https://www.fe.infn.it/coka/ 2026-05-06T12:25:16+00:00 https://www.fe.infn.it/coka/ https://www.fe.infn.it/coka/lib/tpl/dokuwiki/images/favicon.ico text/html 2017-06-06T15:26:32+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:accounting&rev=1496762792&do=diff Requesting an account To request an account, please send an e-mail to <coka-admin@fe.infn.it> with Subject: “Account Request”, using your institutional e-mail address, specifying in the mail body your Name and Surname (plus Name and Surname of your supervisor if applicable), attaching your text/html 2017-09-04T13:20:58+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:hardware&rev=1504531258&do=diff Available Hardware The COKA Cluster is made up of one master node acting as a login front-end and 4 compute nodes named: node01, node02, node03, node04 In the compute nodes hyper-threading is disabled, although it is enabled in the login node. The computing hardware of one node can be summarized by this schema: text/html 2016-04-12T08:27:11+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:interactive&rev=1460449631&do=diff Interactive Session If needed, an interactive shell can be run on a cluster node using the following SURM command (omitting the nodelist argument will let SLURM choose one random free node): srun -u --partition=interactive --nodelist=node02 --pty bash text/html 2016-03-15T18:13:23+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:login&rev=1458065603&do=diff Login to the Cluster In the COKA Cluster we do not use passwords, users authentication is managed by Public-key authentication, thus you need to provide us the public key corresponding to the private key stored in the computer/account you will use to login to the cluster. text/html 2018-02-20T11:35:10+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:partitions&rev=1519126510&do=diff Select a Partition In SLURM, queues are called partitions. A partition can be selected with the --partition or -p argument, to know how to do it in your job submission scripts see: job submission script options. To query the availability of partitions end nodes on the cluster you can use the following SLURM command: text/html 2016-11-07T13:56:52+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:slurm&rev=1478527012&do=diff Submitting a Job with SLURM The workload manager adopted in the COKA cluster is SLURM. The suggested way to submit a job using SLURM is to prepare a Job submission file, also known as job file or submission script. The main purpose of this file is to ask for resources from the job scheduler and to indicate the program to be run. It can be used also to set environment variables, compile your code, etc. text/html 2016-04-01T08:25:06+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:software&rev=1459499106&do=diff Available Software In the COKA Cluster are installed most of the commonly used compilers and libraries. The Environment Modules system is available to manage their usage, in order to select which set of compilers and/or libraries to use. If you need some other software to be installed, you can suggest to add it, writing to: text/html 2016-03-31T09:47:58+00:00 Anonymous (anonymous@undisclosed.example.com) https://www.fe.infn.it/coka/doku.php?id=users:system&rev=1459417678&do=diff Cluster Setup All the cluster nodes are diskless machines, apart from a a single SSD used as a fast/low-latency temporary directory. In the following section a more detailed description of the mounted file-systems will be given for your convenience in order to better exploit the cluster capabilities.