1) Make sure the solution covers “every hop”
Many end-to-end monitoring solutions focus on monitoring end to end round trip times in the datacenter plus the Java or .NET application server. But what about your LDAP, ESB or Proxy server? These overlooked components could be your bottleneck.
2) No code or architecture changes should be tolerated As important as getting your “wing-to-wing” application stack covered there is no need to compromise on ease of installation and maintenance. Today’s more modern tools go in without having to involve network and application teams.
3) Everything should be tied to a business context
If hundreds of alerts are being triggered each day – how do you know which ones are real and which ones are false alarms? When everything is tied to a business context prioritization becomes easy.
This is accomplished by automatically correlating everything that happens throughout the application stack from the moment a specific user conducts a discreet business transaction. Most of the traditional end-to-end solutions out there were not designed to automatically accomplish this correlation.
Whether you are putting together an RFP or simply conducting an initial investigation and information gathering – I advise that you bring up these must-haves during your conversations with vendors. Also – try to get an understanding of whether their underlying technology can really support them.