Bug 1049237

Summary: anaconda fail to fetch kickstart from cdrom
Product: Red Hat Enterprise Linux 7 Reporter: CongLi <coli>
Component: anacondaAssignee: Will Woods <wwoods>
Status: CLOSED CURRENTRELEASE QA Contact: Release Test Team <release-test-team-automation>
Severity: urgent Docs Contact:
Priority: urgent    
Version: 7.0CC: anaconda-maint-list, coli, jstodola, juzhang, mbanas, michen, mkovarik, shuang, wwoods, xhan, xutian, xwei
Target Milestone: rcKeywords: Regression
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: anaconda-19.31.58-1 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-06-13 13:16:11 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:
Attachments:
Description Flags
serial log
none
screenshot - dracut
none
cdrom.log
none
cdrom_ks.cfg.log
none
cdrom.log
none
cdrom.sr0.url.ks.aio_native.log
none
cdrom.sr0.url.ks.aio_threads.log none

Description CongLi 2014-01-07 09:24:17 UTC 
Created attachment 846524 [details]
serial log

Description of problem:
anaconda fail to fetch kickstart from cdrom

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

anaconda-19.31.41-1.el7.x86_64 shipped with RHEL-7.0-20131222.0

How reproducible:
Always

Steps to Reproduce:
1. install a rhel7.0 guest with tree RHEL-7.0-20131222.0:
   with option: -append 'ks=cdrom:/dev/sr1'
2.
3.

Actual results:
anaconda fail to fetch kickstart from cdrom then goes to dracut

serial log (attached):
2014-01-07 17:06:17: [   14.747992] scsi 0:0:0:0: Direct-Access     ATA      QEMU HARDDISK    1.5. PQ: 0 ANSI: 5
2014-01-07 17:06:17: [   14.749536] scsi 0:0:1:0: CD-ROM            QEMU     QEMU DVD-ROM     1.5. PQ: 0 ANSI: 5
2014-01-07 17:06:17: [   14.750971] scsi 1:0:0:0: CD-ROM            QEMU     QEMU DVD-ROM     1.5. PQ: 0 ANSI: 5
2014-01-07 17:06:17: [   14.872655] cirrus 0000:00:02.0: fb0: cirrusdrmfb frame buffer device
2014-01-07 17:06:17: [   14.876471] cirrus 0000:00:02.0: registered panic notifier
2014-01-07 17:06:17: [   14.886473] [drm] Initialized cirrus 1.0.0 20110418 for 0000:00:02.0 on minor 0
2014-01-07 17:06:17: [   14.934121] sr0: scsi3-mmc drive: 4x/4x cd/rw xa/form2 tray
2014-01-07 17:06:17: [   14.939663] cdrom: Uniform CD-ROM driver Revision: 3.20
2014-01-07 17:06:17: [   14.950821] sr1: scsi3-mmc drive: 4x/4x cd/rw xa/form2 tray
2014-01-07 17:06:17: [   15.063815] sd 0:0:0:0: [sda] 41943040 512-byte logical blocks: (21.4 GB/20.0 GiB)
2014-01-07 17:06:17: [   15.078503] sd 0:0:0:0: [sda] Write Protect is off
2014-01-07 17:06:17: [   15.087730] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
2014-01-07 17:06:17: [   15.121189]  sda: unknown partition table
2014-01-07 17:06:17: [   15.124510] sd 0:0:0:0: [sda] Attached SCSI disk

