[torqueusers] Post-job queueing process

[Prakash Velayutham]

Hi,

I have several questions regarding the process that takes place after a 
job is successfully queued into the server's svr_newjobs list (after the 
entire req_quejob, req_rdytocommit and req_commit calls from the client).

1) What actually triggers the scheduler (assuming the default 
pbs_scheduler) to schedule this job? From the following code bit from 
main() routine in pbs_sched.c, I think scheduler just waits on its 
socket to get a connection request to process a request. Am I correct?

if (select(FD_SETSIZE, &fdset, NULL, NULL, NULL) == -1) {
       if (errno != EINTR) {

2) What is the alarm_time in the scheduler for?

3) The find_best_node() routine tries to find nodes for a job only when 
the job itself does not request nodes resource. In this case when the 
scheduler does a pbs_runjob request to the server, it does not specify a 
node to run job on? So how would the scheduler know whether there are 
enough nodes available in the system?

I have some more questions, but the answers to the above would probably 
answer those too.

Thanks,
Prakash