Scyld ClusterWare HPC: Reference Guide | ||
---|---|---|
<< Previous | Next >> |
beostat_count_idle_cpus will count the number of CPUs in the entire cluster that are available to the current user/group and have CPU usage below a given threshold. Note that an easy way to count the total number of CPUs available to a user independent of the usage is to set the threshold arbitrarily large.
int max, fif; max = beostat_count_idle_cpus (9999.0); fif = beostat_count_idle_cpus (0.5); printf ("%d of the %d CPUs available are busy.\n", (max-fif), max); |
Returns the number of CPUs that are both available for the caller and have usage below the threshold. If an error occurs, it will return -1.
This function relies on the Beostat subsystem, which consists of the proc filesystem on the remote node, the sendstats daemon on the remote node, the recvstats daemon on the master node, and two shared memory files in the directory /var/shm. If any part of the system breaks down, this function could fail.
<< Previous | Home | Next >> |
Scyld ClusterWare Beostat Libraries | Up | beostat_count_idle_cpus_on_node |