Dmitry Mishin | 1 Jun 06:41 2006
Picon

New features suggestion

On Wednesday 31 May 2006 18:18, Martin Dobrev wrote:
> Hi  <at> ll,
>
> I was thinking about some kind of kernel/userspace extension to the
> OpenVZ project. Actually the full list of processes on the hostnode is a
> bit disturbing and could lead to some bad things like killing a VPS
> process by accident. It's almost the same with netstat - now we see all
> open ports from the hostnode and the VPSes in the same list. To avoid
> such situations I suggest the following:
>
> 1. To patch the kernel in order to have the possibility to filter easy
> hostnode's and VPS'es processes.
> 2. To patch some userspace programs, just to list the VPS related stuff.
>
> To be clear I'll give the following example with netstat. The program
> will still list the port stats, but _only_ from the hostnode. A new
> program, called vnetstat will list the ports, that are handled by the
> VPSes. In the mean time a new column can show the VPS ID too, just to
> have the possibility to easy find the corresponding server. The same
> idea can be used for ps too.
> Actually there is no need to separate the things in different programs.
> Let's say we'll patch the main program and include the VPS ID column in it.
>
> What do you thing about this? Any comments and suggestions are appreciated!
There was similar development some time ago, check 
http://download.openvz.org/contrib/
It will be great, if you'll port it to current procps version. After that, 
we'll include it to standard OpenVZ package set, as well, as proposed 
vnetstat.

(Continue reading)

Kirill Korotaev | 1 Jun 07:08 2006
Picon

New features suggestion

>>What do you thing about this? Any comments and suggestions are appreciated!
> 
> There was similar development some time ago, check 
> http://download.openvz.org/contrib/
> It will be great, if you'll port it to current procps version. After that, 
> we'll include it to standard OpenVZ package set, as well, as proposed 
> vnetstat.

I will briefly add what this package do:
it adds -E option to both ps and top utils.
when -E is given it starts to print VEID column.
If number is given after -E options, then only processes for this VE are 
filtered/shown.
This is rather outdated, so would be nice to update this package...

Thanks,
Kirill

Kirill Korotaev | 2 Jun 08:52 2006
Picon

New features suggestion

BTW, I also uploaded a couple of useful tools to 
http://download.openvz.org/contrib/utils/ which allow to use fair CPU 
scheduler or UBC separately from virtualization. For example, it is 
possible to limit users.

Would be nice, if someone took a look and cleanup these tools.

Kirill

> Hi  <at> ll,
> 
> I was thinking about some kind of kernel/userspace extension to the
> OpenVZ project. Actually the full list of processes on the hostnode is a
> bit disturbing and could lead to some bad things like killing a VPS
> process by accident. It's almost the same with netstat - now we see all
> open ports from the hostnode and the VPSes in the same list. To avoid
> such situations I suggest the following:
> 
> 1. To patch the kernel in order to have the possibility to filter easy
> hostnode's and VPS'es processes.
> 2. To patch some userspace programs, just to list the VPS related stuff.
> 
> To be clear I'll give the following example with netstat. The program
> will still list the port stats, but _only_ from the hostnode. A new
> program, called vnetstat will list the ports, that are handled by the
> VPSes. In the mean time a new column can show the VPS ID too, just to
> have the possibility to easy find the corresponding server. The same
> idea can be used for ps too.
> Actually there is no need to separate the things in different programs.
> Let's say we'll patch the main program and include the VPS ID column in it.
(Continue reading)

Roberto Pereyra | 2 Jun 13:05 2006
Picon

wireless bridge and orinoco card with Openvz

Hi

I newbie with Openvz and would like to try it.

I have a Orinoco pcmcia wireless card with a pcmcia PCI bridge (Yenta).

This card works only (for me) with Slackware or Ubuntu. The others
distributions not wooks with this pcmcia pc bridge.

A question : If I install Openvz in a CentOS server then  the guest
linux (ie. Slackware or Ubuntu) can access and use the orinoco card ?

