cpu_sup is a process which supervises the CPU load
and CPU utilization. It is part of the OS_Mon application, see
os_mon(6). Available for Unix,
although CPU utilization values (util/0,1) are only
available for Solaris and Linux.
The load values are proportional to how long time a runnable
Unix process has to spend in the run queue before it is scheduled.
Accordingly, higher values mean more system load. The returned
value divided by 256 produces the figure displayed by rup
and top. What is displayed as 2.00 in rup, is
displayed as load up to the second mark in xload.
For example, rup displays a load of 128 as 0.50, and
512 as 2.00.
If the user wants to view load values as percentage of machine
capacity, then this way of measuring presents a problem, because
the load values are not restricted to a fixed interval. In this
case, the following simple mathematical transformation can
produce the load value as a percentage:
PercentLoad = 100 * (1 - D/(D + Load))
D determines which load value should be associated with
which percentage. Choosing D = 50 means that 128 is 60%
load, 256 is 80%, 512 is 90%, and so on.
Another way of measuring system load is to divide the number of
busy CPU cycles by the total number of CPU cycles. This produces
values in the 0-100 range immediately. However, this method hides
the fact that a machine can be more or less saturated. CPU
utilization is therefore a better name than system load for this
measure.
A server which receives just enough requests to never become
idle will score a CPU utilization of 100%. If the server receives
50% more requests, it will still scores 100%. When the system load
is calculated with the percentage formula shown previously,
the load will increase from 80% to 87%.
The avg1/0, avg5/0, and avg15/0 functions
can be used for retrieving system load values, and
the util/0 and util/1 functions can be used for
retrieving CPU utilization values.
When run on Linux, cpu_sup assumes that the /proc
file system is present and accessible by cpu_sup. If it is
not, cpu_sup will terminate.
Types:
UnixProcesses = int()
Reason = term()
Returns the number of UNIX processes running on this machine.
This is a crude way of measuring the system load, but it may
be of interest in some cases.
Returns 0 if cpu_sup is not available.
Types:
SystemLoad = int()
Reason = term()
Returns the average system load in the last minute, as
described above. 0 represents no load, 256 represents the load
reported as 1.00 by rup.
Returns 0 if cpu_sup is not available.
Types:
SystemLoad = int()
Reason = term()
Returns the average system load in the last five minutes, as
described above. 0 represents no load, 256 represents the load
reported as 1.00 by rup.
Returns 0 if cpu_sup is not available.
Types:
SystemLoad = int()
Reason = term()
Returns the average system load in the last 15 minutes, as
described above. 0 represents no load, 256 represents the load
reported as 1.00 by rup.
Returns 0 if cpu_sup is not available.
Types:
CpuUtil = float()
Reason = term()
Returns CPU utilization since the last call to
util/0 or util/1 by the calling process.
Note
The returned value of the first call to util/0 or
util/1 by a process will on most systems be the CPU
utilization since system boot, but this is not guaranteed
and the value should therefore be regarded as garbage. This
also applies to the first call after a restart of
cpu_sup.
The CPU utilization is defined as the sum of the percentage
shares of the CPU cycles spent in all busy processor states
(see util/1 below) in average on all CPUs.
Returns 0 if cpu_sup is not available.
Types:
Opts = [detailed | per_cpu]
UtilSpec = UtilDesc | [UtilDesc]
UtilDesc = {Cpus, Busy, NonBusy, Misc}
Cpus = all | int() | [int()]()
Busy = NonBusy = {State, Share} | Share
State = user | nice_user | kernel
| wait | idle | atom()
Share = float()
Misc = []
Reason = term()
Returns CPU utilization since the last call to
util/0 or util/1 by the calling process, in
more detail than util/0.
Note
The returned value of the first call to util/0 or
util/1 by a process will on most systems be the CPU
utilization since system boot, but this is not guaranteed
and the value should therefore be regarded as garbage. This
also applies to the first call after a restart of
cpu_sup.
Currently recognized options:
-
detailed
-
The returned UtilDesc(s) will be even more
detailed.
-
per_cpu
-
Each CPU will be specified separately (assuming this
information can be retrieved from the operating system),
that is, a list with one UtilDesc per CPU will be
returned.
Description of UtilDesc = {Cpus, Busy, NonBusy, Misc}:
-
Cpus
-
If the detailed and/or per_cpu option is
given, this is the CPU number, or a list of the CPU
numbers.
If not, this is the atom all which implies that
the UtilDesc contains information about all CPUs.
-
Busy
-
If the detailed option is given, this is a list
of {State, Share} tuples, where each tuple
contains information about a processor state that has
been identified as a busy processor state (see below).
The atom State is the name of the state, and
the float Share represents the percentage share of
the CPU cycles spent in this state since the last call to
util/0 or util/1.
If not, this is the sum of the percentage shares of
the CPU cycles spent in all states identified as busy.
If the per_cpu is not given, the value(s)
presented are the average of all CPUs.
-
NonBusy
-
Similar to Busy, but for processor states that
have been identified as non-busy (see below).
-
Misc
-
Currently unused; reserved for future use.
Currently these processor states are identified as busy:
-
user
-
Executing code in user mode.
-
nice_user
-
Executing code in low priority (nice) user mode.
This state is currently only identified on Linux.
-
kernel
-
Executing code in kernel mode.
Currently these processor states are identified as non-busy:
-
wait
-
Waiting. This state is currently only identified on
Solaris.
-
idle
-
Idle.
Note
Identified processor states may be different on different
operating systems and may change between different versions
of cpu_sup on the same operating system. The sum of
the percentage shares of the CPU cycles spent in all busy
and all non-busy processor states will always add up to
100%, though.
Returns {all,0,0,[]} if cpu_sup is not
available.