Waits, Enqueues, and Latches - Expert Oracle RAC Performance Diagnostics and Tuning - page 582

Time Sharing
This is the default class for a process. When processes with the same priority are in run time mode, it changes the
priorities dynamically to use the RR method. As discussed earlier, the scheduler allocates slices of time to a process,
and this depends on the scheduling priority. Under this class, the processes that are scheduled with higher priority get
larger time slices. This is based on the assumption that lower priority processes do not run often and should be given
a larger quantum.
Fixed Priority
The priorities associated with the processes under this class are fixed. They did not dynamically change over the
lifetime of the process. FX is normally present in operating systems such as Solaris and has a range of 0-59.
System
This class is used to schedule kernel processes. Processes in this class are bound, which means they run until they
block or complete. SYS has a range of 60-99.
Real Time
The real time class uses priorities in the range of 100-159. They are higher compared to the time-sharing processes,
which means the real time process will be scheduled before any kernel process.
Round Robin
Like the name “round robin,” the different program/process takes turns using the resources of the computer to limit
each process to a certain short time period, then suspends that process to give another process a turn (or “time slice”).
First in First Out
This is probably the simplest type of scheduling algorithms available; and as the name suggests, the process or
requests for CPU resources are queued in the order in which they arrive. Because of the nature in which they are
received and processed, the rules are fixed, and no prioritization of the processes is possible. Each CPU has a run
queue made up of 140 priority lists that are serviced in FIFO order. Every time a new task is scheduled for execution,
it's added to the end of the task's respective run queue.
CPU utilization by a process can be determined using the “top” command on Linux and UNIX systems or with
more details using the collectl , mpstat , sar , or the iostat commands.
[oracle@ssky1l4p1 ~]$ collectl -i 5
waiting for 5 second sample...
#<----CPU[HYPER]-----><----------Disks-----------><----------Network---------->
#cpu sys inter ctxsw KBRead Reads KBWrit Writes KBIn PktIn KBOut PktOut
1 0 1284 11041 100 8 338 33 75 157 60 144
0 0 1329 10691 68 6 309 28 67 182 69 165
0 0 1169 10590 97 8 123 12 69 96 53 83