Expected results:
anaconda fetch kickstart from cdrom successfully

Additional info:
1. with option 'ks=cdrom:/dev/sr0:/ks.cfg' doesn't work either

2. with option 'ks=hd:sr1:/ks.cfg' works fine

3. Qemu CML:
/root/staf-kvm-devel/autotest-devel/client/tests/virt/qemu/qemu \
    -S  \
    -name 'virt-tests-vm1'  \
    -sandbox off  \
    -M pc  \
    -nodefaults  \
    -vga cirrus  \
    -chardev socket,id=qmp_id_qmpmonitor1,path=/tmp/monitor-qmpmonitor1-20140107-170550-4SULHXAP,server,nowait \
    -mon chardev=qmp_id_qmpmonitor1,mode=control  \
    -chardev socket,id=serial_id_serial0,path=/tmp/serial-serial0-20140107-170550-4SULHXAP,server,nowait \
    -device isa-serial,chardev=serial_id_serial0  \
    -chardev socket,id=seabioslog_id_20140107-170550-4SULHXAP,path=/tmp/seabios-20140107-170550-4SULHXAP,server,nowait \
    -device isa-debugcon,chardev=seabioslog_id_20140107-170550-4SULHXAP,iobase=0x402 \
    -device ich9-usb-uhci1,id=usb1,bus=pci.0,addr=03 \
    -drive id=drive_image1,if=none,cache=none,snapshot=off,aio=native,file=/root/staf-kvm-devel/autotest-devel/client/tests/virt/shared/data/images/RHEL-Server-7.0-64-virtio.qcow2 \
    -device ide-hd,id=image1,drive=drive_image1,bus=ide.0,unit=0 \
    -device virtio-net-pci,mac=9a:e0:e1:e2:e3:e4,id=id3m8m16,netdev=idYSR4Sn,bus=pci.0,addr=04  \
    -netdev tap,id=idYSR4Sn,vhost=on,vhostfd=23,fd=22  \
    -m 2048  \
    -smp 1,maxcpus=1,cores=1,threads=1,sockets=2  \
    -cpu 'Opteron_G2' \
    -drive id=drive_cd1,if=none,snapshot=off,aio=native,media=cdrom,file=/root/staf-kvm-devel/autotest-devel/client/tests/virt/shared/data/isos/linux/RHEL7.0-Server-x86_64.iso \
    -device ide-cd,id=cd1,drive=drive_cd1,bus=ide.0,unit=1 \
    -drive id=drive_unattended,if=none,snapshot=off,aio=native,media=cdrom,file=/root/staf-kvm-devel/autotest-devel/client/tests/virt/shared/data/images/rhel70-64/ks.iso \
    -device ide-cd,id=unattended,drive=drive_unattended,bus=ide.1,unit=0 \
    -device usb-tablet,id=usb-tablet1,bus=usb1.0,port=1  \
    -kernel '/root/staf-kvm-devel/autotest-devel/client/tests/virt/shared/data/images/rhel70-64/vmlinuz'  \
    -append 'ks=cdrom:/dev/sr1 nicdelay=60 console=ttyS0,115200 console=tty0'  \
    -initrd '/root/staf-kvm-devel/autotest-devel/client/tests/virt/shared/data/images/rhel70-64/initrd.img'  \
    -vnc :0  \
    -rtc base=utc,clock=host,driftfix=slew  \
    -boot order=cdn,once=d,menu=off  \
    -no-kvm-pit-reinjection \
    -no-shutdown \
    -enable-kvm
Comment 1 CongLi 2014-01-07 09:27:09 UTC 
Created attachment 846525 [details]
screenshot - dracut
Comment 3 CongLi 2014-01-07 10:11:33 UTC 
(In reply to CongLi from comment #0)
> 2. with option 'ks=hd:sr1:/ks.cfg' works fine

with option 'ks=hd:sr1:/ks.cfg', sometimes failed either.
Comment 4 CongLi 2014-01-07 10:20:14 UTC 
Installation of RHEL-7.0-20131123.0-Server-x86_64 (anaconda-19.31.36-1.el7.x86_64) works fine.
Comment 6 Xiaoqing Wei 2014-01-09 03:37:21 UTC 
Hello Anaconda guys,

Could you pls have a look at this issue ?

Thx.
Xiaoqing.
Comment 7 Will Woods 2014-01-09 20:50:29 UTC 
The option "ks=cdrom:/dev/sr1" is wrong.

In RHEL6 and RHEL7, ks=cdrom:... works like this:

* ks=cdrom

  Probe all cdrom devices and look for '/ks.cfg'

* ks=cdrom:<path>

  Probe all cdrom devices and look for <path>

(See documentation: https://access.redhat.com/site/documentation/en-US/Red_Hat_Enterprise_Linux/6/html/Installation_Guide/s1-kickstart2-startinginstall.html)

In RHEL7 you may also use:

* ks=cdrom:<dev>:<path>

  Look for <path> on the cdrom device <dev>.
  This is identical to "ks=hd:<dev>:<path>".

If you get the wrong device (e.g. the CDROM is in /dev/sr1 but you used "/dev/sr0"), this will fail.

Could you please try the following and tell me if they work?


  ks=cdrom
  ks=cdrom:/ks.cfg
  ks=cdrom:/dev/sr1:/ks.cfg
Comment 8 juzhang 2014-01-10 02:12:58 UTC 
Hi Cong,

Could you have a look comment7 and have a try?

Best Regards,
Junyi
Comment 10 CongLi 2014-01-10 07:18:17 UTC 
Created attachment 848073 [details]
QEMU CML
Comment 11 Will Woods 2014-01-10 21:28:19 UTC 
I've pushed a fix for this into the 'master' git branch; once the appropriate flags are set we can get this pulled into rhel7.
Comment 12 Will Woods 2014-01-24 18:21:03 UTC 
Commit f0adbcb pushed to rhel7-branch.
Comment 14 Martin Banas 2014-01-30 07:34:35 UTC 
Verified, that with:
ks=cdrom -> works
ks=cdrom:/dev/sr1:/ks.cfg -> works
ks=cdrom:/dev/sr1:/another_ks.cfg -> works
ks=cdrom:/another_ks.cfg -> works

ks=cdrom:/dev/sr0 -> doesn't work (expected)

Moving to VERIFIED.

