openvz and qemu

Hi all,

i want to run a few qemu VMs inside VE0. Just wonder how to set
fairsched properties 
for a single process?

	- howto create a new faisched group (for the qemu process)?
	- howto assign weight/rate

or is that not possible at all?

- Dietmar

AW: openvz and qemu

OK, basically found the syscall interface in kernel/fairsched.c

But how do I find out the syscall numbers, i.e how do I 
call sys_fairsched_mknod() with userspace syscall()

- Dietmar

> -----Ursprüngliche Nachricht-----
> Von: users-bounces@... 
> [mailto:users-bounces@...] Im Auftrag von Dietmar Maurer
> Gesendet: Freitag, 01. Februar 2008 07:05
> An: users@...
> Betreff: [Users] openvz and qemu
> 
> Hi all,
> 
> i want to run a few qemu VMs inside VE0. Just wonder how to 
> set fairsched properties for a single process?
> 
> 	- howto create a new faisched group (for the qemu process)?
> 	- howto assign weight/rate
> 
> or is that not possible at all?
> 

Re: AW: openvz and qemu

check vzctl sources. it calls other OVZ-specific syscalls by their numbers,
like sys_setluid, sys_setublimit etc.

For creating fairsched node you'll need to call fairsched_mknod first, then move
the process to this node using fairsched_mvpr

Dietmar Maurer wrote:
> OK, basically found the syscall interface in kernel/fairsched.c
> 
> But how do I find out the syscall numbers, i.e how do I 
> call sys_fairsched_mknod() with userspace syscall()
> 
> - Dietmar
> 
>> -----Ursprüngliche Nachricht-----
>> Von: users-bounces@... 
>> [mailto:users-bounces@...] Im Auftrag von Dietmar Maurer
>> Gesendet: Freitag, 01. Februar 2008 07:05
>> An: users@...
>> Betreff: [Users] openvz and qemu
>>
>> Hi all,
>>
>> i want to run a few qemu VMs inside VE0. Just wonder how to 
>> set fairsched properties for a single process?
>>
>> 	- howto create a new faisched group (for the qemu process)?
>> 	- howto assign weight/rate
>>
>> or is that not possible at all?

Re: Problem with TCP window too large for TCPRCVBUF still present

Hi Vitaliy,

[ Cc'ing to Kirill, who sent the original patch ]

On Tue, Jan 08, 2008 at 11:12:49AM +0300, Vitaliy Gusev wrote:
> On 11 October 2007 21:46:05 Marcin Owsiany wrote:
> > Trying to debug a problem with stalling connections I found this thread:
> > http://forum.openvz.org/index.php?t=msg&goto=9018
> > which describes the exact problems I'm still having with 2.6.18-028.18
> >
> > Has it been fixed in this version? Or do you still need someone to test
> > this patch?
> 
> Yes, this patch still requires testing.

Can I have the patch as an attachment? The forum mangled the formatting
and patch refuses to apply it..

--

-- 
Marcin Owsiany <marcin@...>              http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216

"Every program in development at MIT expands until it can read mail."
                                                              -- Unknown

Re: Problem with TCP window too large for TCPRCVBUF still present

try this one attached plz.

Marcin Owsiany wrote:
> Hi Vitaliy,
> 
> [ Cc'ing to Kirill, who sent the original patch ]
> 
> On Tue, Jan 08, 2008 at 11:12:49AM +0300, Vitaliy Gusev wrote:
>> On 11 October 2007 21:46:05 Marcin Owsiany wrote:
>>> Trying to debug a problem with stalling connections I found this thread:
>>> http://forum.openvz.org/index.php?t=msg&goto=9018
>>> which describes the exact problems I'm still having with 2.6.18-028.18
>>>
>>> Has it been fixed in this version? Or do you still need someone to test
>>> this patch?
>> Yes, this patch still requires testing.
> 
> Can I have the patch as an attachment? The forum mangled the formatting
> and patch refuses to apply it..
> 

--- ./diff.ve7777	2008-02-05 11:25:37.000000000 +0300
+++ ./diff	2008-02-05 11:32:56.000000000 +0300
 <at>  <at>  -1,131 +0,0  <at>  <at> 
-diff --git a/include/net/tcp.h b/include/net/tcp.h
-index 0ff49a5..7e8f200 100644
---- a/include/net/tcp.h
-+++ b/include/net/tcp.h
- <at>  <at>  -1815,8 +1815,9  <at>  <at>  static inline int tcp_win_from_space(int

git 2.6.24 compilation

I have struck a few problems compiling latest git (as of a few hours
ago). Some of the problems are due to the sanitization of kernel
headers, where it seems some openvz definitions have escaped out of
the #ifdef __KERNEL__ clause. Patch for those:

