[Openswan Users] Problem with openswan in embedded system

[Paul Wouters]

On Fri, 10 Dec 2010, Panagiotis Tamtamis wrote:

> I found out that from the logs the following errors:

> 2010/12/03 18:00:24 debug HOOME authpriv pluto[27788]: | cmd( 640):L+PFS+UP+IKEv2ALLOW\'  ipsec _updown:
> 2010/12/03 18:00:25 info HOOME daemon confd_ip[27777]: rule flush cache
> 2010/12/03 18:00:25 err HOOME daemon confd_ip[4115]: Error reading data: Connection reset by peer
> 2010/12/03 18:00:25 warning HOOME authpriv pluto[27788]: \"sample\" #2: route-host output: select() error: Bad file descriptor
> 2010/12/03 18:00:25 warning HOOME authpriv pluto[27788]: \"sample\" #2: route-host command exited with status 255

I have never seen that. I'm not sure what that means.

> And then I tried to run some commands like ip route list or ip addr list
> 
> It seems that the netlink socket in the system does not always provide the data.
> For example if I run the command for one time it might work. But if I do it again it doesn't.
> 
> And from the strace I have some really strange outputs. It seems that the data are indeed written but are not printed.
> And the resource seems to be temporarily unavailable.
> 
> So I guess I will go into this direction to see why I have this behavior in the system and why the route-host command which as it seems
> uses netlink sockets and the confd_ip resource, fails.
> 
> Any hints are always welcomed.


Is this a NOMMU kernel?

Are you severely resource restrained (ram?)

> I will also keep you posted in case I find out more.

Please do. I would like to understand what it is.

Paul