On Wed, 3 Nov 2010, srimugunthan dhandapani wrote:
> Hi all,
> I am using kernel client based Ceph version 0.21.1 on linux kernel
> 2.6.35. I am using a six node cluster with 3 machines and 3 virtual
> machines.
> I wrote a simple program to continuously write to a file. After a few
> iterations, the test gets stuck. Sometimes i am unable to unmount the
> ceph mountpt subsequently.
Can you cat /sys/kernel/debug/ceph/*/mdsc and
/sys/kernel/debug/ceph/*/osdc? That will tell us whether it's an mds or
osd request that is hanging. If /sys/kernel/debug isn't there you
may need to 'mount -t debugfs none /sys/kernel/debug'.
Is there a reason you're using the versions you are? Are those the
versions in ubuntu? (They're a bit out of date.)
Thanks-
sage
> I am using the following code
> --
> int main()
> {
>
> FILE *fp;
> int i;
> int written_before_stat;
> int ret_value;
>
>
> /* some junk buffer*/
> char buffer[1024* SIZEOF_ONE_NODE];
>
> char tbuffer[30];
> struct timeval tv;
>
> time_t curtime;
>
> gettimeofday(&tv, NULL);
> curtime=tv.tv_sec;
>
> strftime(tbuffer,30,"%m-%d-%Y %T.",localtime(&curtime));
> printf("%s%ld\t|",tbuffer,tv.tv_usec);
>
>
> printf("starting....\n");
> printf("infinite loop. Ctrl+C when you wish !!\n");
>
> /* keep writing to the ceph file*/
> /* infinite loop. Ctrl+C when you wish*/
> while(1)
> {
>
> /* the frenetic writing phase */
>
> fp = fopen("/ceph/test_file1","w");
> if(fp == NULL)
> {
> printf("ceph dir write file open err\n");
>
> exit(-1);
> }
> for(i = 0; i < 1024 ;i++)
> fwrite(buffer, sizeof(buffer[0]),
> sizeof(buffer)/sizeof(buffer[0]), fp);
>
> fclose(fp);
> gettimeofday(&tv, NULL);
> curtime=tv.tv_sec;
>
> strftime(tbuffer,30,"%m-%d-%Y %T.",localtime(&curtime));
> printf("%s%ld\t|",tbuffer,tv.tv_usec);
>
> printf("120MB written..\n");
> fflush(stdout);
> usleep(1000);
>
> }
>
>
> }
> --------
> Am i missing something?. Thanks for any help.
>
> Thanks,
> Mugunthan
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>