Herman,

I have worked with a number of customers in that environment that have multiple people working with the products. The numbers always depended on what you were doing and how the whole item was architected. Even the most automated system needs care and feeding by a few people.

The one question I aways brought up early with my customers is: what are you going to do with the notification of an outage? How do you want to be notified? I have seen everything from a team of operators at NetView consoles which open up trouble tickets to a fully automated system. I would ask the customer...how many tickets can a single operator open in an hour? Then, how many outages requireing tickets must you address in an hour? That tells me how many operators there had to be.

A very good practice -- Netview did its discovery and a rule was coded in with a list of 'critical' devices. For traps on non-critical items, they were simply logged for later review. For outages on critical items, an event was sent to TEC. In TEC there was some additional rules to get root cause and then a trouble ticket was automatically generated. In the trouble ticketing system..every critical device was assigned an owner and escalation path. The TT system generated a page to the responsible person, if they person did not log into the TT and acknowledge the ticket, then the next person up the escalation path was paged. All of this happened without operator intervention. When problem was resolved, the person closed the ticket...which went back into TEC and closed the events.

you still need someone adding critical items into the system, care and feeding of databases, writing new TEC rules and managing the trouble ticket system. Certainly more than a one person job.

Stephen Hochstetler shochste AT us.ibm DOT com
International Technical Support Organization at IBM
Office - 512-838-6198 (t/l 678) FAX - 512-838-6931
http://www.redbooks.ibm.com