Scientific Computing

Provision of a cluster system with massively parallel computer systems and machines with particularly large main memory.

Who can use this service

Students no
Individual Employees yes
Central Facilities yes
Administrators no

DESCRIPTION SCIENTIFIC COMPUTING

In order to meet the demand at the university for increased computing power and increased memory requirements, the IT Services offer massively parallel computing systems and computers with large main memory. These computers are available for all employees of the Leibniz University of Hannover free of charge.

  • General description

    In order to meet the demand at the university for increased computing power and increased memory requirements, the IT Services offer massively parallel computing systems and computers with large main memory. These computers are available for all employees of the Leibniz University of Hannover free of charge.

    Besides providing the hardware and systems administration, our HPC consultants offer help with any aspect of your computations. Furthermore, we offer introductory talks on how to efficiently use the cluster and how to get started.

    The service Scientific Computing is always available whenever the computing power at the individual institutes will not suffice any more.

    The service Scientific Computing is closely linked to the service Forschungscluster-Housing, which offers institutes to integrate their hardware into the cluster, thus provisioning the whole software suite that is available on the cluster as well as taking care of the machines. Priority usage rights are granted during working hours to members of the institutes owning the respective machines, and during the night hours, all resources work together on reducing the queue lenght.

  • Scope of services

    The service Scientific Computing offers:

    • Access to the cluster system with currently approx. 350 computing nodes
    • Use of free resources of the service Forschungscluster-Housing (priority use right for the respective owners)
    • Central storage systems (HOME, BIGWORK and PROJECT)
    • Central batch system and software server
    • Installation of special software is generally possible subject to an individual agreement (e.g. system suitability, financing)
    • Telephone and Email support
    • System monitoring
    • Introductory course on basic cluster usage (also on request)
  • Service-specific functional parameters
    • Maximum number of simultaneously running jobs per user: 64
    • Maximum number of simultaneously used CPU cores per user: 768
    • maximum job runtime(walltime): 200 hours
    • maximum number of jobs in the queue per user: 2000
    • Disk quota per user in the HOME area: 10 GB
    • Disk quota per user in the BIGWORK area: 1 TB
    • Disk quota per group in the PROJECT area: 10 TB
    • Access to the login nodes via SSH and graphically possible
    • Access to computing nodes only via the batch system
    • Personal accounts: a maximum of 1 account per user, a maximum of 1 user per account
    • At least one active project in the user information and administration service (BIAS) of the data center per associated institute

PREREQUISITES AND USER ACCOUNTS

  1. To be able to use a cluster system, you must first apply for a project.
  2. You can create usernames and accounts as part of a project (using BIAS).
  3. User accounts give you access to the cluster system.
  • How to apply for a project

    A project is the frame in which you will carry out your work on the cluster. To apply for a project, use the form ORG.BEN 4. In this form, you will need to specify the purpose of the computations you want to carry out on the cluster system as well as the type of work, e.g. a bachelor's or master's thesis or any other project of an institute. By signing the form, you agree to be bound by the terms of use that accompany the application.

    In order to apply for a project, you need to have the formal authority to sign. Students can only get an account while working at an institute.

    ORG.BEN_4.pdf
    PDF, 362 KB
  • Registering and maintaining user names

    In order to register a user name you need to have an active project. The project leader can log in to the BIAS website and create a user name.

    Hint:

    • Each user name should reflect the real name of the user by choosing the suffix accordingly.

LOGIN, ACCESS AND FILE TRANSFERS TO/FROM THE CLUSTER

  • Login on to the cluster

    In order to connect to the computing cluster, log on to:

    login.cluster.uni-hannover.de

    Graphical access is possible using Weblogin or X2Go-Client

  • Access to the cluster system

    In order to create user accounts for the cluster, you first need to register a project at the computing center. In the prerequisites section you can find instructions on which forms you need to fill out and how to then create user names. The hashes of the host key fingerprints of login.cluster.uni-hannover.de are as follows:

