This service will be undergoing maintenance at 00:00 UTC, 2016-08-01. It is expected to last about 1 hours
Bug 677953 - Unable to kickstart with ks.cfg file stored on hard drive
Unable to kickstart with ks.cfg file stored on hard drive
Status: CLOSED NOTABUG
Product: Fedora
Classification: Fedora
Component: kernel (Show other bugs)
rawhide
i686 Unspecified
unspecified Severity medium
: ---
: ---
Assigned To: Kernel Maintainer List
Fedora Extras Quality Assurance
: Reopened
Depends On:
Blocks: F15Beta/F15BetaBlocker
  Show dependency treegraph
 
Reported: 2011-02-16 06:12 EST by Hongqing Yang
Modified: 2011-03-14 04:49 EDT (History)
9 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Environment:
Last Closed: 2011-03-14 04:49:21 EDT
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:


Attachments (Terms of Use)
anaconda log (2.35 KB, text/plain)
2011-02-20 20:14 EST, Hongqing Yang
no flags Details
program log (948 bytes, text/plain)
2011-02-20 20:18 EST, Hongqing Yang
no flags Details
syslog (159.95 KB, text/plain)
2011-02-22 04:20 EST, He Rui
no flags Details
lsmod (1.04 KB, text/plain)
2011-02-22 04:20 EST, He Rui
no flags Details
kickstart from hard drive (16.24 KB, image/png)
2011-03-09 23:24 EST, Hongqing Yang
no flags Details

  None (edit)
Description Hongqing Yang 2011-02-16 06:12:58 EST
Description of problem:

with cmdline  ks=hd:sda1:/ks.cfg   get the following info:
kernel command line:
initrd=initrd.img
BOOT_IMAGE=vmlinuz
ks=hd:sda1:/ks.cfg
anaconda version 15.20 on i686 staring
getting kickstart file
getting kickstart file from harddrive
loading ks from device on sda1 on path /ks.cfg
GetFileFromBlockDevice(sda1, /ks.cfg)
sleeping to wait for USB storage devices
error code: 32
Failed to mount /dev/sda1: No such file or directory.

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


How reproducible:


Steps to Reproduce:
1.boot system
2.enter ks=hd:sda1:/ks.cfg at prompt
3.
  
Actual results:


Expected results:


Additional info:

Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/sda2            229807308  29675652 188458056  14% /
tmpfs                  4126888       524   4126364   1% /dev/shm
/dev/sda1              2015824     74848   1838576   4% /boot
nfs.englab.nay.redhat.com:/fedoratest
                     523860704  20985824 502874880   5% /home/hongqing/mnt
Comment 1 Chris Lumens 2011-02-16 10:08:03 EST
Please attach /tmp/anaconda.log and /tmp/program.log to this bug report.
Comment 2 Hongqing Yang 2011-02-20 20:14:18 EST
Created attachment 479804 [details]
anaconda log
Comment 3 Hongqing Yang 2011-02-20 20:18:05 EST
Created attachment 479805 [details]
program log
Comment 4 Chris Lumens 2011-02-21 10:59:24 EST
Okay, that's consistent with the device just never being found.  Can you also attach /tmp/syslog and the output of lsmod?  Thanks.
Comment 5 He Rui 2011-02-22 04:20:05 EST
Created attachment 480071 [details]
syslog
Comment 6 He Rui 2011-02-22 04:20:40 EST
Created attachment 480072 [details]
lsmod
Comment 7 He Rui 2011-02-22 04:24:40 EST
Reproduced in anaconda 15.20.1.
Comment 8 Chris Lumens 2011-02-25 11:12:16 EST
Looks like the usb-storage module is never getting loaded.  You see no mention of it in lsmod, and there's very little mention of usb anything in your syslog.  It's my understanding that usb-storage should just be getting loaded without any anaconda involvement.  At least, we don't load it anywhere else and we've gotten bug reports from people farther along in the USB process, so it must be working for some.
Comment 9 Chuck Ebbert 2011-02-25 14:39:12 EST
(In reply to comment #5)
> Created attachment 480071 [details]
> syslog

Here sda and its partitions get discovered:
09:03:53,944 INFO kernel:[    5.122413]  sda: sda1 sda2 sda3 sda4 < sda5 sda6 >

And here sda1 gets mounted:
09:06:24,322 INFO kernel:[  155.790703] EXT3-fs (sda1): mounted filesystem with ordered data mode
09:06:24,322 DEBUG kernel:[  155.791610] SELinux: initialized (dev sda1, type ext3), uses xattr
Comment 10 Chuck Ebbert 2011-03-09 22:11:11 EST
sda1 gets discovered and its partition gets mounted, so what bug are you trying to report here?
Comment 11 Hongqing Yang 2011-03-09 23:24:45 EST
Created attachment 483356 [details]
kickstart from hard drive

the sad1 partition is mounted in the log, but the console also outputs error as the attachment. The kick start file cannot be found.
Comment 12 He Rui 2011-03-09 23:29:26 EST
(In reply to comment #9)
> (In reply to comment #5)
> > Created attachment 480071 [details]
> > syslog
> 
> Here sda and its partitions get discovered:
> 09:03:53,944 INFO kernel:[    5.122413]  sda: sda1 sda2 sda3 sda4 < sda5 sda6 >
> 
> And here sda1 gets mounted:
> 09:06:24,322 INFO kernel:[  155.790703] EXT3-fs (sda1): mounted filesystem with
> ordered data mode
> 09:06:24,322 DEBUG kernel:[  155.791610] SELinux: initialized (dev sda1, type
> ext3), uses xattr

When this issue happened, a window would pop up saying unable to find the kickstart file. But I couldn't access to ttys until I clicked cancel on the window to get stage2 started. So I thought sda1 was mounted after I cancelled kickstart install.
Comment 13 James Laska 2011-03-11 12:53:54 EST
Hongqing, when that dialog appears, and no tty's are available, you can press <Ctrl>Z to drop into a shell.  The shell environment will be limited, but you can do some debugging there.
Comment 14 Adam Williamson 2011-03-11 13:06:26 EST
Discussed at 2011-03-11 blocker review meeting. On the face of it, this hits criterion "The installer must be able to use all kickstart delivery methods ", but it may be more a case of a problem with _device discovery_ than a generic issue with the hard disk ks delivery method.

Hongqing, can you give us a few more details on the exact disk at issue here? Is it a USB flash stick or an actual hard disk connected externally? How is it formatted / partitioned?

Can Hongqing (or anyone else) please test with a different USB device, maybe a FAT-formatted USB flash stick?
Comment 15 Hongqing Yang 2011-03-14 04:49:21 EDT
Hi, Adam,

I tested it again, but it is weird, I cannot reproduce it again as I did before. It also not always happened before. I keep the kick start file on a guest hard disk, and install another guest from cdrom. it works now. so I prefer to close it first as it is not a bug.

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