Why Snmp Traps?
In continuation to my last post on highlighting Snmp Trap, mostly sysadmins uses Snmp based polling; MRTG, RRD etc...such sysadmins might ask why Snmp Traps when we already are polling? Polling is part of “monitoring”; another important pieces are analysis and notification based on this analysis….
With loads of data, analysis becomes harder….Snmp Trap is an easy escape….you are basically configuring at device end to do this analysis and notification…But we should not setup everything as Snmp Traps; it will lose its significance…just like “snmp monitoring” and then it puts significant load on device (in case of routers and switches)
Methods, Tools & Technology relationship
Snmp or Traps is not answer to all problems…technology and tools alone can’t do everything; methods play important part ….Having more tools and good technology basically allows “simple methods”; e-g its hard to say who is better resource…unless you have technology (timesheets, peer reviews, managerial reviews) and tools (attendance system, punch systems, online surveys, spreadsheets etc)….with these tools you can easily evaluate who is better resource…..Similarly…Snmp Trap basically is an additional option along with Snmp Polling….you can have simple method that Traps will be given more priority along with reviewing Snmp Polled graphs….reviewing graphs can be skipped…but if given due importance to (well configured) Snmp Trap; chances get reduced….
 Quality Software Project Management by Robert T. Futrell, Donald F. Shafer, Linda I. Safer, Prentic Hall PTR, 2002