Send netdisco-users mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.sourceforge.net/lists/listinfo/netdisco-users
or, via email, send a message with subject or body 'help' to
[email protected]
You can reach the person managing the list at
[email protected]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of netdisco-users digest..."
Today's Topics:
1. Subnet Discovery (Tim Polimeni)
2. Re: Subnet Discovery (Christian Ramseyer)
3. Re: Subnet Discovery (Christian Ramseyer)
--- Begin Message ---
Hello,
Is there a way to allow the CMD on the CLI side to run a discovery on a
/16 subnet? I have a regexp set up to only match on FQDN portions. Any help
to do this would be great.
info discover: aborted - unwise to attempt 65534 jobs at once
--- End Message ---
--- Begin Message ---
Hi Tim
On 14.08.19 18:00, Tim Polimeni wrote:
> Is there a way to allow the CMD on the CLI side to run a discovery
> on a /16 subnet? I have a regexp set up to only match on FQDN portions.
> Any help to do this would be great.
>
> info discover: aborted - unwise to attempt 65534 jobs at once
>
It might not be necessary to discover this whole network explicitly.
Once you discover one or a couple of central devices in you network,
Netdisco will automatically queue discovery of neighbors discovered by
various methods like LLDP, BGP adjacency etc.
However if you really want or have to do it, it's a pretty arbitrary
limit statically defined in the file netdisco-do, line 117:
if (scalar @hostlist > 512) {
info sprintf '%s: aborted - unwise to attempt %s jobs at once',
$action, (scalar @hostlist);
exit 1;
}
For a temporary fix, you can just comment the "exit 1;" line. However
this check is there for a reason. I'd make sure that your organization
owns this whole range, and every NOC involved is prepared to get some
SNMP authfail alerts. Also in a /16 where will be ten thousands of IPs
that don't reply, but the SNMP protocol no way to tell no reply, no snmp
agent running and authentication errors apart - all of these result in a
timeout. So Netdisco will try every configured community times the
timeout times the repeat count on all of these addresses, which might
take a very long time.
I suspect the above is the reason that Oliver added this check in the
first place, and I'd be slightly wary to patch in a --force flag that
allows to bypass the check without messing around in the code. However
if more people think this would be useful, it shouldn't be hard to add
at all.
Cheers
Christian
--- End Message ---
--- Begin Message ---
On 14.08.19 23:32, Tim Polimeni wrote:
> Thanks for the info, yes so the network is a little different and very
> large, we own several /16s. We won’t see any SNMP failures since we use
> the same string in the entire network and my server is allowed via ACL.
> I only want the management switches learned not our core network which
> is why I have the regexp in place to only matches devices that start
> with uer in the FQDN followed by region names.
>
> I don’t discover by LLDP, ospf of BGP as this will discover our core
> network which would in return be thousands of routers which I don’t care
> to discover. I’ll test out the suggestion below.
Ok great, sounds like you know what you're doing :) Good luck and let us
know if it shouldn't work.
Cheers
Christian
--- End Message ---
_______________________________________________
Netdisco mailing list - Digest Mode
[email protected]
https://lists.sourceforge.net/lists/listinfo/netdisco-users
