Bug 215545

Summary: The kdump finishes properly but the vmcore is not complete.
Product: Red Hat Enterprise Linux 5 Reporter: Akira Imamura <aimamura>
Component: kexec-toolsAssignee: Neil Horman <nhorman>
Status: CLOSED CURRENTRELEASE QA Contact:
Severity: medium Docs Contact:
Priority: medium    
Version: 5.0CC: kkomiyam, ktokunag, ntachino
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: 5.0.0 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-02-13 17:27:54 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:
Attachments:
Description Flags
This patch is based on -131.el5 and can fix this problem. none

Description Akira Imamura 2006-11-14 16:30:31 UTC
Description of problem:
The kdump seems to complete properly but the vmcore created by kdump
is incomplete (the file name is vmcore-incomplete). I think the cause of
this problem is that file caches won't be written to the disk before kdump
reboots system. The kdump can create the vmcore completely by running
"umount" before running "reboot".


Version-Release number of selected component (if applicable):
kexec-tools-1.101-131.el5

How reproducible:
This problem occurs easier in the case that the reserved
memory for 2nd kernel is larger.


Steps to Reproduce:
1. Run "service kdump start".
2. Make system panic.
3. See a created vmcore file after rebooting.
  
Actual results:
-r-------- 1 root root 15943680000 Nov 13 20:51 vmcore-incomplete

Expected results:
-r--------    1 root     0       16341865252 Nov 13 20:51 vmcore



Additional info:

Comment 1 Akira Imamura 2006-11-14 16:30:32 UTC
Created attachment 141162 [details]
This patch is based on -131.el5 and can fix this problem.

Comment 2 Neil Horman 2006-11-14 16:55:17 UTC
you're patch is wrong, you still need that unmount in case something goes wrong
during the copy.  you don't want to leave a filesystem mounted if you switchroot
to the rootfs to run init.

I'll fix this and the nfs case shortly.

Comment 3 Neil Horman 2006-11-14 17:31:20 UTC
fixed in -132.el5. thanks

Comment 6 Jay Turner 2007-02-13 17:27:21 UTC
Closing out.