On Thu, 02 Dec 2004 20:01:48 -0800, Kenneth Porter shiva@sewingwitch.com wrote:
I've now got several instances of "[wine-preloader] <defunct>" in my ps listing from a game server that's died in some strange way. Anyone have experience in clearing these short of bouncing the whole box? They're holding onto a couple of TCP ports that I'd like to use. The ports are stuck in CLOSE_WAIT, because the processes are stuck in memory so the sockets can't close.
Does 'kill -9 processnum' not work?
On December 2, 2004 08:15 pm, James Hawkins wrote:
On Thu, 02 Dec 2004 20:01:48 -0800, Kenneth Porter
shiva@sewingwitch.com wrote:
I've now got several instances of "[wine-preloader] <defunct>" in my ps listing from a game server that's died in some strange way. Anyone have experience in clearing these short of bouncing the whole box? They're holding onto a couple of TCP ports that I'd like to use. The ports are stuck in CLOSE_WAIT, because the processes are stuck in memory so the sockets can't close.
Does 'kill -9 processnum' not work?
Probably not. I regularly face them and the only way I know is to reboot
On Thu, 2 Dec 2004 21:28:01 -0800, Bill Medland billmedland@mercuryspeed.com wrote:
On December 2, 2004 08:15 pm, James Hawkins wrote:
On Thu, 02 Dec 2004 20:01:48 -0800, Kenneth Porter
shiva@sewingwitch.com wrote:
I've now got several instances of "[wine-preloader] <defunct>" in my ps listing from a game server that's died in some strange way. Anyone have experience in clearing these short of bouncing the whole box? They're holding onto a couple of TCP ports that I'd like to use. The ports are stuck in CLOSE_WAIT, because the processes are stuck in memory so the sockets can't close.
Does 'kill -9 processnum' not work?
Probably not. I regularly face them and the only way I know is to reboot
-- Bill Medland mailto:billmedland@mercuryspeed.com http://webhome.idirect.com/~kbmed
How can kill -9 not work? It should kill the process without question unless I'm wrong.
Hi,
--- James Hawkins truiken@gmail.com wrote:
Probably not. I regularly face them and the only way I know is to reboot
How can kill -9 not work? It should kill the process without question unless I'm wrong.
I think this is a kernel bug that was fixed in more recent kernels.
Thanks Steven
__________________________________ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail
--On Friday, December 03, 2004 3:18 PM +0900 Mike McCormack mike@codeweavers.com wrote:
The bug present in Linux 2.6.[0-7] and is fixed in Linux 2.6.8 and later.
Running 2.6.8-1.521smp on Fedora Core 2.
Kenneth Porter wrote:
The bug present in Linux 2.6.[0-7] and is fixed in Linux 2.6.8 and later.
Running 2.6.8-1.521smp on Fedora Core 2.
When I say Linux 2.6.8 fixes the bug, I mean the standard kernel shipped by Linus. I've got no idea what Redhat does, and they tend to apply alot of patches to their kernel, so it wouldn't suprise me if it's still broken.
Mike
On December 2, 2004 10:43 pm, Mike McCormack wrote:
Kenneth Porter wrote:
The bug present in Linux 2.6.[0-7] and is fixed in Linux 2.6.8 and later.
Running 2.6.8-1.521smp on Fedora Core 2.
When I say Linux 2.6.8 fixes the bug, I mean the standard kernel shipped by Linus. I've got no idea what Redhat does, and they tend to apply alot of patches to their kernel, so it wouldn't suprise me if it's still broken.
Mike
Yep. RedHat Enterprise Linux 3 and RedHat Desktop 3 2.4.21-20 (heavily modified, as you said!!!) is still broken
Le ven 03/12/2004 à 10:15, Bill Medland a écrit :
On December 2, 2004 10:43 pm, Mike McCormack wrote:
Kenneth Porter wrote:
The bug present in Linux 2.6.[0-7] and is fixed in Linux 2.6.8 and later.
Running 2.6.8-1.521smp on Fedora Core 2.
When I say Linux 2.6.8 fixes the bug, I mean the standard kernel shipped by Linus. I've got no idea what Redhat does, and they tend to apply alot of patches to their kernel, so it wouldn't suprise me if it's still broken.
Mike
Yep. RedHat Enterprise Linux 3 and RedHat Desktop 3 2.4.21-20 (heavily modified, as you said!!!) is still broken
The fixed version is 2.6.8, earlier versions were problematic. I don't think a 2.4.21 variant can be called "still broken" since it predates the fix.
Vincent
On December 3, 2004 08:17 am, Vincent Béron wrote:
Le ven 03/12/2004 à 10:15, Bill Medland a écrit :
On December 2, 2004 10:43 pm, Mike McCormack wrote:
Kenneth Porter wrote:
The bug present in Linux 2.6.[0-7] and is fixed in Linux 2.6.8 and later.
Running 2.6.8-1.521smp on Fedora Core 2.
When I say Linux 2.6.8 fixes the bug, I mean the standard kernel shipped by Linus. I've got no idea what Redhat does, and they tend to apply alot of patches to their kernel, so it wouldn't suprise me if it's still broken.
Mike
Yep. RedHat Enterprise Linux 3 and RedHat Desktop 3 2.4.21-20 (heavily modified, as you said!!!) is still broken
The fixed version is 2.6.8, earlier versions were problematic. I don't think a 2.4.21 variant can be called "still broken" since it predates the fix.
Vincent
Yes, but that's the problem with Red Hat; just 'cause it says it's 2.4 doesn't really mean a thing. RHEL3 basically contains, as far as I understant it, a 2.4 kernel with a load of 2.6 stuff hacked into it. (I think RHEL4 is supposed to have a 2.6 kernel when it comes out)
--On Thursday, December 02, 2004 11:15 PM -0500 James Hawkins truiken@gmail.com wrote:
Does 'kill -9 processnum' not work?
Son of a gun, that did it. I'm so used to avoiding that that it didn't occur to me to try it.
-
Bill Medland -
James Hawkins -
Kenneth Porter -
Mike McCormack -
Steven Edwards -
Vincent Béron