Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision		Previous revision
en:services:application_services:high_performance_computing:file_systems [2022/06/08 09:12] – created en:services:application_services:high_performance_computing:file_systems [2022/06/08 10:07] (current) – [scratch1 SSD Pool]
Line 1: Line 1:
 +====== File Systems ======
 +**This page is still being worked on!**
  
 +We have a number of different file systems with different characteristics. This page aims to give an overview of these, their usage and availability.
 +
 +===== Overview over File systems =====
 +
 +==== Home ====
 +Your home directory is available everywhere, permanent, and comes with backup. Your disk space is limited by your quota, but can be increased. It is comparably slow, however. Please do not use it for I/O intensive workloads. You can find more information about the $HOME file system [[en:services:storage_services:file_service:fileservice_unix:start|here]].
 +
 +==== scratch1 ====
 +This is the shared scratch space, available on ''amp'', ''agq'' and ''agt'' nodes and frontend ''login-mdc.hpc.gwdg.de''. You can use ''-C scratch'' to make sure to get a node with access to shared /scratch. It is very fast, there is currently no automatic file deletion, but also no backup! We may have to delete files manually when we run out of space. You will receive a warning before this happens.\\
 +To copy data there, you can use the machine transfer-mdc.hpc.gwdg.de, but have a look at [[en:services:application_services:high_performance_computing:transfer_data|Transfer Data]] first.
 +
 +==== scratch2 ====
 +This space is the same as scratch described above except it is **ONLY** available on the nodes ''dfa, dge, dmp, dsu'' and ''dte'' and on the frontend ''login-fas.hpc.gwdg.de''. You can use ''-C scratch2'' to make sure to get a node with access to that space.\\
 +To copy data there, you can use the machine transfer-fas.hpc.gwdg.de, but have a look at [[en:services:application_services:high_performance_computing:transfer_data|Transfer Data]] first.
 +
 +==== local ====
 +This is the local hard disk of the node. It is a fast, SSD based option for storing temporary data. Currently, all our nodes have local SSD storage.\\
 +A directory is automatically created for each job at ''/local/jobs/<jobID>'' and the path is exported as the environment variable ''$TMP_LOCAL''.\\
 +Any date saved on the local disk will be removed at the end of the job.
 +
 +==== scratch1 SSD Pool ====
 +The scratch1 system has directory backed by a very fast SSD pool.
 +===== Temporary directories =====
 +There are two different temporary directories created in your jobs for different use cases.
 +
 +''$TMP_LOCAL'' points to the local SSD. This should be used for temporary data, that does not need to be shared across nodes.
 +
 +''$TMP_SCRATCH'' points to a temporary directory on scratch (even SSD-backed on scratch1). This should be used for temporary data, that have to be shared between nodes of a job, or which size exceed the size of our local SSDs.
 +
 +Please keep in mind, that these directories will automatically be deleted at the end of the job. If you want to keep data written to this directory, make sure to copy it at the end of the job. Using [[en:services:application_services:high_performance_computing:running_jobs_slurm:signals|job signals]] can also be helpful to prevent accidental data loss.
 +
 +===== Scratch Cleanup =====
 +To keep our scratch file systems usable, we will soon start a regular clean up. Our current plan is, to delete top-level directories in your scratch directories, that have not been accessed for 30 days.
 +
 +You will receive sufficient warning before any files are deleted.