    • MD5: bf:aa:71:bd:d1:19:d4:4f:b5:60:e7:cb:c4:26:85:a7 (ECDSA)
      SHA256: cYIZZQC96J5bhannZnH2cGQLIBPLQVy29HNc6/vnyFg (ECDSA)
      SHA1: fb:a1:47:ad:97:99:b3:66:86:06:8d:9d:56:46:61:3a:44:9c:a6:cb (ECDSA)
    • MD5: 3b:0a:17:0f:53:85:40:87:c2:be:ed:65:fb:40:59:8a (ED25519)
      SHA256: sG8ZYabQctyGjxPD7X8K2IBxJIE5xHHZ9mQqjlVcjxo (ED25519)
      SHA1: 51:f7:b9:0a:e3:84:8e:d8:2c:a4:e7:7d:42:14:d3:8b:62:2e:a6:5d (ED25519)
    • MD5: bd:b2:04:f9:4e:36:c1:b6:36:8e:d9:03:cc:5d:75:c4 (RSA)
      SHA256: zkVF8Xyxmm7bOEKTNOlvRKe+nG1oHZDeOtY3Un60grg (RSA)
      SHA1: ed:e3:ed:e6:bf:4f:29:9a:2c:72:92:c1:b4:ff:a6:b9:81:f6:6b:45 RSA)
    • MD5: e3:c7:80:67:68:5b:d9:78:df:50:d7:e1:c5:ae:bf:e7 (DSA)
      SHA256: DuWre1exwDsyTzD4yIMy6Oc2CFBCmx5o+l0LmWjMvUc (DSA)
      SHA1: 85:95:f5:c9:9b:2b:29:82:5f:13:70:b9:2b:43:44:84:22:8f:87:40 (DSA)
  • File transfers to/from the cluster

    For file transfers to/from the compute cluster, we recommended using the dedicated transfer node transfer.cluster.uni-hannover.de. It is equiped with a 10GBs network connection and imposes no limitation on CPU time (as opposed to the login nodes).

    Linux: we suggest using any of the following commands to copy files to the cluster: scp, sftp or the graphical tool filezilla (only  sftp protocol). We would also recommend to use the rsync command, which is able to resume interrupted file transfers and can also transfer only the files that have changed.

    Windows: you can - for example - use WinSCP or FileZilla (only scp or sftp protocols) to transfer files from your Windows system to the Cluster.

GETTING STARTED WITH THE CLUSTER SYSTEM

In order to facilitate your start with the cluster system, we provide an introductory talk, a handbook and a cheatsheet. May we suggest to begin with the video to get a first overview. Please continue with the cluster system handbook before starting your work — or at least have a look. Most of the frequently asked questions are already answered there.

BASIC COURSE AND REPORTING PROBLEMS

  • Cluster system introductory course

    All members of the Leibniz Universität Hannover can join a basic introductory course (free of charge) for the LUIS cluster system, which is offered every second month. During the course, you will be learning about the capabilities of the system, how to set up batch jobs, work with software modules and various file systems as well as how to transfer files to and from the cluster system. Preliminary experience in working with computing clusters is not required, although some familiarity with Linux and/or the command line will help a great deal.

    The course is offered every second month and takes about 2 hours.

    All courses have been postponed!

    If you are interested, please send us an email at cluster-help@luis.uni-hannover.de and show up that day. We do not send out confirmations. Should a course have to be cancelled, this will be announced on the Cluster-News mailing list.

    Meeting point: Foyer at LUIS, Schloßwender Straße 5

    In case you arrive late, we will be in room C009 in building 1210/1211.
    You should have a working cluster account and know your password.

    Talk_cluster_basics.pdf
    PDF, 421 KB
  • Reporting problems on the cluster system

    First of all, check your batch script and your program for possible errors.  It is also a good idea to have a look at the batch system documentation.

    The email adress cluster-help@luis.uni-hannover.de is being monitored by several consultants.

    Whenever you have a question about a job or the cluster system, you should provide the following information:

    • Your username (nh...)
    • Job ID(s)
    • A description of the problem
    • Any error output from the program, if available
    • The batch script (as an attachment) and the directory on the cluster
    • The job output report e.g.: myjob.o12345 (as an attachment)

CONTACT

Cluster-Team
Monday to Friday 8:00 a.m. – 7:00 p.m.
Office hours
Due to the corona pandemic the IT service desk is currently only available from 8 a.m. to 5 p.m. (via email)
Cluster-Team
Monday to Friday 8:00 a.m. – 7:00 p.m.
Office hours
Due to the corona pandemic the IT service desk is currently only available from 8 a.m. to 5 p.m. (via email)
Cluster, Nodes, CPUs, Cores/Node ...

RESEARCH CLUSTER HOUSING

Scientific Computing is closely linked to the Research Cluster Housing service, with which we offer institutes the opportunity to integrate and operate their hardware in the cluster. The resources of Research Cluster Housing are an important part of the cluster system, but the owners of the machines have priority in their use.