Bug 1327867 - Introspection failed as the TFTP port does not get whitelisted in the firewall rules
Summary: Introspection failed as the TFTP port does not get whitelisted in the firewal...
Keywords:
Status: CLOSED WORKSFORME
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: instack-undercloud
Version: 7.0 (Kilo)
Hardware: x86_64
OS: Linux
unspecified
high
Target Milestone: ---
: 7.0 (Kilo)
Assignee: James Slagle
QA Contact: Udi Shkalim
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2016-04-17 08:38 UTC by Udi Shkalim
Modified: 2017-02-09 10:18 UTC (History)
11 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2017-02-09 10:18:31 UTC
Target Upstream Version:


Attachments (Terms of Use)
facter and iptables rules (7.80 KB, text/plain)
2016-04-17 08:38 UTC, Udi Shkalim
no flags Details
undercloud_installation_log (1.13 MB, text/plain)
2016-04-18 11:48 UTC, Udi Shkalim
no flags Details

Description Udi Shkalim 2016-04-17 08:38:40 UTC
Created attachment 1148051 [details]
facter and iptables rules

Description of problem:
On OSPD 7.3 Introspection of nodes failed due to TFTP Timeout. When dropping all of the iptables rules introspection pass.

From the overcloud node:

Intel(R) Boot Agent GE v1.3.24                                                  
Copyright (C) 1997-2008, Intel Corporation                                      
                                                                                
CLIENT MAC ADDR: 44 1E A1 73 3D 43  GUID: 36303930 3935 435A 3332 313142363630 
.           
CLIENT IP: 192.0.2.114  MASK: 255.255.255.0  DHCP IP: 192.0.2.1                 
GATEWAY IP: 192.0.2.1                                                           
PXE-E32: TFTP open timeout                                                      
PXE-E32: TFTP open timeout                                                      
PXE-E32: TFTP open timeout                                                      
PXE-M0F: Exiting Intel Boot Agent.                                                                   
Boot failed, continue to look for Boot Media                                    
Wait for input.....                                   


Version-Release number of selected component (if applicable):
python-ironicclient-0.5.1-12.el7ost.noarch
openstack-ironic-discoverd-1.1.0-8.el7ost.noarch
ipxe-bootimgs-20150821-1.git4e03af8e.el7.noarch
openstack-ironic-common-2015.1.2-2.el7ost.noarch
python-ironic-discoverd-1.1.0-8.el7ost.noarch
ipxe-roms-qemu-20150821-1.git4e03af8e.el7.noarch
openstack-ironic-api-2015.1.2-2.el7ost.noarch
openstack-ironic-conductor-2015.1.2-2.el7ost.noarch


How reproducible:
On my env 100%

Steps to Reproduce:
1. Install 7.3 undercloud and follow the installation guide
2. When you get to introspection stage, run:  openstack baremetal introspection bulk start
3. Introspection timeout

Actual results:
Introspection of nodes failed

Expected results:
Introspection of nodes success.

Additional info:
sosreport and baremetal spec is attached.

Comment 3 Dmitry Tantsur 2016-04-18 07:39:10 UTC
Hi! Please provide logs from the openstack-ironic-discoverd and openstack-ironic-discoverd-dnsmasq services.

Comment 5 Udi Shkalim 2016-04-18 08:28:28 UTC
Hi, Please see this new link

http://ikook.tlv.redhat.com/uploads/general/sosreport-ushkalim-20160417112034.tar.xz

Comment 6 Udi Shkalim 2016-04-18 11:48:23 UTC
Created attachment 1148181 [details]
undercloud_installation_log

Comment 7 Tzach Shefi 2016-04-19 13:13:41 UTC
Adding info about my deployment's possibly related issue. 
Undercloud (ospd 8 2016-03-29.3) passed BM introspection, but overcloud deploy of 4 BMs fails with "no valid host". 
Out of the 4 BM nodes, only one deploys OS. Looking at other node console I see no DHCP address is given on BM reboot hence no OS deployed.  

Deleted ironic nodes introspcted them a few times, this works fine, but overcloud deploy fails over and over. 

One more tip/hint I can think of I've used this undercloud for a while now deploying and deleting quite a few stacks on it successfully before this hit me. Maybe this causes the issue.

Comment 8 Dmitry Tantsur 2017-02-07 11:45:58 UTC
Hi, sorry, this report has fallen out of our team's radar. Is this still an issue?

Comment 9 Udi Shkalim 2017-02-09 09:14:50 UTC
Hi,

I have not installed osp7 lately. Haven't seen this in the latest ops 10/11 though.

Comment 10 Dmitry Tantsur 2017-02-09 10:18:31 UTC
Got it, please reopen if you see it again.


Note You need to log in before you can comment on or make changes to this bug.