chapter_22

Exercises 22

1. Implement basic job center functionality with the following interface:

• job_centre:start_link() -> true.
• job_centre:add_job(F) -> JobNumber.
• job_centre:work_wanted() -> {JobNumber, F} | no.
• job_centre:job_done(JobNumber)

Solution in the exercise_1/ directory.

erl
1> job_centre:start_link().
{ok,<0.67.0>}
2> job_centre:add_job(fun() -> io:format("working on job") end).
1
3> {JobID, JobFun} = job_centre:work_wanted().
{1,#Fun<erl_eval.20.99386804>}
4> JobFun().
working on jobok
5> job_centre:job_done(JobID).
ok

2. Add function named job_centre:statistics/0 that reports the status of the jobs in the queue.

Solution in the exercise_2/ directory.

erl
1> job_centre:start_link().
{ok,<0.67.0>}
2> job_centre:add_job(fun() -> io:format("Working on job 1") end).
1
3> job_centre:add_job(fun() -> io:format("Working on job 2") end).
2
4> {JobID, JobFun} = job_centre:work_wanted().
{1,#Fun<erl_eval.20.99386804>}
5> JobFun().
Working on job 1ok
6> job_centre:job_done(JobID).
not_found
7> job_centre:statistics().
[{1,in_progress},{2,pending}]

3. Add code to monitor the workers. If a worker dies makesure the jobs it was doing are returned to the queue of waiting jobs.

Solution in the exercise_3/ directory.

erl
1> job_centre:start_link().
{ok,<0.67.0>}
2> job_centre:add_job(fun() -> io:format("Working on job 1") end).
1
3> job_centre:add_job(fun() -> io:format("Working on job 2") end).
2
4> spawn(fun() ->
    {JobID, JobFun} = job_centre:work_wanted(),
    JobFun(),
    exit(self(), test_failure)
    end).
Working on job 1<0.71.0>
Worker process <0.71.0> crashed with reason test_failureDOWN message from unknown process
5> {JobID, JobFun} = job_centre:work_wanted().
{1,#Fun<erl_eval.20.99386804>}
6> JobFun().
Working on job 1ok

4. Check for lazy workers. Change the work wanted function to return {JobNumber, JobTime, F} where JobTime is the number of seconds that the worker as to complete the job by. At JobTime - 1, the server should send a hurry_up message to the worker if it has not finished the job. And at time JobTime + 1, it should kill the worker process with an exit(Pid, youre_fired) call.

Solution in the exercise_4/ directory.

erl
1> job_centre:start_link().
{ok,<0.67.0>}
2> job_centre:add_job(fun() -> io:format("Working on job 1") end).
1
3> job_centre:add_job(fun() -> io:format("Working on job 2") end).
2
4> spawn(fun() ->
    {JobID, JobFun} = job_centre:work_wanted(),
    JobFun(),
    receive
      Msg ->
        io:format("~p", [Msg])
    after 3000 ->
        ok
    end
  end).
Working on job 1<0.71.0>
Worker process <0.71.0> crashed with reason test_failureDOWN message from unknown process
5> job_centre:statistics().
[{1,pending},{2,pending}]

5. Implement a trade union server to monitor the workers. Check that they are not fired without being sent a warning.

Not going to implement this as it's not a design pattern I've seen regularly used in OTP applications.