This command defines a limit into the current node. It is a means of providing simple load management by say limiting the number of tasks submitted to a specific server.
Typically you either define limits on suite level or define a separate suite to hold limits so that they can be used by multiple suites.
| Code Block |
|---|
suite limits |
...
limit sgi 10 |
...
limit mars 10 |
...
endsuite |
...
suite obs |
...
family |
...
limits limit hpcd 20 endfamily extern /limits:sgi |
...
task |
...
t1 inlimit sgi task t2 inlimit /obs/limits:hpcd endsuite |
Using Limits
In order to use limits, you have to first define a limit in the suite definition file and then you must also assign families/tasks to use this limit, e.g.
| Code Block |
|---|
suite x
limit fast 1
family f
inlimit /x:fast
task t1
task t2 |
Where fast is the "name" of the limit and the number after the name defines the maximum number of tasks that can run simultaneously using this limit. That's why you do not need a trigger between tasks "t1" and "t2".
There is no need to change tasks. The jobs are run in the order they appear in the suite definition. Be aware that the command order may be used to modify the order after the suite definition has been defined.
The following is a sequence if the jobs run in the normal order
And if you manually now re-run the "t1" (you go over the limit!) 
Using integer limits is a bit different. In the following example, we define limit disk to be 50 (megabytes) and task using 20 (megabytes) each. This means that only two of them can be running any given time.
| Code Block |
|---|
suite x
limit disk 50
family f
inlimit disk 20
task t1
task t2
task t3 |
Where disk is the "name" of the limit, 50 is the maximum value this limit can be.
endsuiteAnchor