Skip to content

Commit cd5878a

Browse files
committed
fix
Signed-off-by: oliver könig <okoenig@nvidia.com>
1 parent 16f771d commit cd5878a

File tree

1 file changed

+5
-1
lines changed

1 file changed

+5
-1
lines changed

nemo_run/core/execution/dgxcloud.py

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -405,7 +405,11 @@ def fetch_logs(
405405
job_subdir = self.job_dir[len(nemo_run_home) + 1 :] # +1 to remove the initial backslash
406406
self.pvc_job_dir = os.path.join(self.pvc_nemo_run_dir, job_subdir)
407407

408-
cmd.extend([f"{self.pvc_job_dir}/logs/output-{i}.log" for i in range(self.nodes)])
408+
while files := len(glob.glob(f"{self.pvc_job_dir}/logs/output-*.log")) < self.nodes:
409+
logger.info(f"Waiting for {self.nodes - len(files)} log files to be created...")
410+
time.sleep(3)
411+
412+
cmd.extend(files)
409413

410414
logger.info(f"Attempting to stream logs with command: {cmd}")
411415

0 commit comments

Comments
 (0)