Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.
For bugs related to Red Hat Enterprise Linux 5 product line. The current stable release is 5.10. For Red Hat Enterprise Linux 6 and above, please visit Red Hat JIRA https://issues.redhat.com/secure/CreateIssue!default.jspa?pid=12332745 to report new issues.

Bug 432421

Summary: [Stratus 5.3 feat] kdump: add support to mkdumprd to support md targets
Product: Red Hat Enterprise Linux 5 Reporter: Panagiotes M. Petrakis <peter.petrakis>
Component: kexec-toolsAssignee: Neil Horman <nhorman>
Status: CLOSED DUPLICATE QA Contact:
Severity: medium Docs Contact:
Priority: medium    
Version: 5.3CC: andriusb, chas.horvath, smcgrath
Target Milestone: rcKeywords: FutureFeature, OtherQA
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Enhancement
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2008-02-12 16:08:46 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 432518    
Attachments:
Description Flags
PATCH: Adds support for dumping to an MD device none

Description Panagiotes M. Petrakis 2008-02-11 22:13:54 UTC 
Description of problem:
1. Currently mkdumprd support for dumping to an md is incomplete

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


How reproducible:
Always

Steps to Reproduce:
1. Crash a kernel with the dump destination being an md
2. The proper raid modules are not loaded in the kexec'd kernel
  
Actual results:
init fails to mount file system, gives up, and reboots the system.

Expected results:
The target md is mounted and the dump is written to the target directory.

Additional info:
Attached is a patch which augments mkdumprd to detect that the md exists on the
system (there was bug with the test) and start the md's. It requires the
addition of the nash binary.
Comment 1 Panagiotes M. Petrakis 2008-02-11 22:13:54 UTC 
Created attachment 294603 [details]
PATCH: Adds support for dumping to an MD device
Comment 2 Neil Horman 2008-02-12 16:08:46 UTC 
I think this is a dupe of bz 317231, which is fixed in kexec-tools-1.102pre-10.el5

*** This bug has been marked as a duplicate of 317231 ***
Comment 3 Andrius Benokraitis 2008-02-12 16:12:54 UTC 
Panagiotes,

I spoke with Neil and he asked that you try the latest package posted on his
people page:

http://people.redhat.com/nhorman/rpms/latest_kexec-tools.tbz2

This package will be included in RHEL 5.2 Beta.

This looks like a dupe of bug 317231.
Comment 4 Panagiotes M. Petrakis 2008-02-12 18:53:27 UTC 
Hi,

I verified that this does the job but it seems to introduced a new bug...

++ echo ro root=/dev/md2 nmi_watchdog=0 console=tty0 console=ttyS0,115200
nosoftlockup piix.intel_via_libata=1 crashkernel=64M@16M
++ sed -e 's/crashkernel=[0-9]\+[MmKkGg]@[0-9]\+[MmGgKk]//'
+ KDUMP_COMMANDLINE='ro root=/dev/md2 nmi_watchdog=0 console=tty0
console=ttyS0,115200 nosoftlockup piix.intel_via_libata=1 '
+ KDUMP_COMMANDLINE='ro root=/dev/md2 nmi_watchdog=0 console=tty0
console=ttyS0,115200 nosoftlockup piix.intel_via_libata=1  irqpoll maxcpus=1
reset_devices'
++ /sbin/kexec --args-linux -p '--command-line=ro root=/dev/md2 nmi_watchdog=0
console=tty0 console=ttyS0,115200 nosoftlockup piix.intel_via_libata=1  irqpoll
maxcpus=1 reset_devices' --initrd=/boot/initrd-2.6.18-78.el5kdump.img
/boot/vmlinuz-2.6.18-78.el5
+ KEXEC_OUTPUT='Command line overflow'
+ '[' 1 == 0 ']'
+ /usr/bin/logger -p info -t kdump Command line overflow                       
             


Once I trimmed down the kernel args some it was fine and built the initrd,
though the same set of arguments worked fine before.

We can close this bug as a duplicate. Thank you for your help!
Comment 5 Neil Horman 2008-02-12 19:48:53 UTC 
Thats not a 'new' bug per-se.  Its been plaguing kexec for some time actually. 
I've got a fix pending upstream.  As soon as its fixed I'll be backporting it to
RHEL5.  Thanks!
Comment 6 Andrius Benokraitis 2008-02-12 20:36:47 UTC 
Neil - thanks for the info - is there a BZ placeholder for possibly the RHEL 5
back port?
Comment 7 Neil Horman 2008-02-12 21:14:36 UTC 
For the command line overflow bug?  Yeah, its bz 428310