If the communication with the cluster fails in checking the queue, then the full job is resubmitted.
This should be avoided, and at least demanded a few times.
To be fixed before pull request:
72
A report of the error:
kex_exchange_identification: read: Connection reset by peer
kex_exchange_identification: Connection closed by remote host
Error with cmd: ssh daint 'squeue -u $USER'
EXITSTATUS: 255; attempt = 1
Exception in thread Thread-10:
Traceback (most recent call last):
File "/home/lmonacelli/anaconda3/lib/python3.9/threading.py", line 973, in _bootstrap_inner
self.run()
File "/home/lmonacelli/anaconda3/lib/python3.9/threading.py", line 910, in run
self._target(*self._args, **self._kwargs)
File "/home/lmonacelli/anaconda3/lib/python3.9/site-packages/sscha/Cluster.py", line 1452, in compute_single_jobarray
subs, indices, labels = self.batch_submission(structures, calc, jobs_id, ".pwi",
File "/home/lmonacelli/anaconda3/lib/python3.9/site-packages/sscha/Cluster.py", line 922, in batch_submission
while not self.check_job_finished(job_id):
File "/home/lmonacelli/anaconda3/lib/python3.9/site-packages/sscha/Cluster.py", line 1000, in check_job_finished
if data[0] == job_id:
IndexError: list index out of range
If the communication with the cluster fails in checking the queue, then the full job is resubmitted. This should be avoided, and at least demanded a few times.
To be fixed before pull request:
72
A report of the error: