9:46 a.m.
----- "Louis Bouchard" <louis.bouchard(a)hp.com> wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello,
> Is /usr/bin/crashdc itself a shell script?
>
> Dave
Yes, crashdc itself is a shell script, as well as the run-crashdc-*
which are intended to be invoked by the kdump mechanisms (kdump_post in
RHEL, KDUMP_POSTSCRIPT in SLES).
I'm guessing the run-*.sh scripts are easy to modify in order for
a user to tailor them to their needs?
Anyway, it sounds reasonable to me.
Can you work up a patch to the crash.spec file to create an additional
sub-package? Presume a new crashdc subdirectory off the top-level
to contain the files so that tar.gz users will get them as well.
Thanks,
Dave
9:55 a.m.
New subject: crashdc : automatic data collection for new vmcores in text format
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Dave Anderson a écrit :
I'm guessing the run-*.sh scripts are easy to modify in order for
a user to tailor them to their needs?
Anyway, it sounds reasonable to me.
Can you work up a patch to the crash.spec file to create an additional
sub-package? Presume a new crashdc subdirectory off the top-level
to contain the files so that tar.gz users will get them as well.
Thanks,
Dave
The idea that I'm following is to keep crashdc generic so it can be used
on the cmdline and used in an automated fashion. The run-*.sh are where
distro specific stuff gets put in (i.e. location of the dump directory,
debuginfo files etc). This is the script that is invoked by the kdump
mechanism.
Right now, it's working relatively well on RHEL5. SLES10 is not working
: crash just hangs after invoking gdb when executed from within the
kexec kernel. I will probably come back to you with some queries on that
later (with debugging info).
As for SLES11, the kdump mechanism has changed quite a lot so I must do
some more work on this, but I'm hoping to have something working soon.
As soon as I get something relatively stable, I'll propose a patch.
Thanks very much,
Kind Regards,
- --
Louis Bouchard, Linux Support Engineer
Team lead, EMEA Linux Competency Center,
Linux Ambassador, HP
HP Services 1 Ave du Canada
HP France Z.A. de Courtaboeuf
louis.bouchard(a)hp.com 91 947 Les Ulis
http://www.hp.com/go/linux France
http://www.hp.com/fr
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAkq82csACgkQDvqokHrhnCzbjQCfS59NXsq7GKLQFfawro8BS5HJ
0tMAoNBakShYQ61vIfwFZeoM43amus4b
=NUs2
-----END PGP SIGNATURE-----
10:25 a.m.
New subject: crashdc : automatic data collection for new vmcores in text format
On Fri, Sep 25 2009 at 10:55am -0400,
Louis Bouchard <louis.bouchard(a)hp.com> wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Dave Anderson a écrit :
>
> I'm guessing the run-*.sh scripts are easy to modify in order for
> a user to tailor them to their needs?
>
> Anyway, it sounds reasonable to me.
>
> Can you work up a patch to the crash.spec file to create an additional
> sub-package? Presume a new crashdc subdirectory off the top-level
> to contain the files so that tar.gz users will get them as well.
>
> Thanks,
> Dave
>
The idea that I'm following is to keep crashdc generic so it can be used
on the cmdline and used in an automated fashion. The run-*.sh are where
distro specific stuff gets put in (i.e. location of the dump directory,
debuginfo files etc). This is the script that is invoked by the kdump
mechanism.
Seems like you should have a config file also; so that each of the
run-*.sh don't contain path information, etc.. just the distro specific
mechanism.
make the configuration common (e.g. /etc/sysconfig/crashdc ?). Have
templates as you like for each distro. But config info does not belong
hidden in scripts in /usr/bin/
Mike
10:40 a.m.
New subject: crashdc : automatic data collection for new vmcores in text format
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello Mike,
Mike Snitzer a écrit :
Seems like you should have a config file also; so that each of the
run-*.sh don't contain path information, etc.. just the distro specific
mechanism.
make the configuration common (e.g. /etc/sysconfig/crashdc ?). Have
templates as you like for each distro. But config info does not belong
hidden in scripts in /usr/bin/
Mike
Interesting comment and I agree that those should fit in configuration
files.
I will also do my best to keep the run-*.sh scripts exempt from
configuration specific information. Actually, those scripts should not
be in /usr/bin at all. I would prefer to have them somewhere like in
/var/crash/scripts (/var/log/dump/scripts in SLES10).
Matter of fact, those run-*.sh scripts are there to deal with things like :
- In the kexec kernel, the namelist is /boot/vmlinux{version} in RHEL
and /root/boot/vmlinux{version} in SLES11 which mounts the root FS under
/root in the initrd realm.
- SLES needs to know where to find the .debug file which is part of the
debuginfo RPM. RHEL doesn't need that
- Differences in implementation of kdump post script mechanisms.
They're different everywhere (RHEL5, SLES10 and SLES11). I don't want
the user to have to bother with that. They should only have to add one
line to the /etc/kdump.conf|/etc/sysconfig/kdump file to have this working.
So thanks for your comments : they come in early enough in the
development so I can change things around easily.
Kind regards,
- --
Louis Bouchard, Linux Support Engineer
Team lead, EMEA Linux Competency Center,
Linux Ambassador, HP
HP Services 1 Ave du Canada
HP France Z.A. de Courtaboeuf
louis.bouchard(a)hp.com 91 947 Les Ulis
http://www.hp.com/go/linux France
http://www.hp.com/fr
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAkq85IgACgkQDvqokHrhnCwh3ACcDcTZjOkvhol1JXHcVj6Jmkip
KWYAoPWiC5vQ+e95W+U7cffFcx5y3RQH
=SwCt
-----END PGP SIGNATURE-----
5683
days inactive
5683
days old
devel@lists.crash-utility.osci.io
3 comments
3 participants
participants (3)
-
Dave Anderson -
Louis Bouchard -
Mike Snitzer