--- linux-2.6.24-openvz/include/linux/capability.h.old	2008-02-07 14:51:17.000000000 +0100
+++ linux-2.6.24-openvz/include/linux/capability.h	2008-02-07 16:57:20.000000000 +0100
 <at>  <at>  -15,7 +15,10  <at>  <at> 

 #include <linux/types.h>
 #include <linux/compiler.h>
+
+#ifdef __KERNEL__
 #include <linux/spinlock.h>
+#endif

 struct task_struct;

--- linux-2.6.24-openvz/include/linux/if_tun.h.old	2008-02-07 14:51:17.000000000 +0100
+++ linux-2.6.24-openvz/include/linux/if_tun.h	2008-02-07 17:16:16.000000000 +0100
 <at>  <at>  -18,15 +18,15  <at>  <at> 
 #ifndef __IF_TUN_H
 #define __IF_TUN_H

-#include <linux/skbuff.h>
-#include <linux/if_ether.h>
-#include <linux/netdevice.h>
-
 /* Uncomment to enable debugging */
 /* #define TUN_DEBUG 1 */

Re: git 2.6.24 compilation

> ERROR: "sys_open" [kernel/cpt/vzrst.ko] undefined!
> ERROR: "sys_open" [kernel/cpt/vzcpt.ko] undefined!
> make[1]: *** [__modpost] Error 1
> make: *** [modules] Error 2
> + exit 1

this seems to have been fixed recently, and now checkpointing
compiles on amd64 (doesn't seem to work yet though:)

venom:/home/eyck# vzctl chkpnt 32
Setting up checkpoint...
	suspend...
	dump...
Can not dump VE: Invalid argument
Error: unknown restart block ffffffff802b0f9a
Checkpointing failed
and:
CPT ERR: ffff81006ed18000,101 :unknown restart block ffffffff802b0f9a
CPT ERR: ffff81006ed19800,32 :unknown restart block ffffffff802b0f9a
CPT ERR: ffff810070d20800,64 :unknown restart block ffffffff802b0f9a

--

-- 
Key fingerprint = 40D0 9FFB 9939 7320 8294  05E0 BCC7 02C4 75CC 50D9
 Total Existance Failure

problems wit 2.6.24 from git

I just downloaded the latest version from jit with:

git clone git://git.openvz.org/pub/linux-2.6.24-openvz
linux-2.6.24-openvz.org

(hope that is correct)

configuration, compilation worked without problem (compiling for amd64).

but when i try to start a VE I get the following error:

any idea whats wrong?

> vzctl start 101

sysfs: duplicate filename '0' can not be created
WARNING: at fs/sysfs/dir.c:424 sysfs_add_one()
Pid: 3251, comm: vzctl Tainted:  GF       2.6.24 #1

Call Trace:
 [<ffffffff802ea26c>] sysfs_add_one+0x54/0xbd
 [<ffffffff802ea7a7>] create_dir+0x4f/0x87
 [<ffffffff802ea961>] sysfs_create_dir+0x47/0x5c
 [<ffffffff8032474d>] kobject_get+0x12/0x17
 [<ffffffff80324c7a>] kobject_add+0xf3/0x1a6
 [<ffffffff80240c47>] user_kobject_create+0x77/0xa7
 [<ffffffff80241090>] alloc_uid+0xe1/0x185
 [<ffffffff802411b7>] copy_user_ns+0x71/0xde
 [<ffffffff8024ecf7>] create_new_namespaces+0x103/0x1da
 [<ffffffff8024efc1>] copy_namespaces+0x31/0x90

2.6.24 and fairsched syscalls

I thought vzctl uses fairsched syscalls to set cpu resources - but I
cant find then in the kernel (2.6.24 from git) - what do i miss?

- Dietmar

AW: AW: openvz and qemu

> check vzctl sources. it calls other OVZ-specific syscalls by 
> their numbers, like sys_setluid, sys_setublimit etc.
> 
> For creating fairsched node you'll need to call 
> fairsched_mknod first, then move the process to this node 
> using fairsched_mvpr

OK, I wrote a small (perl) script to test those syscalls, and everything
work as expected, for example:

..
fairsched_mknod (0, 500, 400);
fairsched_vcpus (400, 1);
set_cpulimit (400, 25);
fairsched_mvpr ($$, 400);
fork_and_do_something(); # runs with 25% cpu
wait
fairsched_rmnod (400);
..

work perfectly.

But when I call fairsched_rmnod(400) from another process while there is
still a process running inside 400 the system freezes. I get a kernel
panic somewhere inside move_task_off_dead_cpu.

any idea how to fix that?

- Dietmar