diff --git a/NEWS b/NEWS index ac16da3f31a20997e3c413ba4841796d5f7d3710..206e519dbb1a5ef41b4352f690f68eab51397465 100644 --- a/NEWS +++ b/NEWS @@ -7,8 +7,6 @@ documents those changes that are of interest to users and admins. -- Support large (over 2GB) files on 32-bit linux systems -- changed all writes to safe_write in srun -- added $float to globals.example in the testsuite - -- Limit attempted BGL jobs scheduled at once to 100 (too slow with - dynamic scheduling to handle hundreds of jobs at once). -- Set job's num_proc correctly for jobs that do not have exclusive use of it's allocated nodes. -- Change in support for test suite: 'testsuite/expect/globals.example' diff --git a/src/slurmctld/job_scheduler.c b/src/slurmctld/job_scheduler.c index 1477f659b4100fca8980ed26dd146bba920f2606..6cbbb5cba6a979e8a61cf167dbdcc6e64799532f 100644 --- a/src/slurmctld/job_scheduler.c +++ b/src/slurmctld/job_scheduler.c @@ -143,7 +143,7 @@ extern bool job_is_completing(void) int schedule(void) { struct job_queue *job_queue; - int i, error_code, failed_part_cnt, job_queue_size, job_cnt = 0; + int i, j, error_code, failed_part_cnt, job_queue_size, job_cnt = 0; struct job_record *job_ptr; struct part_record **failed_parts; /* Locks: Read config, write job, write node, read partition */ @@ -153,8 +153,6 @@ int schedule(void) uint16_t quarter = (uint16_t) NO_VAL; uint16_t nodecard = (uint16_t) NO_VAL; char tmp_char[256]; -#else - int j; #endif lock_slurmctld(job_write_lock); /* Avoid resource fragmentation if important */ @@ -176,20 +174,13 @@ int schedule(void) job_ptr = job_queue[i].job_ptr; if (job_ptr->priority == 0) /* held */ continue; -#ifdef HAVE_BG - /* Dynamic partitioning on BGL is slow, so we don't - * want to try scheduling too many jobs or the - * slurmctld will stop responding for too long. */ - if (i >= 100) - break; -#else for (j = 0; j < failed_part_cnt; j++) { if (failed_parts[j] == job_ptr->part_ptr) break; } if (j < failed_part_cnt) continue; -#endif + error_code = select_nodes(job_ptr, false); if (error_code == ESLURM_NODES_BUSY) { #ifndef HAVE_BG /* keep trying to schedule jobs in partition */