Verified on:
RHEL-7.0-20140129.n.0:
anaconda-19.31.54-1.el7
Comment 15 xhan 2014-02-11 08:00:20 UTC 
(In reply to Martin Banas from comment #14)
> Verified, that with:
> ks=cdrom -> works

How many cdrom devices are there in your test environment? 
My test result is failed.
ks=cdrom -> failed
ks=cdrom:/ks.cfg ->failed.

Tested version:
anaconda-19.31.56-1.el7.x86_64.rpm (shipped with RHEL-7.0-20140206.0)
  
> ks=cdrom:/dev/sr1:/ks.cfg -> works
> ks=cdrom:/dev/sr1:/another_ks.cfg -> works
> ks=cdrom:/another_ks.cfg -> works
> 
> ks=cdrom:/dev/sr0 -> doesn't work (expected)
> 
> Moving to VERIFIED.
> 
> Verified on:
> RHEL-7.0-20140129.n.0:
> anaconda-19.31.54-1.el7
Comment 16 Martin Banas 2014-02-11 08:39:06 UTC 
Hi,
I had 2 cdroms attached.
Comment 17 Martin Banas 2014-02-11 08:59:22 UTC 
Could you please attach some log files? Probably best would be to boot with: rd.debug and gather the logs from the console. Please move the bug then back to ASSIGNED.

I tried it again now and I'm still not able to reproduce it.

Thanks!
Comment 18 xhan 2014-02-11 10:09:40 UTC 
Created attachment 861738 [details]
cdrom.log

The log is generated by ks=cdrom cmdline.
Comment 19 xhan 2014-02-11 10:11:05 UTC 
Created attachment 861740 [details]
cdrom_ks.cfg.log

This log file is generated by cmdline ks=cdrom:/ks.cfg.
Comment 20 xhan 2014-02-11 10:15:04 UTC 
I use the kernel commandline "-append 'ks= ... console=ttyS0,115200 rd.debug".
  ks=cdrom          ---> cdrom.log          
  ks=cdrom:/ks.cfg  ---> cdrom_ks.cfg.log 
   
Are these capture files enough? If not, let me your method to save log.
Comment 21 Will Woods 2014-02-11 20:48:52 UTC 
The problem with ks=cdrom:/ks.cfg is here:

  dracut-pre-trigger[416]: //lib/dracut/hooks/pre-trigger/50-kickstart-genrules.sh: line 16: when_any_cdrom_appears: command not found

A required commit didn't get pulled into rhel7-branch. It's now been pushed:

  https://git.fedorahosted.org/cgit/anaconda.git/commit/?h=rhel7-branch&id=70dc76d

So that should be fixed in the next anaconda build.

The same patch should also fix the "ks=cdrom" case, but I can't tell for sure, since the log seems to be incomplete. (there's no mention of parse-anaconda-kickstart.sh at all.)
Comment 22 xhan 2014-02-12 02:32:40 UTC 
Created attachment 862093 [details]
cdrom.log

attach the cdrom.log again.
Comment 23 xhan 2014-02-12 02:58:19 UTC 
Met the anaconda dracut problem with ks=cdrom:/dev/sr0:/ks.cfg which aio is assigned as native.

If being changed to aio=threads, the test passed.

Please refer the log files: cdrom.sr0.url.ks.aio_native.log and cdrom.sr0.url.ks.aio_threads.log
Comment 24 xhan 2014-02-12 02:59:02 UTC 
Created attachment 862096 [details]
cdrom.sr0.url.ks.aio_native.log
Comment 25 xhan 2014-02-12 02:59:29 UTC 
Created attachment 862097 [details]
cdrom.sr0.url.ks.aio_threads.log
Comment 27 Xu Tian 2014-02-12 03:05:58 UTC 
(In reply to xhan from comment #23)
> Met the anaconda dracut problem with ks=cdrom:/dev/sr0:/ks.cfg which aio is
> assigned as native.
> 
> If being changed to aio=threads, the test passed.
> 
> Please refer the log files: cdrom.sr0.url.ks.aio_native.log and
> cdrom.sr0.url.ks.aio_threads.log

Hi xhan,

It looks above issue is a bug of qemu-kvm component, it's better to open a new bug in qemu-kvm component to track above issue.

Thanks,
Xu
Comment 28 Jan Stodola 2014-04-08 14:09:12 UTC 
Retested with anaconda-19.31.75-1.el7 and following boot options:

ks=cdrom
ks=cdrom:/ks.cfg
ks=cdrom:/dev/sr1:/ks.cfg

All worked as expected, the when_any_cdrom_appears() function is present in usr/lib/anaconda-lib.sh of the installation initrd.

Moving to VERIFIED.
Comment 29 Ludek Smid 2014-06-13 13:16:11 UTC 
This request was resolved in Red Hat Enterprise Linux 7.0.

Contact your manager or support representative in case you have further questions about the request.