Re: [Darshan-users] Logs not generated inspite throwing warning message
Ah, Ok. This is the problem then. The log directory has to be on a shared file system (e.g. nfs, lustre, etc.) that is visible to all nodes. The log is created and written using collective I/O operations, so all of the ranks need to be able to access the same file simultaneously. I'll open a ticket to investigate if there is a way to produce a more helpful error message in this case. thanks, -Phil On 6/21/21 5:50 PM, Hassan Asghar wrote:
it is not shared directory, but i have install the same version of darshan on the other nodes as well.
-- Regards, *Hassan Asghar* MS Computer Science Mobile: +92-321-1741755 Personal Website: http://bit.ly/2OQ0egs <http://bit.ly/2OQ0egs> Google Scholar: https://bit.ly/3lVYuxF <https://bit.ly/3lVYuxF>
On Mon, Jun 21, 2021 at 10:30 PM Phil Carns <[email protected] <mailto:[email protected]>> wrote:
Hi Hassan,
Is /darshan-logs/ a shared directory?
thanks,
-Phil
On 6/21/21 1:04 PM, Hassan Asghar wrote:
Yes, I have set the log path and all permission are given when i run program on local machine it execute successfully but when i exceute on three nodes it show error.
On Tue, 22 Jun 2021, 1:18 am Harms, Kevin, <[email protected] <mailto:[email protected]>> wrote:
Hassan,
did you create the /darshan-logs directory? Looks like it might not be writeable.
try: /darshan-logs/2021/6/21
if you want to try a quick test, set DARSHAN_LOGFILE to the path/name of the log, like:
DARSHAN_LOGFILE=/path/I/can/write/to/logfile.darshan
kevin
________________________________________ From: Darshan-users <[email protected] <mailto:[email protected]>> on behalf of Hassan Asghar <[email protected] <mailto:[email protected]>> Sent: Monday, June 21, 2021 12:35 AM To: [email protected] <mailto:[email protected]> Subject: [Darshan-users] Logs not generated inspite throwing warning message
I am running darshan with mpich on a three-node cluster and my job executed successfully but not generated logs. error warning is given below at the end of the execution.
[haxxanasghar@gpuserver2 ~]$ mpiexec -n 64 -f machinefile -env LD_PRELOAD /usr/l ocal/lib/libdarshan.so ./MADbench2 640 80 4 8 8 4 4
MADbench 2.0 IO-mode no_pe = 64 no_pix = 640 no_bin = 80 no_gang = 4 sblocksize = 8 fblocksize = 8 r_mod = 4 w_mod = 4 IOMETHOD = POSIX IOMODE = SYNC FILETYPE = UNIQUE REMAP = CUSTOM
S_cc 0.02 [ 0.02: 0.03] S_bw 0.01 [ 0.01: 0.01] S_w 0.03 [ 0.03: 0.04] ------- S_total 0.07 [ 0.07: 0.07]
W_cc 0.04 [ 0.03: 0.05] W_bw 0.22 [ 0.20: 0.23] W_r 0.02 [ 0.02: 0.02] W_w 0.01 [ 0.01: 0.01] ------- W_total 0.28 [ 0.28: 0.29]
C_cc 0.01 [ 0.01: 0.01] C_bw 0.00 [ 0.00: 0.00] C_r 0.01 [ 0.01: 0.01] ------- C_total 0.02 [ 0.02: 0.02]
dC[0] = 0.00000e+00 darshan_library_warning: unable to create log file /darshan-logs/2021/6/21/haxxa nas_MADbench2_id29484-29484_6-21-52165-1341319014666854881.darshan_partial. [haxxanasghar@gpuserver2 ~]$
-- Regards,
Hassan Asghar MS Computer Science Mobile: +92-321-1741755 Personal Website: http://bit.ly/2OQ0egs <http://bit.ly/2OQ0egs> Google Scholar: https://bit.ly/3lVYuxF <https://bit.ly/3lVYuxF>
_______________________________________________ Darshan-users mailing list [email protected] <mailto:[email protected]> https://lists.mcs.anl.gov/mailman/listinfo/darshan-users <https://lists.mcs.anl.gov/mailman/listinfo/darshan-users>
_______________________________________________ Darshan-users mailing list [email protected] <mailto:[email protected]> https://lists.mcs.anl.gov/mailman/listinfo/darshan-users <https://lists.mcs.anl.gov/mailman/listinfo/darshan-users>
Ok, thank you so much. I try nfs. On Tue, 22 Jun 2021, 2:58 am Phil Carns, <[email protected]> wrote:
Ah, Ok. This is the problem then. The log directory has to be on a shared file system (e.g. nfs, lustre, etc.) that is visible to all nodes. The log is created and written using collective I/O operations, so all of the ranks need to be able to access the same file simultaneously.
