---
layout: default
navsection: userguide
title: "Storing and Retrieving data using Keep"
...

This tutorial introduces you to the Arvados file storage system.


*This tutorial assumes that you are "logged into an Arvados VM instance":{{site.baseurl}}/user/getting_started/ssh-access.html#login, and have a "working environment.":{{site.baseurl}}/user/getting_started/check-environment.html*

The Arvados distributed file system is called *Keep*.  Keep is a content-addressable file system.  This means that files are managed using special unique identifiers derived from the _contents_ of the file, rather than human-assigned file names (specifically, the md5 hash).  This has a number of advantages:
* Files can be stored and replicated across a cluster of servers without requiring a central name server.
* Systematic validation of data integrity by both server and client because the checksum is built into the identifier.
* Minimizes data duplication (two files with the same contents will result in the same identifier, and will not be stored twice.)
* Avoids data race conditions (an identifier always points to the same data.)

h1. Putting Data into Keep

We will start with downloading a freely available VCF file from the "Personal Genome Project (PGP)":http://www.personalgenomes.org subject "hu599905":https://my.personalgenomes.org/profile/hu599905 to a staging directory on the VM, and then add it to Keep.

In the following tutorials, replace <b><code>you</code></b> with your user id.

First, log into the Arvados VM instance and set up the staging area:

notextile. <pre><code>~$ <span class="userinput">mkdir /scratch/<b>you</b></span></code></pre>

Next, download the file:

<notextile>
<pre><code>~$ <span class="userinput">cd /scratch/<b>you</b></span>
/scratch/<b>you</b>$ <span class="userinput">curl -o var-GS000016015-ASM.tsv.bz2 'https://warehouse.personalgenomes.org/warehouse/f815ec01d5d2f11cb12874ab2ed50daa+234+K@ant/var-GS000016015-ASM.tsv.bz2'</span>
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  216M  100  216M    0     0  10.0M      0  0:00:21  0:00:21 --:--:-- 9361k
</code></pre>
</notextile>

{% include 'notebox_begin' %}

If you have your own data, for example @MyData.vcf@, you can use @scp@ or @rsync@ to copy from your local workstation to the shell VM (run this on your local workstation):

notextile. <pre><code>~$ <span class="userinput">scp MyData.vcf <b>you@shell</b>.arvados:/scratch/<b>you</b>/MyData.vcf</span></code></pre>

{% include 'notebox_end' %}

Now use @arv keep put@ to add your VCF data to Keep, then delete the local copy of the file:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep put var-GS000016015-ASM.tsv.bz2</span>
c1bad4b39ca5a924e481008009d94e32+210
/scratch/<b>you</b>$ <span class="userinput">rm var-GS000016015-ASM.tsv.bz2</span>
</code></pre>
</notextile>

The output value @c1bad4b39ca5a924e481008009d94e32+210@ from @arv keep put@ is the Keep locator.  This enables you to access the file you just uploaded, and is explained in the next section.

h2(#dir). Putting a directory

You can also use @arv keep put@ to add an entire directory:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">mkdir tmp</span>
/scratch/<b>you</b>$ <span class="userinput">echo "hello alice" > tmp/alice.txt</span>
/scratch/<b>you</b>$ <span class="userinput">echo "hello bob" > tmp/bob.txt</span>
/scratch/<b>you</b>$ <span class="userinput">echo "hello carol" > tmp/carol.txt</span>
/scratch/<b>you</b>$ <span class="userinput">arv keep put tmp</span>
0M / 0M 100.0% 
887cd41e9c613463eab2f0d885c6dd96+83
</code></pre>
</notextile>

The locator @887cd41e9c613463eab2f0d885c6dd96+83@ represents a collection with multiple files.

h1. Getting Data from Keep

h2. Using Workbench

You may access collections through the "Collections section of Arvados Workbench":https://{{ site.arvados_workbench_host }}/collections located at "https://{{ site.arvados_workbench_host }}/collections":https://{{ site.arvados_workbench_host }}/collections .  You can also access individual collections and individual files within a collection.  Some examples:

