Perfect. Working now with my pure TCP/IP network;-) <TR ><th align="left" nowrap width="250"> <a href="/192.168.11.100.html" >192.168.11.100</a> <img src="/clock.gif" border="0" alt="NTP Server"><td align="center"> </td></th> <TD ALIGN=RIGHT>192.168.11.100</TD><TD ALIGN=RIGHT>00:10:5A:53:E6:A7</TD> <TD ALIGN=RIGHT NOWRAP> </TD><TD ALIGN=LEFT><IMG ALIGN=ABSMIDDLE SRC="/gaugeS.jpg" ALT="Sent 63%" WIDTH=189 HEIGHT=12><IMG ALIGN=ABSMIDDLE SRC="/gaugeR.jpg" ALT="Received 33%" WIDTH=99 HEIGHT=12> </TD> <TD ALIGN=RIGHT NOWRAP>3COM CORPORATION</TD><TD ALIGN=RIGHT> </TD><TD ALIGN=RIGHT>5</TD><TD ALIGN=RIGHT NOWRAP>44:46</A></TD><TD ALIGN=RIGHT NOWRAP> </TD></TR>
and the debug Wed Feb 25 21:38:09 2004 CMPFCTN_DEBUG: setResolvedName(0x085d8980) 0 -> 6 192.168.11.100 - hash.c(1160) For me ist is working now without problems with and without -n. But I'm not really interested in the NetBIOS name resolution. See the debug of another systems name resolution phases: Wed Feb 25 22:00:24 2004 CMPFCTN_DEBUG: setResolvedName(0x085ded50) 0 -> 3 ETHER - pbuf.c(664) Wed Feb 25 22:04:31 2004 CMPFCTN_DEBUG: setResolvedName(0x085ded50) 3 aether -> 6 192.168.11.240 - pbuf.c(3216) The first is the entry of the NetBIOS name broadcast, the second entry is an entry generate at the moment I pinged that host. Curious in that way that the first broadcast packet contained the MAC, the IP-Adress and the name of the host. Why not a 6 at the first packet? And the NetBIOS name a 5? And the IP-Address a 4 and for folks like me using the -n parameter switching off both 5 and 6? Otherwise I fear there are folks around who will say something like 'the NetBIOS Name Service is a name service and of equal value as the DNS resolution'. Think it is a valid argument (not for me, I'm interested in the IPs only an can live with NetBIOS prioritized at 3). Cheers Markus _____________________ On Wednesday 25 February 2004 17:29, Burton M. Strauss III wrote: > So it's a 6! FLAG_HOST_SYM_ADDR_TYPE_IP. So I'm looking for a place where > hRN transitions to 6, i.e. setResolvedName(... So it's a type 6! > FLAG_HOST_SYM_ADDR_TYPE_IP but the underlying value isn't set... There are > only 4, shouldn't be that hard to nail down. Especially as the debug line > is in the log: > > setResolvedName(0x085dd730) 0 -> 6 192.168.11.100 - hash.c(1160) > > (Don't you just LOVE it when the debugging stuff works!) > > Let's see in your 1st message you said "Tried it using the -n Parameter" > > I think I've found something odd - no clue if it's your problem - there's > one case where hostResolvedName (was) set w/o going through the function, > and that didn't test for the -n flag, just updated the name from the cache. > With -n set, that cache might not be loaded and so it would be updated to > blank. > > > I'll commit the patch in a few minutes, give it a try... > > For that to be the problem, however, you would have to also have a messed > up dns cache file. > > So please do the following. > > 1. Bring down the updated reportUtils.c, compile and test that > 2. Run dumpgdbm against your dnsCache.db (ntop must be down) > > $ dumpgdbm /usr/share/ntop/dnsCache.db | grep "'3232[23]" > > (dumpgdbm is at http://article.gmane.org/gmane.linux.ntop.general/3557, I > should probably throw the current version up @ SourceForge). > > 3. Delete dnsCache.db and rerun ntop > > Let me know what you see. > > -----Burton > _______________________________________________ Ntop mailing list [EMAIL PROTECTED] http://listgateway.unipi.it/mailman/listinfo/ntop