With Openvz I have access to pci devices like Orinoco card and bridge ?

Thanks in advance.

roberto

--

-- 
Ing. Roberto Pereyra
ContenidosOnline
Buscas un buen servicio de telefon?a IP, prueba
http://www.contenidosonline.com.ar

Cliff Wells | 3 Jun 04:00 2006

Noob questions

Hi,

I've decided to try out OpenVZ on my home system prior to using it on
some servers.  My plan was to continue to use my normal desktop but run
some services (httpd at least) in a VE.

I ran into a couple issues, one that I have a fix for, and one that
perhaps someone could shed some light on for me.

To start, XFS support was not compiled in, therefore I rebuilt from the
src.rpm and added this (just a note).

General system info:
FC4 (updated from the usual suspects: freshrpms, dag, etc)
Athlon 1700+ (kernel built for i686)
XFS filesystem on aic7xxx
NVidia binary driver (dual head using twinview, xinerama off)
ovzkernel-2.6.16-026test012.1 (rebuilt from src rpm)

2)  I'm using the binary NVidia driver.  It was initially unable to
build because it couldn't locate Makefile.cpu.  This was easily resolved
by copying this file
from /usr/src/redhat/BUILD/ovzkernel-2.6.16/linux-2.6.16/arch/i386/Makefile.cpu 
to /lib/modules/2.6.16-026test012.1/build/arch/i386/

3) After successfully getting X to start, I am unable to use the mouse.
It moves only along a vertical axis in the center of the screen (this is
the issue I could use help with). The mouse in question is a Logitech
cordless USB mouse.

(Continue reading)

Vasily Tarasov | 5 Jun 03:46 2006
Picon

wireless bridge and orinoco card with Openvz

At the moment support of these devices is not compiled in openVZ kernels.
However, you can modify config file appropriately and recompile kernel with
support of devices in question.

HTH,
Vasily.

Roberto Pereyra wrote:

> Hi
>
> I newbie with Openvz and would like to try it.
>
> I have a Orinoco pcmcia wireless card with a pcmcia PCI bridge (Yenta).
>
> This card works only (for me) with Slackware or Ubuntu. The others
> distributions not wooks with this pcmcia pc bridge.
>
> A question : If I install Openvz in a CentOS server then  the guest
> linux (ie. Slackware or Ubuntu) can access and use the orinoco card ?
>
> With Openvz I have access to pci devices like Orinoco card and bridge ?
>
> Thanks in advance.
>
> roberto
>
>
>

(Continue reading)

Vasily Tarasov | 5 Jun 03:59 2006
Picon

Noob questions

Cliff Wells wrote:

>Hi,
>
>I've decided to try out OpenVZ on my home system prior to using it on
>some servers.  My plan was to continue to use my normal desktop but run
>some services (httpd at least) in a VE.
>
>I ran into a couple issues, one that I have a fix for, and one that
>perhaps someone could shed some light on for me.
>
>To start, XFS support was not compiled in, therefore I rebuilt from the
>src.rpm and added this (just a note).
>
>General system info:
>FC4 (updated from the usual suspects: freshrpms, dag, etc)
>Athlon 1700+ (kernel built for i686)
>XFS filesystem on aic7xxx
>NVidia binary driver (dual head using twinview, xinerama off)
>ovzkernel-2.6.16-026test012.1 (rebuilt from src rpm)
>
>2)  I'm using the binary NVidia driver.  It was initially unable to
>build because it couldn't locate Makefile.cpu.  This was easily resolved
>by copying this file
>from /usr/src/redhat/BUILD/ovzkernel-2.6.16/linux-2.6.16/arch/i386/Makefile.cpu 
>to /lib/modules/2.6.16-026test012.1/build/arch/i386/
>
>3) After successfully getting X to start, I am unable to use the mouse.
>It moves only along a vertical axis in the center of the screen (this is
>the issue I could use help with). The mouse in question is a Logitech
(Continue reading)

Kirill Korotaev | 5 Jun 05:25 2006
Picon

Noob questions

Also check, that you have CONFIG_INPUT_EVDEV option enabled.

