oph:cluster:problems
Questa è una vecchia versione del documento!
Known problems
In this section we list a set of known problems that may occur when using the cluster, with possible quick solutions. If you encounter problems not listed here, or if the solutions below do not solve your problem, please direct your issue exclusively to difa.csi@unibo.it
- no write permission to own
$HOME
: rarely happens at first login; just logout and login again - can't share data/programs in
$HOME
with other users: it's usually way better to use another storage to share data, but if you really need it, ask admins to relax permissions on your home - connection refused error but my user have just been given access: you have to wait till after 7.00am of the following day, when a script updates access lists and accounting
- An ORTE daemon has unexpectedly failed after launch error on multi-node mpi jobs: try using srun instead of mpirun to launch the job step
oph/cluster/problems.1691477318.txt.gz · Ultima modifica: 2023/08/08 06:48 da diego.zuccato@unibo.it