I'll open a ticket to investigate if there is a way to produce a more helpful error message in this case.
thanks,
-Phil On 6/21/21 5:50 PM, Hassan Asghar wrote:
it is not shared directory, but i have install the same version of darshan on the other nodes as well.
-- Regards,
*Hassan Asghar* MS Computer Science Mobile: +92-321-1741755 Personal Website: http://bit.ly/2OQ0egs Google Scholar: https://bit.ly/3lVYuxF
On Mon, Jun 21, 2021 at 10:30 PM Phil Carns <[email protected]> wrote:
Hi Hassan,
Is /darshan-logs/ a shared directory?
thanks,
-Phil On 6/21/21 1:04 PM, Hassan Asghar wrote:
Yes, I have set the log path and all permission are given when i run program on local machine it execute successfully but when i exceute on three nodes it show error.
On Tue, 22 Jun 2021, 1:18 am Harms, Kevin, <[email protected]> wrote:
Hassan,
did you create the /darshan-logs directory? Looks like it might not be writeable.
try: /darshan-logs/2021/6/21
if you want to try a quick test, set DARSHAN_LOGFILE to the path/name of the log, like:
DARSHAN_LOGFILE=/path/I/can/write/to/logfile.darshan
kevin
________________________________________ From: Darshan-users <[email protected]> on behalf of Hassan Asghar <[email protected]> Sent: Monday, June 21, 2021 12:35 AM To: [email protected] Subject: [Darshan-users] Logs not generated inspite throwing warning message
I am running darshan with mpich on a three-node cluster and my job executed successfully but not generated logs. error warning is given below at the end of the execution.
[haxxanasghar@gpuserver2 ~]$ mpiexec -n 64 -f machinefile -env LD_PRELOAD /usr/l
ocal/lib/libdarshan.so ./MADbench2 640 80 4 8 8 4 4
MADbench 2.0 IO-mode no_pe = 64 no_pix = 640 no_bin = 80 no_gang = 4 sblocksize = 8 fblocksize =
8 r_mod = 4 w_mod = 4 IOMETHOD = POSIX IOMODE = SYNC FILETYPE = UNIQUE REMAP = CUSTOM
S_cc 0.02 [ 0.02: 0.03] S_bw 0.01 [ 0.01: 0.01] S_w 0.03 [ 0.03: 0.04] ------- S_total 0.07 [ 0.07: 0.07]
W_cc 0.04 [ 0.03: 0.05] W_bw 0.22 [ 0.20: 0.23] W_r 0.02 [ 0.02: 0.02] W_w 0.01 [ 0.01: 0.01] ------- W_total 0.28 [ 0.28: 0.29]
C_cc 0.01 [ 0.01: 0.01] C_bw 0.00 [ 0.00: 0.00] C_r 0.01 [ 0.01: 0.01] ------- C_total 0.02 [ 0.02: 0.02]
dC[0] = 0.00000e+00 darshan_library_warning: unable to create log file /darshan-logs/2021/6/21/haxxa
nas_MADbench2_id29484-29484_6-21-52165-1341319014666854881.darshan_partial. [haxxanasghar@gpuserver2 ~]$
-- Regards,
Hassan Asghar MS Computer Science Mobile: +92-321-1741755 Personal Website: http://bit.ly/2OQ0egs Google Scholar: https://bit.ly/3lVYuxF
_______________________________________________ Darshan-users mailing [email protected]https://lists.mcs.anl.gov/mailman/listinfo/darshan-users
_______________________________________________ Darshan-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/darshan-users
participants (2)
-
Hassan Asghar -
Phil Carns