final edits

This commit is contained in:
Natalie Elphick 2024-04-12 17:32:41 -07:00
parent b0dd9c589e
commit d482023d75
4 changed files with 189 additions and 174 deletions

View file

@ -585,10 +585,16 @@ document.addEventListener('DOMContentLoaded', function(e) {
</section>
<section id="introductions" class="slide level2">
<h2>Introductions</h2>
<p><strong>Natalie Elphick</strong><br />
Bioinformatician I</p>
<p><strong>Alex Pico (TA)</strong><br />
Bioinformatics Core Director</p>
<p>Instructor:</p>
<p>      <strong>Natalie Elphick</strong><br />
      <em>Bioinformatician I</em></p>
<p>TAs:</p>
<p>      <strong>Alex Pico</strong><br />
      <em>Bioinformatics Core Director</em><br />
      <strong>Ayushi Agrawal</strong><br />
      <em>Bioinformatician III</em><br />
      <strong>Min-Gyoung Shin</strong><br />
      <em>Bioinformatician III</em></p>
</section>
<section id="target-audience" class="slide level2">
<h2>Target Audience</h2>
@ -634,7 +640,7 @@ specifications</a>)<br />
<li>Funded and administered cooperatively by UCSF campus IT and key
research groups</li>
</ul>
<p><a href="https://wynton.ucsf.edu">https://wynton.ucsf.edu</a></p>
<p><a href="https://wynton.ucsf.edu" class="uri">https://wynton.ucsf.edu</a></p>
</section></section>
<section>
<section id="node-types-and-logging-in" class="title-slide slide level1">
@ -781,8 +787,8 @@ requests</li>
<li>Certain input/output patterns can be problematic</li>
</ul>
</section>
<section id="beegfs---io-patterns" class="slide level2 small-bullets">
<h2 class="small-bullets">BeeGFS - I/O patterns</h2>
<section id="beegfs---io-patterns" class="slide level2">
<h2>BeeGFS - I/O patterns</h2>
<ul>
<li>Anything that requires lots of metadata operations can feel slow
<ul>
@ -791,21 +797,18 @@ and directory searches (<strong>conda</strong>)</li>
</ul></li>
<li>Keep the number of reads and writes to a single directory to a
reasonable number</li>
<li>If using conda, putting the conda application inside a Apptainer
(formerly singularity) container will result in better performance</li>
</ul>
</section>
<section id="beegfs---tips" class="slide level2">
<h2>BeeGFS - Tips</h2>
<section id="beegfs---takehome-message" class="slide level2 small-bullets">
<h2 class="small-bullets">BeeGFS - Takehome Message</h2>
<ul>
<li>Prefer fewer, large files over many small ones</li>
<li>Distribute reading and writing over several directories
<ul>
<li>Including compute job output and error files</li>
</ul></li>
<li>Use local scratch (/scratch) when possible</li>
<li>Distribute reading and writing over several directories</li>
<li>Use local scratch (<strong>/scratch</strong>) when possible</li>
<li>Dont include anything in <strong>/wynton</strong> in your default
LD_LIBRARY_PATH</li>
<li>If using conda, putting the conda application inside a Apptainer
(formerly singularity) container will result in better performance</li>
</ul>
</section>
<section id="storage-1" class="slide level2 small-bullets">
@ -872,18 +875,22 @@ contacting Gladstone IT<br />
knowledge base page</a></li>
</ul>
</section>
<section id="storage-advice" class="slide level2">
<h2>Storage Advice</h2>
<section id="storage-advice" class="slide level2 small-bullets">
<h2 class="small-bullets">Storage Advice</h2>
<ul>
<li>Always back up anything you store under
<strong>/wynton</strong></li>
<li>Back up your data on <strong>/gladstone</strong> if you have access
to it
<li>If you have access to it keep all of your data on
<strong>/gladstone</strong>
<ul>
<li>A large number of jobs reading and writing to these directories will
<li>A large number of jobs reading and writing to these directories may
be slower since it is NFS mounted not BeeGFS</li>
</ul></li>
<li>Use the scratch directories to store temporary files</li>
<li>Use the scratch directories to store temporary files
<ul>
<li>e.g. A large amount of .fastq that you do not need after the
alignment step</li>
</ul></li>
</ul>
</section></section>
<section>
@ -906,6 +913,13 @@ be slower since it is NFS mounted not BeeGFS</li>
</ul>
<pre><code>{local}$ scp alice@dt1.wynton.ucsf.edu:/path/to/local_file.tsv /destination/path</code></pre>
</section>
<section id="hands-on" class="slide level2">
<h2>Hands-on</h2>
<ul>
<li>Use scp to copy this <a href="https://www.dropbox.com/scl/fi/463ymz88q89d2co90kj30/candidatus_carsonella_ruddii_complete_genome.fasta?rlkey=9x64iek2yy149sh2i1r9sse9y&amp;dl=1">file</a>
into your home directory on Wynton</li>
</ul>
</section>
<section id="gui-sftp-clients" class="slide level2 small-bullets">
<h2 class="small-bullets">GUI SFTP Clients</h2>
<ul>
@ -955,8 +969,7 @@ UCSF Box</a> instructions</li>
</section>
<section id="poll-1" class="slide level2">
<h2>Poll 1</h2>
<p>Which of these can you <strong>not</strong> log in to from your
computer?</p>
<p>Poll 1 - Which of these can you <strong>not</strong> SSH in to?</p>
<ol type="1">
<li>Login Nodes</li>
<li>Development Nodes</li>
@ -967,7 +980,7 @@ computer?</p>
<section id="poll-2" class="slide level2">
<h2>Poll 2</h2>
<p>The <strong>/wynton</strong> directory is backed up on a nightly
basis so do not need to back up the data you store here.</p>
basis, so there is no need to back up anything stored here.</p>
<ol type="1">
<li>True</li>
<li>False</li>
@ -1136,7 +1149,7 @@ run:</li>
<ul>
<li>Please take some time to fill out the workshop survey if you are not
attending part 2:<br />
<a href="https://www.surveymonkey.com/r/F75J6VZ">https://www.surveymonkey.com/r/F75J6VZ</a></li>
<a href="https://www.surveymonkey.com/r/F75J6VZ" class="uri">https://www.surveymonkey.com/r/F75J6VZ</a></li>
</ul>
</section>
<section id="upcoming-data-science-training-program-workshops" class="slide level2">

