I wrote regression tests for my two reported cases: 1. sysUpTime gets strange if the clock changes after net-snmp starts; 2. subagent fails to reconnect until the clock catches back up if the clock changes after it sets its reconnect timer.
I haven't yet gotten the test for #2 to fail in my previous configuration, so I still have some test development to do there. Unsurprisingly, #1 fails with the old code and passes with the new. My test for master-subagent time sync still passes, so nothing went funny about the agentx subagent's starttime value. I haven't changed my agent infrastructure to use starttimeM yet, although the window for trouble there is so much smaller because the subagent code calls netsnmp_set_agent_uptime() on every request. Bill ------------------------------------------------------------------------------ Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ _______________________________________________ Net-snmp-coders mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
