Content
1. CINECA HPC system
The CINECA HPC system is a HPC infrastructure based in Italy (Bologna) at CINECA.
To get access to the system you need to do the following:
- Register to the the system at this URL: https://userdb.hpc.cineca.it/user/register
- Once you receive your credentials, connect to the UserDB portal (https://userdb.hpc.cineca.it/user/r)
- Go to the “HPC Access” section (you find the link among the Available Services, in the left menu)
- Complete the three sub-sections:
- – Documents for HPC (upload a document, fill the mandatory fields, and subscribe the agreement on the “User Responsibilities”)
- – Institution, this is already OK (you filled the required fields)
- – Personal Data, this is already OK (you filled the fields and subscribed the agreement
Once finished, in the HPC Access page a Submit button will appear. Click on it to request an HPC access (that is, username and password to log into the clusters) After a short verification of the data you inserted you will receive your credentials in two different e-mails.
It will require approximately two hours from the email receiving for the credentials to be active on the servers
2. CINECA servers
Up to date there exist two main servers: GALILEO and PICO. For Bioinformatics work the suggested system is PICO, which has already installed a wide variety of tools
3. PICO
To access PICO one can use standard ssh command line
ssh username@login.pico.cineca.it
For a detailed description of PICO’s infrastructure see here: PICO userguide
Resources saldo
To have an overview of the present availability of resources one can use this command
saldo -b
that will return both space and CPU hrs saldo.
For DRES filesystem the command is
cindata
Several projects might be active and ready to use. Please refer to Stefano Calza to get an updated overview of which project is better rely on for the present work.
All projects share a common DRES filesystem that can be used for storing data and sharing data between projects.
Data management
Detailed information on data management in CINECA can be found here: Data management
Modules
PICO system, just like UppMax, is based on modules to provide additional features. To inspect available module run
module avail
To load advanced modules run
module load profile/advanced
Inspecting the module list will now show a whole new set of tools. E.g. to load R
module load gnu/4.8.3 module load r/3.2.2
4. PBS
The batch scheduler installed in CINECA machines is similar to UppMax’s one (SLURM) though with some minor differences. To create a batch files and find out all the details please refer to PBS userguide: PBS userguide