About this list Date view Thread view Subject view Author view Attachment view

From: Mario Lorenz (ml_at_vdazone.org)
Date: Thu 23 Jan 2003 - 11:35:00 GMT


Hello Alexey,

On Thu, Jan 23, 2003 at 08:48:34AM +0300, Lyashkov Alexey wrote:

> this RPM based on my version vserver patch.
> you can download lastes version her from my home page www.ttn.ru/~shadow/ and apply
> manually.
> if you have bugs send to me bug report and i fix this.
>

I have downloaded the latest version of your patch from ttn.ru, recompiled
the kernel and redone the tests. Same behaviour.

Now furhter investigation makes me wonder if this is a problem with your
patch or with the man page of chbind.

chbind supposedly locks only a process and its children into a specific
IP/Interface. It doesnt. It seems it locks the whole context (and in case
of context 0, thats not really useful).

To witness:
(freshly rebooted box, no vservers started)

# ip addr show
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
    link/ether 00:e0:7d:98:0b:8d brd ff:ff:ff:ff:ff:ff
    inet 192.168.123.125/24 brd 192.168.123.255 scope global eth0

# telnet 192.168.123.100
Trying 192.168.123.100...
Connected to 192.168.128.102.
[...]

(Everything working as expected. Now bring up secondary interface)

# /sbin/ifconfig eth0:1 192.168.123.126 netmask 255.255.255.255
# ip addr show
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
    link/ether 00:e0:7d:98:0b:8d brd ff:ff:ff:ff:ff:ff
    inet 192.168.123.125/24 brd 192.168.123.255 scope global eth0
    inet 192.168.123.126/32 brd 192.168.123.255 scope global eth0:1

(set up chbind, _EXIT_ children so there is no chbind anymore, then take
 down secondary IF)

# /usr/sbin/chbind --ip 192.168.123.126 /bin/bash
ipv4root is now 192.168.123.126
# exit
# /sbin/ifconfig eth0:1 down
# ip addr show
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
   link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
   link/ether 00:e0:7d:98:0b:8d brd ff:ff:ff:ff:ff:ff
   inet 192.168.123.125/24 brd 192.168.128.255 scope global eth0
            
# telnet 192.168.123.100
Trying 192.168.123.100...
telnet: connect to address 192.168.123.100: Invalid argument

Now what ?

# /usr/sbin/chbind --ip 0.0.0.0 /bin/bash
ipv4root is now 0.0.0.0
# exit
# telnet 192.168.123.100
Trying 192.168.123.100...
Connected to 192.168.128.102.
[...]

Thats vserver-0.22.

Who is wrong ? Me, the Kernel patch, the chbind manpage ? And how is this
best corrected ?

Best regards,

Mario


About this list Date view Thread view Subject view Author view Attachment view
[Next/Previous Months] [Main vserver Project Homepage] [Howto Subscribe/Unsubscribe] [Paul Sladen's vserver stuff]
Generated on Thu 23 Jan 2003 - 12:02:41 GMT by hypermail 2.1.3