Bug 789145 - Anaconda fails to download kickstart file when specified using "ks=http://" or "ks=nfs:" on UV systems.
Summary: Anaconda fails to download kickstart file when specified using "ks=http://" o...
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: anaconda
Version: 6.2
Hardware: x86_64
OS: Linux
high
medium
Target Milestone: rc
: ---
Assignee: Radek Vykydal
QA Contact: Release Test Team
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2012-02-09 23:07 UTC by Tony Ernst
Modified: 2012-07-10 23:04 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2012-07-10 15:35:24 UTC
Target Upstream Version:


Attachments (Terms of Use)
Screenshot of anaconda error. (24.18 KB, image/png)
2012-02-09 23:07 UTC, Tony Ernst
no flags Details
Log files from /tmp (23.64 KB, application/x-bzip2)
2012-02-09 23:08 UTC, Tony Ernst
no flags Details
/tmp/anaconda.log (2.69 KB, text/plain)
2012-02-10 14:17 UTC, Tony Ernst
no flags Details
/tmp/syslog (221.71 KB, text/plain)
2012-02-10 14:21 UTC, Tony Ernst
no flags Details

Description Tony Ernst 2012-02-09 23:07:07 UTC
Created attachment 560729 [details]
Screenshot of anaconda error.

Description of problem:

When doing a kickstart install of RHEL6.2 GA, specifying the kickstart file with "ks=http://" or "ks=nfs:" fails on SGI UV systems, which use EFI. (It works on other non-EFI x86_64 systems.)

Anaconda complains:

    Welcome to Red Hat Enterprise Linux for x86_64
    
    Error downloading kickstart file
    
    Unable to download the kickstart file.  Please modify the
    kickstart parameter below or press Cancel to proceed as an
    interactive installation.
    
    http://128.162.237.12/~tee/RHEL62/ks-x86_64.cfg

The only protocol that works is "ks=hd:sda4:". This seems to imply that the network isn't working. But if I then do ^Z to get a shell prompt, I am able to verify that the network is up by downloading the same kickstart file.

    bash-4.1# wget http://128.162.237.12/~tee/RHEL62/ks-x86_64.cfg
    --2001-01-03 17:14:31--  http://128.162.237.12/~tee/RHEL62/ks-x86_64.cfg
    Connecting to 128.162.237.12:80... connected.
    HTTP request sent, awaiting response... 200 OK
    Length: 25345 (25K) [text/plain]
    Saving to: `ks-x86_64.cfg'
    
    100%[======================================>] 25,345   --.-K/s  in 0.002s  
    
    2001-01-03 17:14:31 (9.84 MB/s) - `ks-x86_64.cfg' saved [25345/25345]


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

RHEL6.2 GA

How reproducible:

100%

Steps to Reproduce:
1. Try to do a kickstart install using "ks=http://" or "ks=nfs:" on an SGI UV system.
  
Actual results:

Anaconda complains "Error downloading kickstart file".

Expected results:

The kickstart file should be downloaded, and the install started.

Additional info:

Comment 1 Tony Ernst 2012-02-09 23:08:57 UTC
Created attachment 560730 [details]
Log files from /tmp

Here are all of the log files from /tmp after the error occurs.

Comment 3 Brian Lane 2012-02-10 01:47:20 UTC
Please attach the logs as individual plain/text attachments. This makes it alot easier to evaluate them, and allows bugzilla to search them.

Comment 4 Tony Ernst 2012-02-10 14:17:57 UTC
Created attachment 560900 [details]
/tmp/anaconda.log

Comment 5 Tony Ernst 2012-02-10 14:21:27 UTC
Created attachment 560901 [details]
/tmp/syslog

Comment 6 Tony Ernst 2012-02-10 14:25:13 UTC
I've reattached all of the files from /tmp. Please note that there was also an empty zero-length ks.cfg in /tmp.

Comment 7 Chris Lumens 2012-02-10 14:34:51 UTC
In the syslog, note the relatively short amount of time between when ethX comes up and when it attempts to fetch the kickstart file.  I wonder if there's some sort of timing problem here where the interface isn't really quite ready yet.  Your test from the shell works because enough time has elapsed.

Comment 8 Tony Ernst 2012-02-10 15:12:48 UTC
Hmm. You're right. After the error, if I wait a minute and try again, it works.

So when it says "Waiting for NetworkManager to configure eth0", it must not really be waiting until the interface is up.

Comment 9 RHEL Program Management 2012-05-03 05:11:39 UTC
Since RHEL 6.3 External Beta has begun, and this bug remains
unresolved, it has been rejected as it is not proposed as
exception or blocker.

Red Hat invites you to ask your support representative to
propose this request, if appropriate and relevant, in the
next release of Red Hat Enterprise Linux.

Comment 10 David Cantrell 2012-07-03 18:45:04 UTC
No, the waiting for NetworkManager message is waiting for NetworkManager to tell us it's done.  "Done" is defined by NetworkManager in that sense, which may or may not translate to "wait until the network link is up and traffic is flowing".  A workaround that may work in your case is the linksleep= parameter.  Add linksleep=5 or something like that to your installer boot command line and see if that works.

Comment 11 Tony Ernst 2012-07-03 19:02:47 UTC
I'm no longer set up to test this, but I'll make a note of the linksleep= parameter in case it ever comes up again. I've lowered the Severity, since there is an easy workaround (waiting a few seconds and retrying). Thanks.

Comment 12 RHEL Program Management 2012-07-10 07:08:18 UTC
This request was not resolved in time for the current release.
Red Hat invites you to ask your support representative to
propose this request, if still desired, for consideration in
the next release of Red Hat Enterprise Linux.

Comment 13 David Cantrell 2012-07-10 15:27:09 UTC
devel_ack- per comment #11.

Comment 14 RHEL Program Management 2012-07-10 15:35:24 UTC
Development Management has reviewed and declined this request.
You may appeal this decision by reopening this request.

Comment 15 RHEL Program Management 2012-07-10 23:04:22 UTC
This request was erroneously removed from consideration in Red Hat Enterprise Linux 6.4, which is currently under development.  This request will be evaluated for inclusion in Red Hat Enterprise Linux 6.4.


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