Kyle,
The LDM doesn't do anything special to determine what interface (i.e.,
IP address) to use when connecting to another LDM: it leaves that job to
the operating system. You can see what the O/S is going to do via the
command "netstat -nr". You can then use the route(8) utility's "add"
command to tell the O/S to use a specific interface when connecting to a
particular given host. If you use an interface for which the upstream
LDM has an ALLOW entry, then things should work correctly.
Regards,
Steve Emmerson
LDM Developer
Kyle Schaffrick wrote:
> Hello,
>
> I'm attempting to set up an LDM instance for the first time and have run
> into some issues. The machine is CentOS Linux; I'm using LDM 6.6.5.
>
> When I start the LDM daemon, it fails to receive any products (according
> to 'ldmadmin watch'), and logs messages of the form (host names redacted
> at operators' request):
>
> Aug 29 04:20:51 ldm-server upstream.com[28268] NOTE: LDM-6 desired
> product-class: 20080829032051.817 TS_ENDT {{WMO, ".*"}}
> Aug 29 04:20:51 ldm-server upstream.com[28268] ERROR: Disconnecting
> due to LDM failure; nullproc_6 failure to upstream.com; RPC:
> Unable to receive; errno = Connection reset by peer
>
> at "pseudo-random" intervals, about 1 or 2 per second. 'ldmping' also
> reports "Connection reset by peer".
>
> The machine I'm running it on has a number of IP addresses set up on its
> interface, and according to the upstream provider, their server is
> rejecting our requests because the connections are coming from one of
> the other IPs on the interface, i.e., not the one they ALLOWed.
>
> I've set $ip_addr to the correct address in ldmadmin-pl.conf, LDM emits
> a log message during startup that it is listening on that address, and
> 'netstat -l' also corroborates this. However, all signs seem to point to
> the client code somehow failing to bind it's *outgoing* socket to that
> address.
>
> I'm not sure where to go from here as far as troubleshooting this. It
> bears mentioning that I'm also getting an error message during startup
> regarding a nonexistent file /etc/killall (which past traffic on this
> list indicated was a problem with the configure script), but this
> in itself doesn't seem to be causing any immediate problems.
>
> Any pointers on investigating this would be greatly appreciated.
>
> -Kyle
> _______________________________________________
> ldm-users mailing list
> ldm-users@xxxxxxxxxxxxxxxx
> For list information or to unsubscribe, visit:
> http://www.unidata.ucar.edu/mailing_lists/
