Bug 1297853

Summary:	Old Ipxe roms cause a host of problems.
Product:	Red Hat Enterprise Linux 7	Reporter:	Jeremy <jmelvin>
Component:	ipxe	Assignee:	Miroslav Rezanina <mrezanin>
Status:	CLOSED WORKSFORME	QA Contact:	Raviv Bar-Tal <rbartal>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	7.4	CC:	achernet, areis, dgilbert, huding, jen, jmelvin, juzhang, mrezanin, weliao, xfu
Target Milestone:	rc	Keywords:	RFE
Target Release:	---	Flags:	achernet: needinfo+
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	ipxe-20150821-1.git4e03af8e.el7	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-09-11 12:59:01 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Jeremy 2016-01-12 16:02:40 UTC

Description of problem: Requesting newer ipxe roms in the next release. The problem here showed itself in openstack:
virtual machines with the pxe_ssh Ironic driver, and the virtualized nodes get stuck in "wait call-back" when attempting to deploy an overcloud. This happens when the VM has more than 1 network interface attached during the PXE process.

If more than 1 interface is attached, the VM will use the first NIC, vnet0, to obtain a DHCP lease, but then use the MAC from vnet1 in the URL when attempting to download the MAC specific PXE config file from http://x.x.x.x/pxelinux.cfg/xx-xx-xx-xx-xx-xx.

If the only attached interface is the provisioning network interface, the correct MAC is used for both steps and the OS install proceeds.

Version-Release number of selected component (if applicable):
ipxe-roms-qemu-20130517-7.gitc4bce43.el7.noarch

How reproducible:

Steps to Reproduce:
1.
2.
3.

Actual results:

Expected results:

Additional info:I have confirmed that updating the ipxe-roms-qemu package on the virtualization host resolves this issue:

---> Package ipxe-roms-qemu.noarch 0:20130517-7.gitc4bce43.el7 will be updated
---> Package ipxe-roms-qemu.noarch 0:20140303-1.gitff1e7fc7.fc20 will be an update

[root@716198-controller06 ~]# cd /usr/share/ipxe
[root@716198-controller06 ipxe]# ls -al
total 368
drwxr-xr-x. 2 root root 101 Jan 12 09:33 .
drwxr-xr-x. 117 root root 4096 Jan 12 09:33 ..
-rw-r--r--. 1 root root 73216 Mar 3 2014 10222000.rom
-rw-r--r--. 1 root root 73216 Mar 3 2014 10ec8029.rom
-rw-r--r--. 1 root root 74752 Mar 3 2014 10ec8139.rom
-rw-r--r--. 1 root root 72704 Mar 3 2014 1af41000.rom
-rw-r--r--. 1 root root 73728 Mar 3 2014 8086100e.rom

With the updated iPXE ROMs, the VMs were able to succesfully PXE boot even with multiple NICs attached at boot time.

Can we get these ROMs updated in the next RHEL release? They are over 2 1/2 years old.

Comment 7 Raviv Bar-Tal 2016-09-11 12:59:01 UTC

The bug was not reproduce by Wei nor by myself.
In any case the ipxe rpms are update in new releases to ipxe-roms-qemu-20150821-1.git4e03af8e.el7.noarch which is much newer then the one solve the origianl problem for Jeremy
ipxe-roms-qemu.noarch0:20140303-1.gitff1e7fc7.fc20