View file

@ -585,10 +585,14 @@ document.addEventListener('DOMContentLoaded', function(e) {
</section>
<section id="introductions" class="slide level2">
<h2>Introductions</h2>
<p><strong>Natalie Elphick</strong><br />
Bioinformatician I</p>
<p><strong>Alex Pico (TA)</strong><br />
Bioinformatics Core Director</p>
<p>Instructor:</p>
<p>      <strong>Natalie Elphick</strong><br />
      <em>Bioinformatician I</em></p>
<p>TAs:</p>
<p>      <strong>Alex Pico</strong><br />
      <em>Bioinformatics Core Director</em><br />
      <strong>Michela Traglia</strong><br />
      <em>Senior Statistician</em></p>
</section>
<section id="target-audience" class="slide level2">
<h2>Target Audience</h2>
@ -939,8 +943,8 @@ cancelled</li>
</section>
<section id="poll-3" class="slide level2">
<h2>Poll 3</h2>
<p>Anything that you can run on a compute node can be run on a
development node.</p>
<p>Any submitted job to compute nodes can also be run on development
nodes.</p>
<ol type="1">
<li>True</li>
<li>False</li>
@ -1155,6 +1159,8 @@ into Wynton</li>
</section>
<section id="bioinformatics-questions" class="slide level2">
<h2>Bioinformatics Questions</h2>
<p>For any bioinformatics specific questions feel free to reach out to
the Gladstone Bioinformatics Core.</p>
<ul>
<li>Email
<ul>
@ -1174,8 +1180,11 @@ into Wynton</li>
<section id="thank-you" class="slide level2">
<h2>Thank You!</h2>
<ul>
<li>Please take some time to fill out the workshop survey:<br />
<a href="https://www.surveymonkey.com/r/F75J6VZ">https://www.surveymonkey.com/r/F75J6VZ</a></li>
<li><p>Please take some time to fill out the workshop survey:<br />
<a href="https://www.surveymonkey.com/r/F75J6VZ">https://www.surveymonkey.com/r/F75J6VZ</a></p></li>
<li><p>Want some additional Wynton training?<br />
Check out the UCSF library <a href="https://calendars.library.ucsf.edu/event/12197724">Introduction to
Wynton HPC Cluster</a> Workshop</p></li>
</ul>
</section>
<section id="upcoming-data-science-training-program-workshops" class="slide level2">