On 11/11/09 2:53 PM, Dave Shield wrote: > 2009/11/11 Mike Bowie <[email protected]>: >> Just a quick inquiry to see if anyone in the net-snmp team is looking at >> (or perhaps leaning towards) NetBSD as a target. I've been asking >> around in the NetBSD and pkgsrc communities for quite some time now and >> no one seems to have a solution to some issues specific to the platform. > > Ask your questions here, and we'll see what we can do to help. > > Note that it will be extremely useful if you can indicate which version > of the Net-SNMP agent you are using. I'm sure I could eventually work > out what version is shipped with various releases of NetBSD. > But it's probably simpler if you just tell us directly!
Thanks for the response Dave... much appreciated. pkgsrc is currently building 5.4.2.1. At the most basic level, the issue is that querying snmpd returns values which look to be uninitialized, for example: $ snmpwalk -v1 -c public 127.0.0.1 | grep "should be" IP-MIB::ipForwarding.0 = Wrong Type (should be INTEGER): Counter32: 4601253 IP-MIB::ipDefaultTTL.0 = Wrong Type (should be INTEGER): Counter32: 4601253 IP-MIB::ipReasmTimeout.0 = Wrong Type (should be INTEGER): Counter32: 4601253 TCP-MIB::tcpRtoAlgorithm.0 = Wrong Type (should be INTEGER): Counter32: 4294967295 TCP-MIB::tcpRtoMin.0 = Wrong Type (should be INTEGER): Counter32: 4294967295 TCP-MIB::tcpRtoMax.0 = Wrong Type (should be INTEGER): Counter32: 4294967295 TCP-MIB::tcpMaxConn.0 = Wrong Type (should be INTEGER): Counter32: 4294967295 TCP-MIB::tcpCurrEstab.0 = Wrong Type (should be Gauge32 or Unsigned32): Counter32: 4294967295 Since all our NetBSD systems are amd64, I initially thought we were seeing 32bit truncation issues, but the issue turned out to exist on both i386 and amd64. The behavior exists on all the NetBSD builds we have running, which are from 4.0.1 through 5.0.1. There are other error messages output to syslog, along the lines of: Nov 11 23:00:23 netbsd5-0-stable-amd64 snmpd[6022]: netsnmp_assert !"cache == valid" failed mibII/tcp.c:228 tcp_handle r() There's one other I was getting, but I'm not able to find my notes on it right now... and I seem to recall that some google'ing revealed it as benign. I'd pull it now, but the build I have handy has all kinds of debug monkeying in it, so it's just spewing my sprintf's all over the console. I hope that gives some insight; please let me know if any other details would be of interest. Cheers, Mike. ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Net-snmp-coders mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
