Re: Problem using net-snmp as subagent (Error in packet)
Frank Fock <[email protected]>
| Newsgroups | gmane.network.net-snmp.user |
|---|---|
| Organization | AGENT++ |
| Message-ID | <[email protected]> |
Hi Dave, Sorry, the below "issue" is a false alarm. The user has encountered a firewall issue. So NET-SNMP is properly using the authoritative engine ID for discovery! Cheers, Frank Frank Fock wrote: > Hi Dave, > > I have not forgotten to put together the AgentX > issues that I have found while testing NET-SNMP > interoperability. I think I can provide that > report next weekend. > > Meanwhile I came across another issue related > to NET-SNMP: > > SNMP4J(-Agent) supports engine ID discovery. > A SNMP4J user reported, that he used a 5.3.1 > NET-SNMP tool to send an inform request to > a SNMP4J based manager application. > > The NET-SNMP tool sends a GET PDU > to discover SNMP4J's engine ID. > SNMP4J sends back a REPORT PDU with its > engine ID in the authoritative engine ID > field and the contextEngineID field is empty. > NET-SNMP seems to ignore the authoritative > engine ID and again sends a discover GET PDU. > > If the contextEngineID is set to the > authoritative engine ID then NET-SNMP > seems to recognize the engine ID and sends the > INFORM PDU. I am writing "seems to" because > this has been tested with another tool which > sends its REPORT from the same port as it > received the GET PDU whereas SNMP4J is using > a different port (and no contextEngineID). > > From my point of view this is a bug because > RFC 3414 states in §3.2 that specifying the > contextEngineID in a REPORT PDU is optional. > > What do you think about this? > > Cheers, > Frank > > Dave Shield wrote: >> >> We're currently in the (early) process of releasing 5.2.4, >> so it's an ideal time to fix that particular line. Fixes for >> the 5.3.x and 5.4.x lines will take a little longer, but >> I can get the patch applied so it's ready to roll out. >> >> >>> BTW, I have implemented a few workarounds into >>> SNMP4J-AgentX and AgentX++ to cope with some bugs >>> of the NET-SNMP subagent. Are interested in the >>> details? >> >> Most Definitely! >> I've been pulling together the outstanding bugs on our >> tracker system, so it would be good to get as much of >> this sorted as possible. A clear analysis of where things >> have gone wrong is invaluable. >> >>> There seem also to be 64bit OS related bugs >>> when encoding an IpAddress on AgentX. >> >> That wouldn't surprise me either. >> Let us have what you've got, and we'll see what we >> can do. >> >> Dave >> > -- AGENT++ http://www.agentpp.com http://www.mibexplorer.com http://www.mibdesigner.com ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Net-snmp-users mailing list [email protected] Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users