Difference between revisions of "Job and File Storage Space"

From Statistics Cluster
Jump to navigation Jump to search
(Created page with "== Disk Space Organization == Below is a list of of different forms of disk space available on the cluster to the user for reading and writing. (Unless noted otherwise, there ...")
 
 
(12 intermediate revisions by the same user not shown)
Line 1: Line 1:
== Disk Space Organization ==
+
== Disk Space Organization Overview ==
 
Below is a list of of different forms of disk space available on the cluster to the user for reading and writing. (Unless noted otherwise, there is no backup and thus no warranty about the security of the data.)
 
Below is a list of of different forms of disk space available on the cluster to the user for reading and writing. (Unless noted otherwise, there is no backup and thus no warranty about the security of the data.)
 
<br>
 
<br>
 
{| class="wikitable" style="text-align: left; background: #A9A9A9"
 
{| class="wikitable" style="text-align: left; background: #A9A9A9"
 
| style="background: #f9f9f9; width: 150px" | /home/<i><b>username</b></i>
 
| style="background: #f9f9f9; width: 150px" | /home/<i><b>username</b></i>
| style="background: #f9f9f9; width: 750px" | user home directory with regular backups; 90 MB default quota.
+
| style="background: #f9f9f9; width: 750px" | user home directory with regular backups; 90 MB default SHARED quota.  
 
|-  
 
|-  
 
| style="background: #f9f9f9; width: 150px" | /home/<i><b>username</b></i>/jobs
 
| style="background: #f9f9f9; width: 150px" | /home/<i><b>username</b></i>/jobs
| style="background: #f9f9f9; width: 750px" | link to local space on stats.phys ideal for launching Condor jobs. No backups, no per-user quota; ~150 GB space total  
+
| style="background: #f9f9f9; width: 750px" | only available on stats.phys.uconn.edu ideal for launching Condor jobs. No backups, no per-user quota; ~150 GB space total  
 
|-
 
|-
 
| style="background: #f9f9f9; width: 150px" | /scratch
 
| style="background: #f9f9f9; width: 150px" | /scratch
| style="background: #f9f9f9; width: 75px" | temporary space shared by <b>all</b> users; ~100 GB total; <b>NO BACKUP, DO NOT PERMANENTLY STORE RESULTS HERE</b>
+
| style="background: #f9f9f9; width: 750px" | temporary space shared by <b>all</b> users; ~100 GB total; <b>NO BACKUP, DO NOT PERMANENTLY STORE RESULTS HERE</b>
|-
 
| style="background: #f9f9f9; width: 150px" | /local
 
| style="background: #f9f9f9; width: 750px" | ~30 GB local space available on every Statistics node that is automatically used for staging by Condor for executables and data.
 
 
|}
 
|}
 +
=== Results ===
 +
If you have files that you will not be using and do not want to delete, please move them to your own personal storage off of the cluster. The jobs directory is NOT backed up so any results that are important should be moved off of the cluster. The cluster is where you stage your work but it is not a recommended permanent home for your results.
 +
== /home/username ==
 +
Upon logging in, the user lands in his or her home directory: /home/<i><b>username</b></i>. This modest amount of space is intended as private space for development, testing of code etc. Results may safely be kept here for some time as this space is backed up. It is not intended for large data files or the many files generated by Condor jobs.
  
Upon logging in, the user lands in his or her home directory: /home/<i><b>username</b></i>. This modest amount of space is intended as private space for development, testing of code etc. Results may safely be kept here for some time as this space is backed up. It is not intended for large data files or the many files generated by Condor jobs.
+
The quota for this directory is shared among all statistics users. If one person fills their home directory it will prevent other users from being able to write to theirs. For large files, please use the jobs directory.
  
 +
== /home/username/jobs ==
 
A special directory /home/<b><i>username</i></b>/jobs is a more efficient space for launching jobs. It is actually a link to local space on stats.phys.uconn.edu (and is not accessible from processing nodes.) It is not bound by quotas.
 
A special directory /home/<b><i>username</i></b>/jobs is a more efficient space for launching jobs. It is actually a link to local space on stats.phys.uconn.edu (and is not accessible from processing nodes.) It is not bound by quotas.
  
