Systemsoftware
2-40
System- and communication configuring D7-SYS - SIMADYN D
Edition 03.2001
The actual service utility processing is realized in a 32 ms sampling time
(the next sampling time below 35 ms is selected; the sampling times
specified at the SER blocks are not significant for processing). In the
cyclic task used, the service blocks are provided a certain computation
time, which may use as maximum the basic clock cycle T0. The ratio of
the basic clock cycle T0 to the used task defines the CPU performance
available and therefore the system loading.
Example 1:
Basic clock cycle T0 = 1 ms; selected sampling time = 32 ms. Every
32 ms, 1 ms is reserved for the service utility. Thus, the system load is
calculated of
1ms / 32 ms = 0.03125 = 3.125%
Example 2:
Basic clock cycle T0 = 2ms; selected sampling time = 16 ms. Every
16 ms, 2 ms are reserved for the service utility. Thus, a system load is
obtained from
2ms / 16 ms = 0.125 = 12.5%
This available computation time is used by all service blocks to the same
extent, i. e. as long as the time is sufficient, if possible, all of the SER
blocks are processed once. An SER block processes a maximum of one
task per clock cycle. For cyclic tasks, for each clock a maximum of one
response telegram is received. The advantage of this mode of operation
is that for cyclic tasks, equidistant-timed responses are obtained.
If the reserved computation time is not fully utilized, because, for
example, there is no task to be processed, then this time is made
avasilable to the system.
For multiple configuring with simultaneous access to system resources
which are only available once (e. g. change memory of the memory
module), resources are assigned to the first component which makes the
request. All others are rejected and output at the latest after 1 second, an
error message ("resource occupied") via the data interface.
In a fault condition (exception), i. e. for initialization errors or online faults,
the system goes into the stop mode. Thus, there are special conditions
for the service utility. It is then no longer computed in a cyclic task, but
runs continuously, started from an exception administrator. Under fault
conditions, the service utility cannot be connected to the configured user.
In order to still permit system diagnostics, the CPU’s own diagnostics
interface is connected. The DUST1 protocol runs here (refer to the
Chapter "Operating statuses of a CPU module").
System loading,
response times
Behavior under
fault conditions
