John P. Looney wrote:
exec .1.3.6.1.4.1.2021.51 memfree /usr/local/bin/memfree$
exec .1.3.6.1.4.1.2021.100 disk1 /usr/local/bin/diskused /$
The .100 OID is already used by "ucd-snmp/versioninfo", so you cannot
just hi-jack that one - but how would anybody know that ?!
I think a playing-ground subtree was introduced at some point. If you
use that one, you'll be safe from such collisions. (Or you could use an
OID within your own private subtree.)
Anyway, the above explains why your setup fails (duplicate registration),
but this shouldn't crash the agent! The reason for that is a double-free
of the handler registration object:
1) netsnmp_register_mib() => netsnmp_subtree_free()
2) netsnmp_register_old_api()
I guess the one in netsnmp_register_old_api() could just be removed, but
I haven't checked all possible error paths in the code (!!).
-Tommy
-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
Net-snmp-coders mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
