Hi Brett, SNMP4J sends a GET request to discover an engine ID when you use Snmp.discoverAuthoritativeEngineID(..).
If you see an INFORM request, then that belongs to something different. The second request is indeed not necessary when using Snmp.discoverAuthoritativeEngineID(..). I will optimize it for the 1.10 release. The NET-SNMP implementation did not correctly implement the discovering process, but I thought in the latest release this had been fixed. Best regards, Frank Brett Hall wrote: > Hi, I'm using snmp4j 1.9.3.b (also tried 1.9.3.d). > > To define Localized USM users when sending V3 Informs, I'm using > Snmp.discoverAuthoritativeEngineID() > > I encountered an unexpected issue discovering a NET-SNMP 5.4.1 > snmptrapd. My discover request takes the full timeout value to return. > > I do get the engineId result. > > Connecting up wireshark I noticed that two GET requests are sent out by > SNMP4J. NET-SNMP ignores the second GET. > > Request 1: Inform with zero length engine ID noAuth/noPriv. > Response 1: Receive Report usmStatsUnknownEngineIDs and the discovered > engineID > > Request 2: Inform with discovered engine id, noAuth/noPriv. > Response 2: No response. PDU times out > > 1. Should the second request be sent at all? > Snmp.ReportProcessor is coded to resend PDU in response to Report. > Seems right for auto-discovery but not for manual discovery. > > 2. NET-SNMP seems also at fault. > - NET-SNMP 5.4.1 (snmptrapd) times out. > - MG-Soft MIB Browser (Trap Ringer) - Ok and returns > unSupportedSecurity level response. > > Thoughts? Thanks. > Brett > _______________________________________________ > SNMP4J mailing list > [email protected] > http://lists.agentpp.org/mailman/listinfo/snmp4j -- AGENT++ http://www.agentpp.com http://www.snmp4j.com http://www.mibexplorer.com http://www.mibdesigner.com _______________________________________________ SNMP4J mailing list [email protected] http://lists.agentpp.org/mailman/listinfo/snmp4j
