Ok, so that does resolve error message (/usr/sbin/vserver: virtual: 
command not found), but it appears my problem is larger.

I'm still seeing a failure to bind to the second IP if the first has 
already been bound.  For example, with Apache:

[Tue Jun 17 11:06:27 2003] [crit] (98)Address already in use: make_sock: 
could not bind to port 80

Same flavored message with AOLserver.  They're definitely trying to bind 
the second IP, and a netstat -ant shows ONLY the first IP already in use.  
Attempts to connect to the service on the second IP fail, so I don't think 
the service on the first IP actually is listening on both.

I stopped and restarted the vserver with no change in symptoms.

Thanks for any further ideas,

Cathy



On Tue, 17 Jun 2003, Cathy Sarisky wrote:

> 
> Absolutely spot-on, Herbert, thank you.  *Blush* I went through
> /usr/sbin/vserver with a fine-toothed comb looking for SOMETHING that only
> happened with multiple IPs, and missed the missing # in my config file.
> 
> On Tue, 17 Jun 2003, Herbert Poetzl wrote:
> 
> > On Tue, Jun 17, 2003 at 04:02:17AM -0400, Cathy Sarisky wrote:
> > > Hi list,
> > > 
> > > I have a vserver with 3 IPs assigned to it.
> > > 
> > > Starting or entering the vserver gives the following error message (in 
> > > triplicate):
> > > 
> > > /usr/sbin/vserver: virtual: command not found
> > 
> > # grep -rnI virtual /usr/sbin/vserver
> > 2:# This is a script to control a virtual server
> > 180:        echo " build   : Create a virtual server by copying the packages"
> > 182:        echo " enter   : Enter in the virtual server context and starts a 
> > shell"184:    echo " exec    : Exec a command in the virtual server context"
> > 185:        echo " suexec  : Exec a command in the virtual server context uid"
> > 190:        echo " running : Tells if a virtual server is running"
> > 247:                # Create a minimal dev so the virtual server can't grab
> > 299:# Select the IP number assigned to the virtual server
> > 332:# virtual server at boot time
> > 377:        echo Starting the virtual server $1
> > 500:        echo Stopping the virtual server $1
> > 
> > I guess the '#' from line 332 is missing, but do
> > not necessary look in /usr/sbin/vserver, because
> > line 332 is actually copied into each <name>.conf
> > in /etc/vservers ... probably you untentionally
> > deleted the '#' from there ...
> > 
> > hth,
> > Herbert
> > 
> > > It does set ipv4root correctly, and binding 0.0.0.0 results in it 
> > > listening on all 3 IPs.
> > > 
> > > However, attempting to bind the second IP with the first one bound gives a 
> > > message that the address is already in use.
> > > 
> > > This is vanilla 2.4.20 with Jacques' ctx-17 patch, no other modifications,
> > > with RedHat 7.3 on the host and virtual server.  (Custom compiled kernel -
> > > for highmem - but nothing different from other servers where multiple IPs 
> > > works ok.)
> > > 
> > > I'm using vserver-0.22-1, installed from the rpm. 
> > 
> > I assumed, the one from Jacques page ...
> > 
> > > Has anyone seen something similar?  Suggestions?  I don't see where in 
> > > /usr/sbin/vserver this error message is coming from.
> > > 
> > > Thanks!
> > > Cathy
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > 
> 
> 

Reply via email to