I agree with the list suggested by metlin (258108) and augmented by dubl-u (51156) and "Anonymous Coward on Saturday August 25, @10:41PM":
Stability
Reliability
Response time
Network load
Security
Efficiency
User satisfaction
Crisis preparedness
I would add, separate from "Response Time [to problems]", some measure of how change requests have been handled, both in responsiveness and quantity. A significant portion of our sysadmin time is spent on bringing new servers or applications online, or adding new features or upgrades.
Many of the responses I read seemed to assume a fairly static configurations, which doesn't seem realistic to me. (Or maybe such changes are handled by other staff in their organization).
Slashdot Mirror
I agree with the list suggested by metlin (258108) and augmented by dubl-u (51156) and "Anonymous Coward on Saturday August 25, @10:41PM":
Stability
Reliability
Response time
Network load
Security
Efficiency
User satisfaction
Crisis preparedness
I would add, separate from "Response Time [to problems]", some measure of how change requests have been handled, both in responsiveness and quantity. A significant portion of our sysadmin time is spent on bringing new servers or applications online, or adding new features or upgrades.
Many of the responses I read seemed to assume a fairly static configurations, which doesn't seem realistic to me. (Or maybe such changes are handled by other staff in their organization).