpping fails to find the MAC of the IP of the
netdump-server, the netdump startup script errors out.  This isn't a problem if
the client is on the same subnet/vlan as the server, but if it isn't we
unnecisarily see errors:

netdump: cannot arp 144.254.226.60
netdump: cannot find 144.254.226.60 in arp cache
netdump: can't resolve 144.254.226.60 MAC address
netdump server address resolution                          [FAILED]

Most enterprises *don't* have a netdump-server per subnet, they have one per
location.  In this setup we have to hardcode the MAC address of the gateway into
/etc/sysconfig/netdump.  We need all this, which causes netdump to fail, to get
the MAC that was hardcoded anyway?

Why rely on arping working?  Why not leave the MAC address in
/etc/sysconfig/netdump optional, even for netdump-across-subnet scenarios, and
just do this:

arping netdump_server_IP && use_server_MAC || arping default_gateway_IP &&
use_gateway_MAC

??


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

netdump in RHEL3 and RHEL4

Comment 1 Jeff Moyer 2005-06-27 19:42:40 UTC

I'll take a look.

Comment 4 Jeff Moyer 2005-08-01 15:40:40 UTC

I've built a fix for this into netdump-0.7.10-3, and posted the i386 package here:
  http://people.redhat.com/jmoyer/bz161513/

Note that this rpm is being provided for testing only.  It has not seen any QA
other than the testing I performed in my environment.  Please provide feedback.

Comment 5 Ray Bellis 2005-08-17 09:59:12 UTC

The change to /etc/rc.d/init.d/netdump works for me - the syslog server I wanted
to dump to is completely offsite, so I needed to be able to put the default GW's
MAC address in /etc/sysconfig/netdump, but previously couldn't without modifying
the startup script.

Will this make it into the next RHEL 3 update?

Comment 6 Jeff Moyer 2005-08-17 11:50:49 UTC

Thanks for the testing feedback.  This fix is not going to be in Update 6,
unfortunately.

Comment 7 Joshua Jensen 2005-08-26 20:36:53 UTC

How about U7 ??

Comment 16 Joshua Jensen 2005-12-28 16:28:40 UTC

PING.... How about U7?

Comment 18 Red Hat Bugzilla 2006-03-07 18:24:34 UTC

An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2006-0038.html