Bug 619345

Summary:	RFE: Enable coredump capture in the appliance
Product:	Red Hat Enterprise Linux 6	Reporter:	Matthew Booth <mbooth>
Component:	libguestfs	Assignee:	Richard W.M. Jones <rjones>
Status:	CLOSED WONTFIX	QA Contact:	Virtualization Bugs <virt-bugs>
Severity:	medium	Docs Contact:
Priority:	low
Version:	6.2	CC:	mbooth, mshao, rjones, virt-maint
Target Milestone:	rc	Keywords:	FutureFeature
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Enhancement
Doc Text:		Story Points:	---
Clone Of:	619334	Environment:
Last Closed:	2011-06-12 18:24:48 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:	619334
Bug Blocks:

Description Matthew Booth 2010-07-29 10:24:50 UTC

+++ This bug was initially created as a clone of Bug #619334 +++

Description of problem:

It should be possible to enable coredump capture in the appliance.  This
would allow us to more easily debug certain problems without requiring
'printf' statements as now.

It would work something like this:

(1) Add a handle variable called 'coredump' /
  LIBGUESTFS_COREDUMP=<path>

(2) This sets a kernel parameter for the appliance.

(3) If the kernel parameter is set then the daemon enables coredumps:

  (a) echo /sysroot/<path>/core.%t.%e > /proc/sys/kernel/core_pattern
  (b) Set ulimit core unlimited

(4) After something coredumps (even the daemon) this ought to
  capture the coredump on the currently mounted device.

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

1.5.x

--- Additional comment from rjones on 2010-07-29 06:05:35 EDT ---

I wonder if we can coredump to a raw disk?  Then the user could
add a "coredump capture disk".

--- Additional comment from mbooth on 2010-07-29 06:15:30 EDT ---

Unless we plan to turn this on by default, this is really a developer tool. I think we can assume to a significant extent that the user understands exactly how it works and what the implications of using it are. This would include, for example, choosing a sensible value for LIBGUESTFS_COREDUMP for the current image file. I'd keep this one simple.

Comment 1 Matthew Booth 2010-07-29 10:25:52 UTC

Cloning this here because I think this will be an essential support tool.

Comment 2 Richard W.M. Jones 2010-11-24 09:06:12 UTC

Postponing to 6.2 since we have no idea how to implement this
upstream.

6.1 will include the unsupported debug core_pattern feature
which Marek is using.

Comment 3 Richard W.M. Jones 2011-03-17 10:58:00 UTC

I posted a patch to the kernel which works rather
nicely.  However at present it seems unlikely to
be accepted.

http://lkml.org/lkml/2011/3/16/201

Comment 4 Richard W.M. Jones 2011-03-24 17:17:11 UTC

Note to self on why using core_pattern is unlikely
to work for us:
https://rwmj.wordpress.com/2010/08/03/linux-core_pattern-fail/#content

Comment 6 Richard W.M. Jones 2011-06-12 18:06:28 UTC

After further "discussions" with kernel developers, they're
not interested in fixing the broken core_pattern interface.
Barring some major revolution, this can't be done.

Comment 7 RHEL Program Management 2011-06-12 18:24:48 UTC

Development Management has reviewed and declined this request.  You may appeal
this decision by reopening this request.