Re: Is job x running

When I am having a big heated discussion at work, I use a rubber chicken which I keep in my desk for just such occassions. The person holding the chicken is the only person who is allowed to talk. If you don't hold the chicken you cannot speak. You can only indicate that you want the chicken and wait until you get it before you speak. Once you have finished speaking, you can hand the chicken back to the moderator who will hand it to the next person to speak. This ensures that people do not speak over each other, and also have their own space to talk.

Replace Chicken with Mutex and person with thread and you basically have the concept of a mutex.

... for a data warehouse load process, we've needed multiple conditions, what could run together. We wrote a SRVPGM (not too complcated) with following procedures

getMutex(name, wait){
if syncobj with name not exists create
try to lock the syncobj with specified wait
if not possible in time throw exception (send escape message via QMHSNDPM
store locks in some array
}

releaseAll(){
free all locks
}

If you need more than one lock (we needed this for some sophisticated schedules) be carefull with deadlocks. It's easier to avoid deadlocks, than handling. simply define an order to get more than one: they have to be sorted by alphabet. e.g. if someone has a lock to 'HUGO', he could get 'MARTHA'. If he tries to get 'ADAM' throw an exception.
We used a special lockfile and inserted a record with name as one keyfield and held the lock using commitment controll to release all we specified commit (so this must run in a named ACTGRP only used for this SRVPGM. Advantage of this is, that you could see the sequence of aquired locks in journal (if you use a FlipFlop mechanism := an int column created with contents 1 multiplied by -1 for eych update).
If you need more flexibility, to release only a single lock, still holding others (releaseMutex(name)) , some lightweight syncobjects are better.

The QUSLOBJ and the write some info to a DTAARA/FILE I wouldn't recommend. The first is not correct, if more than one is trying to run at the same time, maybe none of them will win, worst case blocking each other. The second is incorrect too, if a job dies you would have to do some manual repair before it could run next time.

Dieter