Kirill

> Cliff Wells wrote:
> 
>> Hi,
>>
>> I've decided to try out OpenVZ on my home system prior to using it on
>> some servers.  My plan was to continue to use my normal desktop but run
>> some services (httpd at least) in a VE.
>>
>> I ran into a couple issues, one that I have a fix for, and one that
>> perhaps someone could shed some light on for me.
>>
>> To start, XFS support was not compiled in, therefore I rebuilt from the
>> src.rpm and added this (just a note).
>>
>> General system info:
>> FC4 (updated from the usual suspects: freshrpms, dag, etc)
>> Athlon 1700+ (kernel built for i686)
>> XFS filesystem on aic7xxx
>> NVidia binary driver (dual head using twinview, xinerama off)
>> ovzkernel-2.6.16-026test012.1 (rebuilt from src rpm)
>>
>> 2)  I'm using the binary NVidia driver.  It was initially unable to
>> build because it couldn't locate Makefile.cpu.  This was easily resolved
>> by copying this file
>> from 
>> /usr/src/redhat/BUILD/ovzkernel-2.6.16/linux-2.6.16/arch/i386/Makefile.cpu 
(Continue reading)

Pradeep Padala | 6 Jun 20:06 2006
Picon

Measuring and Adjusting CPU utilization

Hi,

I am trying to measure the CPU utlilization of the VZ containers, and change
the cpu share dynamically. I have poured over most of the documentation, and
looked at the code as well, and it seems like there's no utility that can
directly show the current CPU utilization of a container (some thing like
30% of CPU). A search on the user list got me a message, where someone
suggested using loadavg. However, It seems like the loadavg is not showing
the proper utilization (or showing the total CPU utilization). This is what
I am doing.

I setup a container with 1000 units limit (total CPU units: ~10000). I wrote
a small do {; }while(1); loop and ran it in the container, now I do cat
/proc/loadavg in both the container and on the host node. Since, the
container is only using 1000 units, I should see something like 100% loadavg
in the container, and 10% loadavg in the hostnode. But, I see 100% at both
places. Am I doing something wrong? How do I get the current cpu utilization
of a container?

Thanks,
--

-- 
Pradeep Padala
http://ppadala.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://openvz.org/pipermail/users/attachments/20060606/07844d15/attachment.html

Kir Kolyshkin | 6 Jun 21:16 2006

Measuring and Adjusting CPU utilization

Looks like you misunderstand the concept of cpuunits. cpuunits is not a 
hard limit, but just a suggestion, and a CPU time is shared 
proportionally to the values given. So, if you will have 9 VEs and the 
host system with cpuunits set to 1000 for all of them, and run the loop 
in all of them, each VE will use 10% of the CPU time.

In case you will stop the loop running in 5 VEs so there will be 4 such 
VEs (plus the host system) left, each of them will use 20% of the CPU. 
So, all the CPU time is distributed between VEs which will need it, 
according with their proportional cpuunits.

More to say, the concept of "total CPU units" is purely fiction, and is 
here just for the convenience. People do want to set CPU units is terms 
of processor's megaherts, and this is what cpuunits does. But in fact it 
is not a megaherts but just a relative weights. I.e. all the cpuunits 
values are relative to each other, it doesn't matter what the actual 
numbers are -- what matters is a number given to a VE in relation to the 
sum of all cpuweights (which is expressed as "total CPU units" just for 
the convenience).

So, cpuunits, if you do not oversell them, are a CPU guarantee, not a 
limit. If you want CPU limit -- use cpulimit parameter.

Pradeep Padala wrote:

> Hi,
>
> I am trying to measure the CPU utlilization of the VZ containers, and 
> change the cpu share dynamically. I have poured over most of the 
> documentation, and looked at the code as well, and it seems like 
(Continue reading)


Gmane