--> [acer]lanman has joined #net-snmp
--> [intra]lanman has joined #net-snmp
--> ChanServ has joined #net-snmp
--> [intra]lanman has joined #net-snmp
--> [intra]lanman has joined #net-snmp
--> intralanman has joined #net-snmp
--> jsafrane has joined #net-snmp
--> jsafrane has joined #net-snmp
--> hardaker has joined #net-snmp
--> [intra]lanman has joined #net-snmp
--> exothermc has joined #net-snmp
--> [intra]lanman has joined #net-snmp
--> [intra]lanman has joined #net-snmp
[2009/06/22 08:30:44] #net-snmp <fenestro> anyone ever run into trouble with sysUpTime skew with agentx master/client? I'm vaguely blaming queueing delay on the agentx socket (e.g., if I take more than 10ms to service the agentX socket then the recovered starttime value will be off by enough to matter)
--> T_X has joined #net-snmp
--> kisielk has joined #net-snmp
--> frostburn has joined #net-snmp
--> magfr has joined #net-snmp
[2009/06/22 11:36:48] #net-snmp <rstory> fenestro: it shouldn't skew on a master.. unless the get request also contains values from an agentx client, then it could be off by the time used by the subagent to process the request..
[2009/06/22 11:37:12] #net-snmp <rstory> and i don't think there's much we could do about that..
[2009/06/22 12:25:44] #net-snmp <Hobart> Can anyone suggest what I might be doing wrong if my calls to netsnmp.dll keep throwing "Cannot find module (IP-MIB): At line 0 in (none)"
[2009/06/22 12:26:08] #net-snmp <Hobart> ...even when the first invocation of the DLL is an attempt to do AddMibDir to tell it where to find all of them?
[2009/06/22 12:27:05] #net-snmp <Hobart> per hardaker's suggestion, I turned on debugTokens ALL and doDebugging 1, so I can see all the trace messages leading up to it
--> Cru has joined #net-snmp
[2009/06/22 13:08:01] #net-snmp <Cru> hi
[2009/06/22 13:08:45] #net-snmp <Cru> I've got this problem with CentOS5's Net-SNMP snmpd: http://home.ircnet.de/cru/zeug/snmpd.png
[2009/06/22 13:10:23] #net-snmp <Cru> since a few days it reports 100.0% CPU usage on the server, but the usage is in fact low -- the "gaps" when not showing 100% occur when snmpd is restarted -- it seems to report a correct value on the first query, then switching back to continously reporting 100.0%
[2009/06/22 13:13:36] #net-snmp <Cru> the version is 5.3.2.2 -- any suggestions how to fix this?
[2009/06/22 14:10:57] #net-snmp <fenestro> rstory: sorry, I meant sysUpTime skew between master and subagent. The subagent is reporting a sysUpTime value which is *sometimes* off by one from the master's, and since the master reports sysUpTime and the subagent implements this timeFiltered table, I sometimes can't provide the precise TimeFIlter semantics (e.g., get sysUpTime, then something changes, then walk column.UpTime and you don't get the changed row
--> mike_irc has joined #net-snmp
--> blotchy has joined #net-snmp
--> hardaker has joined #net-snmp
--> hardaker2 has joined #net-snmp
--> varun_c has joined #net-snmp
--> varun_c_ has joined #net-snmp
--> chands has joined #net-snmp
--> varun_c has joined #net-snmp
--> varun_c_ has joined #net-snmp
--> snmpbot has joined #net-snmp
--> jsafrane has joined #net-snmp