* "https://{{ site.arvados_workbench_host }}/collections/c1bad4b39ca5a924e481008009d94e32+210":https://{{ site.arvados_workbench_host }}/collections/c1bad4b39ca5a924e481008009d94e32+210
* "https://{{ site.arvados_workbench_host }}/collections/887cd41e9c613463eab2f0d885c6dd96+83/alice.txt":https://{{ site.arvados_workbench_host }}/collections/887cd41e9c613463eab2f0d885c6dd96+83/alice.txt

h2. Using arv-get

You can view the contents of a collection using @arv keep ls@:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep ls c1bad4b39ca5a924e481008009d94e32+210</span>
var-GS000016015-ASM.tsv.bz2
</code></pre>

<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep ls 887cd41e9c613463eab2f0d885c6dd96+83</span>
alice.txt
bob.txt
carol.txt
</code></pre>
</notextile>

Use @-s@ to print file sizes rounded up to the nearest kilobyte:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep ls -s c1bad4b39ca5a924e481008009d94e32+210</span>
221887 var-GS000016015-ASM.tsv.bz2
</code></pre>
</notextile>

Use @arv keep get@ to download the contents of a collection and place it in the directory specified in the second argument (in this example, @.@ for the current directory):

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep get c1bad4b39ca5a924e481008009d94e32+210/ .</span>
</code></pre>
</notextile>

You can also download indvidual files:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">arv keep get 887cd41e9c613463eab2f0d885c6dd96+83/alice.txt .</span>
</code></pre>
</notextile>

With a local copy of the file, we can do some computation, for example computing the md5 hash of the complete file:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">md5sum var-GS000016015-ASM.tsv.bz2</span>
44b8ae3fde7a8a88d2f7ebd237625b4f  var-GS000016015-ASM.tsv.bz2
</code></pre>
</notextile>

h2. Using arv-mount

Use @arv-mount@ to take advantage of the "File System in User Space / FUSE":http://fuse.sourceforge.net/ feature of the Linux kernel to mount a Keep collection as if it were a regular directory tree.

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">mkdir mnt</span>
/scratch/<b>you</b>$ <span class="userinput">arv-mount --collection c1bad4b39ca5a924e481008009d94e32+210 mnt &</span>
/scratch/<b>you</b>$ <span class="userinput">cd mnt</span>
/scratch/<b>you</b>/mnt$ <span class="userinput">ls</span>
var-GS000016015-ASM.tsv.bz2
/scratch/<b>you</b>/mnt$ <span class="userinput">md5sum var-GS000016015-ASM.tsv.bz2</span>
44b8ae3fde7a8a88d2f7ebd237625b4f  var-GS000016015-ASM.tsv.bz2
/scratch/<b>you</b>/mnt$ <span class="userinput">cd ..</span>
/scratch/<b>you</b>$ <span class="userinput">fusermount -u mnt</span>
</code></pre>
</notextile>

You can also mount the entire Keep namespace in "magic directory" mode:

<notextile>
<pre><code>/scratch/<b>you</b>$ <span class="userinput">mkdir mnt</span>
/scratch/<b>you</b>$ <span class="userinput">arv-mount mnt &</span>
/scratch/<b>you</b>$ <span class="userinput">cd mnt/c1bad4b39ca5a924e481008009d94e32+210</span>
/scratch/<b>you</b>/mnt/c1bad4b39ca5a924e481008009d94e32+210$ <span class="userinput">ls</span>
var-GS000016015-ASM.tsv.bz2
/scratch/<b>you</b>/mnt/c1bad4b39ca5a924e481008009d94e32+210$ <span class="userinput">md5sum var-GS000016015-ASM.tsv.bz2</span>
44b8ae3fde7a8a88d2f7ebd237625b4f  var-GS000016015-ASM.tsv.bz2
/scratch/<b>you</b>/mnt/c1bad4b39ca5a924e481008009d94e32+210$ <span class="userinput">cd ../..</span>
/scratch/<b>you</b>$ <span class="userinput">fusermount -u mnt</span>
</code></pre>
</notextile>

Using @arv-mount@ has several significant benefits:

* You can browse, open and read Keep entries as if they are regular files.
* It is easy for existing tools to access files in Keep.
* Data is downloaded on demand, it is not necessary to download an entire file or collection to start processing

<hr>

You are now ready to proceed to the next tutorial, "running a crunch job.":tutorial-job1.html