max user processes inside jobs too low

[fadedmaple]

The value of max user processes is only 1024 on all pbsnodes.

ulimit -u 
1024

Threads count as processes in *unix, so it is not possible to create more than 1024 pthreads, which is such a low limit on such powerful hardware.

A four core VM on other pulbic cloud.

ulimit -u
95180

[fadedmaple]

Sorry, I got an error the first time I clicked post and it told me flood, I thought it didn't publish successfully and clicked again

[fadedmaple]

#include <pthread.h>
#include <stdio.h>
#include <unistd.h>
#define MAXTHREADS 131072
#define THREADSTACK 2097152

static unsigned long long thread_nr = 0;

pthread_mutex_t mutex_;

void* inc_thread_nr(void* arg) {
    (void*)arg;
    pthread_mutex_lock(&mutex_);
    thread_nr ++;
    pthread_mutex_unlock(&mutex_);

    /* printf("thread_nr = %d\n", thread_nr); */

    sleep(60);
}

int main(int argc, char *argv[])
{
    pthread_t       pid[MAXTHREADS];
    pthread_attr_t  attrs;
    int  err, i;
    int  cnt = 0;

    pthread_attr_init(&attrs);
    pthread_attr_setstacksize(&attrs, THREADSTACK);

    pthread_mutex_init(&mutex_, NULL);

    for (cnt = 0; cnt < MAXTHREADS; cnt++) {

        err = pthread_create(&pid[cnt], &attrs, (void*)inc_thread_nr, NULL);
        if (err != 0)
            break;
    }

    pthread_attr_destroy(&attrs);

    for (i = 0; i < cnt; i++)
        pthread_join(pid[i], NULL);

    pthread_mutex_destroy(&mutex_);

    printf("Maximum number of threads per process is %d (%lld)\n", cnt, thread_nr);
}

This can be verified by a simple tool. Test on the 64 core 256G  icelake node. qsub -l nodes=1:icx:ppn=2

Maximum number of threads per process is 1021 (1021)

[fadedmaple]

No one else has the same need?

[AthiraM_Intel]

Hi,

Thanks for posting in Intel communities.

Could you please let us know which DevCloud (oneAPI/Edge/FPGA) you are using?

Thanks.

[fadedmaple]

I am using oneAPI

[AthiraM_Intel]

Hi,

We are checking on it internally, will get back to you soon with an update.

Thanks

[fadedmaple]

Any update? I think this number should be increased to at least 10240

[AthiraM_Intel]

Hi,

Sorry for the delay. We have contacted the concerned team. Could you please let us know your usage scenario so that we can proceed further.

Thanks

[AthiraM_Intel]

Hi,

We have not heard back from you. Could you please let us know your usage scenario?

Thanks

[AthiraM_Intel]

Hi,

We have not heard back from you. This thread will no longer be monitored by Intel. If you need further assistance, please post a new question.

Thanks