15.10 - Process Scheduling CPU Utilization Columns - Teradata Database

Teradata Database Resource Usage Macros and Tables

Teradata Database
Release Number
April 2016
English (United States)
Last Update

These columns count all CPU activities, including activities performed for virtual processors.

The CPU utilization columns are aggregates representing all CPUs on the node. CPU utilization by user code is further subdivided by the vproc tables.

  • CPU idle time = CPUIdle + CPUIoWait
  • CPU busy time = CPUUServ + CPUUExec

Theoretically, the values of these four columns, for any given interval, account for total CPU time on the node. That is, these columns should total to 100 * Secs * number of CPUs on the node, since each CPU is always in exactly one of these four states. In practice, there is occasionally a very small plus or minus difference from this theoretical total.

Column Name Mode Description Data Type
CPUIdle count Time in centiseconds the CPU is idle and not waiting for I/O. FLOAT
CPUIoWait count Time in centiseconds CPU is waiting for I/O completion.
This represents another variety of Idle, since the CPU is only recorded as being in this state if there are no processes eligible for execution. This is because if there were any such process, the CPU would be immediately dispatched for that process.
CPUUServ count Time in centiseconds CPU is busy executing kernel system calls or servicing I/O and timer hardware interrupts. FLOAT
CPUUExec count Time in centiseconds CPU is busy executing user execution code, that is, time spent in a user state on behalf of a process. FLOAT