 +
<b>This is the recommended location for storing files that are in use.</b>
 +
 +
== /scratch ==
 
For this and other purposes a temporary directory /scratch is also available from all nodes. This is a larger working space suitable for collaborative work and launching of jobs. Big files that need to be transfered to the cluster can be copied directly to the scratch space by specifying /scratch as a destination in your scp client. For example, compression-enabled transfer of a large file directly to a working directory under the cluster's scratch space with the console-based scp client looks as follows:
 
For this and other purposes a temporary directory /scratch is also available from all nodes. This is a larger working space suitable for collaborative work and launching of jobs. Big files that need to be transfered to the cluster can be copied directly to the scratch space by specifying /scratch as a destination in your scp client. For example, compression-enabled transfer of a large file directly to a working directory under the cluster's scratch space with the console-based scp client looks as follows:
  
<i>scp -C WarAndPeace.txt lntolstoy@stats.phys.uconn.edu:/scratch/LeosSpace/novel_parse</i>
+
<i>scp -C WarAndPeace.txt Intolstoy@stats.phys.uconn.edu:/scratch/LeosSpace/novel_parse</i>
 
+
{|class="wikitable" style="text-align: center"
Being a collaborative space, however, means that it should be kept organized for the sake of all other clusters users (including members of the adjoining Physics and Geophysics clusters). This space may be cleaned up by administrators if any files appear to be abandoned.
+
|style="background: #f2f2f2; width: 200px"| scp
 
+
|style="background: #f2f2f2; width: 200px"|  -C
The shared spaces discussed so far reside on a file server and accessible to the stats nodes over the network via nfs. While this is conveneint for the purposes of a shared space for testing and job distribution, network latency, bandwidth limitation and congestion may create a bottleneck for data-intensive calculations. To resolve this problem, local space is available on each node in the form of an on-board hard disk. It is mounted on each node under /local. Note that use of this space requires a job to copy the necessary files and clean up at the end.
+
|style="background: #f2f2f2; width: 200px"| WarAndPeace.txt
 +
|style="background: #f2f2f2; width: 200px"| lntolstoy@stats.phys.uconn.edu:/scratch/LeosSpace/novel_parse
 +
|-
 +
|style="width: 200px"| secure copy command
 +
|style="width: 200px"| specific option for scp
 +
|style="width: 200px"| local file in present directory
 +
|style="width: 200px"| remote file location on stats server
 +
|}

Latest revision as of 18:50, 9 November 2016

Disk Space Organization Overview

Below is a list of of different forms of disk space available on the cluster to the user for reading and writing. (Unless noted otherwise, there is no backup and thus no warranty about the security of the data.)

/home/username user home directory with regular backups; 90 MB default SHARED quota.
/home/username/jobs only available on stats.phys.uconn.edu ideal for launching Condor jobs. No backups, no per-user quota; ~150 GB space total
/scratch temporary space shared by all users; ~100 GB total; NO BACKUP, DO NOT PERMANENTLY STORE RESULTS HERE

Results

If you have files that you will not be using and do not want to delete, please move them to your own personal storage off of the cluster. The jobs directory is NOT backed up so any results that are important should be moved off of the cluster. The cluster is where you stage your work but it is not a recommended permanent home for your results.

/home/username

Upon logging in, the user lands in his or her home directory: /home/username. This modest amount of space is intended as private space for development, testing of code etc. Results may safely be kept here for some time as this space is backed up. It is not intended for large data files or the many files generated by Condor jobs.

The quota for this directory is shared among all statistics users. If one person fills their home directory it will prevent other users from being able to write to theirs. For large files, please use the jobs directory.

/home/username/jobs

A special directory /home/username/jobs is a more efficient space for launching jobs. It is actually a link to local space on stats.phys.uconn.edu (and is not accessible from processing nodes.) It is not bound by quotas.

This is the recommended location for storing files that are in use.

/scratch

For this and other purposes a temporary directory /scratch is also available from all nodes. This is a larger working space suitable for collaborative work and launching of jobs. Big files that need to be transfered to the cluster can be copied directly to the scratch space by specifying /scratch as a destination in your scp client. For example, compression-enabled transfer of a large file directly to a working directory under the cluster's scratch space with the console-based scp client looks as follows:

scp -C WarAndPeace.txt Intolstoy@stats.phys.uconn.edu:/scratch/LeosSpace/novel_parse

scp -C WarAndPeace.txt lntolstoy@stats.phys.uconn.edu:/scratch/LeosSpace/novel_parse
secure copy command specific option for scp local file in present directory remote file location on stats server