I am in situation where I have servers that do not use the SLES 9 stock and vetted kernel, they use a later 2.6 kernel from kernel.org. The server exhibits behavior of instability when booting because the kernel crashes mysteriously and there are ext3 boot-up errors as well. Keep in mind this is a production environment that someone put into jeopardy by compiling a pristine kernel without knowing all the facts of what was needed by the OS and it applications. I just would like to know who else practices this in a production environment. And why?? Thanks, LDB -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thu, 2008-07-03 at 22:02 -0400, LDB wrote:
I am in situation where I have servers that do not use the SLES 9 stock and vetted kernel, they use a later 2.6 kernel from kernel.org. The server exhibits behavior of instability when booting because the kernel crashes mysteriously and there are ext3 boot-up errors as well. Keep in mind this is a production environment that someone put into jeopardy by compiling a pristine kernel without knowing all the facts of what was needed by the OS and it applications.
I just would like to know who else practices this in a production environment. And why??
Thanks,
LDB
I would think the only reason to recompile a kernel now a days is for some legacy support of a really old piece of hardware or for off-the-wall hardware. But if you're going to run a production server, you should probably go with well documented/supported hardware. I'm sure there are other reasons to do it, though. Just my $.02 -- Michael S. Dunsavage -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Michael S. Dunsavage wrote:
I would think the only reason to recompile a kernel now a days is for some legacy support of a really old piece of hardware or for off-the-wall hardware. But if you're going to run a production server, I might not be running a server but I am recompiling the kernel on my son's computer right now in order to correct the timer resolution problem for Rosegarden. At work I am using SLES 10 out of the box for a small lab setup. For that I don't have an reason to recompile.
Damon Register -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Sunday 2008-07-06 at 06:05 -0400, Damon Register wrote:
I might not be running a server but I am recompiling the kernel on my son's computer right now in order to correct the timer resolution problem for Rosegarden.
I'm curious on that. Could you expand a bit, please? - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFIcJ4LtTMYHG2NR9URAlwNAJ41AvbuAkgyExUWkzE3/3IoiRRGYgCfTzsP Vp68OyJPj7/DGOETWl/0xZs= =CCwv -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Thursday 2008-07-03 at 22:02 -0400, LDB wrote:
I am in situation where I have servers that do not use the SLES 9 stock and vetted kernel, they use a later 2.6 kernel from kernel.org. The server exhibits behavior of instability when booting because the kernel crashes mysteriously and there are ext3 boot-up errors as well. Keep in mind this is a production environment that someone put into jeopardy by compiling a pristine kernel without knowing all the facts of what was needed by the OS and it applications.
I just would like to know who else practices this in a production environment. And why??
Possibly something was not working and the new kernel solved that. Like some unsupported hardware, or some feature needed for another app. Whoever did that should have documented it. You should be able to put the original kernel back easily, but without knowing what it is that (possibly) did not work... - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFIbrvqtTMYHG2NR9URAn/RAJ9oJtmUKitS6HRCELY4neYePRvgHgCfWgVb iVkCZZIipRzVwBm4YgVuMeI= =TKOd -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Carlos E. R. wrote:
The Thursday 2008-07-03 at 22:02 -0400, LDB wrote:
I am in situation where I have servers that do not use the SLES 9 stock and vetted kernel, they use a later 2.6 kernel from kernel.org. The server exhibits behavior of instability when booting because the kernel crashes mysteriously and there are ext3 boot-up errors as well. Keep in mind this is a production environment that someone put into jeopardy by compiling a pristine kernel without knowing all the facts of what was needed by the OS and it applications.
I just would like to know who else practices this in a production environment. And why??
Possibly something was not working and the new kernel solved that. Like some unsupported hardware, or some feature needed for another app. Whoever did that should have documented it.
You should be able to put the original kernel back easily, but without knowing what it is that (possibly) did not work...
-- Cheers, Carlos E. R.
I wish that the case. :) It was mostly HP and Penguin hardware and nothing special in hardware to drive. :)) The excuse was, and I quote, "It was the latest stable 2.6 kernel". This guy just customized EVERYTHING -without documentation - for no damn reason. Sorry if I sound bitter, but you would not believe the unjustified customizations. :) LDB -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2008-07-04 at 23:37 -0400, LDB wrote:
I wish that the case. :)
:-}
It was mostly HP and Penguin hardware and nothing special in hardware to drive. :))
The excuse was, and I quote, "It was the latest stable 2.6 kernel". This guy just customized EVERYTHING -without documentation - for no damn reason. Sorry if I sound bitter, but you would not believe the unjustified customizations. :)
Arf! I can understand that on a personal computer, but I'd never experiment on a production server with sles, that costs a packet. You pay to be safe... only if something doesn't work or some new feature is needed you can justify the experiment. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD4DBQFIb3b6tTMYHG2NR9URAoFqAJjNh7M5jpLeInKc15cIpEEOIOokAJsFMB4W wLtYjnFIoRpD4Kb2GKJ44Q== =ntzU -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (5)
-
Carlos E. R. -
Damon Register -
LDB -
LDB -
Michael S. Dunsavage