-
Notifications
You must be signed in to change notification settings - Fork 4
/
hackathonFAQ.html
180 lines (162 loc) · 10.1 KB
/
hackathonFAQ.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
<!DOCTYPE html>
<html class="no-js" lang="en">
<head>
<title>Brain Image Library: Analysis Ecosystem</title>
<meta http-equiv="x-ua-compatible" content="ie=edge" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Raleway:400,800" />
<link rel='stylesheet' href="https://maxcdn.bootstrapcdn.com/font-awesome/4.7.0/css/font-awesome.min.css" />
<link rel="stylesheet" href="/css/bootstrap.css" />
<link rel="stylesheet" href="/css/styles.css" />
<link rel="stylesheet" href="/my.css" />
<script src="/js/jquery.min.js"></script>
</head>
<body class="bg">
<div class="container">
<p> </p>
</div>
<div class="container">
<!--Navigation bar-->
<div id="topmenu">
</div>
<script>
$(function(){
$("#topmenu").load("menu.html");
});
</script>
<!--end of Navigation bar-->
<div class="jumbotron">
<H1>Get Your Brain Straight Hackathon</h1>
This page describes site-specific information about the Brain Image Library that hackathon
participants may find useful. If your question is NOT answered here please check to see if
it is answered on the <a href="https://insightsoftwareconsortium.github.io/GetYourBrainStraight/HCK01_2022_Virtual/">
main hackathon page</a>.
<h2>BIL Helpdesk</h2>
The Brain Image Library has a helpdesk, that you should feel free to contact with any problems/issues/or questions.
To submit a troubleticket, send email to: <i>bil-support@psc.edu</i> . It would be helpful to include that you are
a hackathon participant in the email message and include the word "Hackathon" in the subject line.
<h2>Username/Password</H2>
If you requested an account, you should have received an email from grants@psc.edu with your username and
instructions on how to <a href="https://www.brainimagelibrary.org/password.html">change your password</a>.
<i>All BIL systems use the same username and password.</i>
If you can not find the email or do not know what your username is, please email bil-support@psc.edu for assistance.
<H2>Hackathon Data</H2>
The BIL Filesystem, which is a lustre based filesystem, is mounted on all BIL systems at: /bil .
The hackathon data itself can be found at: /bil/proj/hackathon/2022_GYBS/ Please read the README
file in this directory for instructions on where to find the data and where to place your results.
<H2>BIL Hackathon Documentation</H2>
We highly recommend that participants <a href="https://hackmd.io/@biomed-apps/B1B8mQCb5">review the BIL
hackathon documentation</a>. The BIL Hackathon documentation goes over several items to be aware of using
the BIL analysis ecosystem, including how to connect using <a href="software/X2Go.html">X2Go</a>
and connecting through a terminal program using ssh. If you haven't used them before, LMOD and SLURM
are two <i>important</i>items to review and understand
from the tutorial:
<ul>
<li>SLURM is an open source, fault-tolerant, and highly scalable cluster management and job
scheduling system.</li>
<li>LMOD is software to manage software emvironments. It is used to make software available for use.</li>
</ul>
<p> <i> Note 1: There are a few sample SLURM scripts(aka jobs) that use LMOD Modules that you can examine and
submit on the BIL analysis ecosystem at: /bil/proj/hackathon/2022_GYBS/BIL/slurm</i>
<br /></p>
<p><i>Note 2: The BIL SLURM system has been 100% reserved for the hackathon. You will need to specify the reservation and
the hackathon charge account for your SLURM submissions. (e.g.1 interact -A tra220018p --reservation hackathon)
(e.g.2 #SBATCH -A tra220018p --reservation hackathon) <b>If you have trouble using SLURM and/or interact, please contact us at bil-support@psc.edu</b></i>
<br /></p>
<p>The hackathon documentation also goes into detail outlining several ways to containerize codes and also
gives examples on how to run select software packages installed on the BIL system including MATLAB.
</p>
<h2>Analysis Ecosystem</H2>
<p>The Brain Image Library analysis ecosystem provides an integrated computational and visualization system to
explore, visualize, and and access data without having to download it. The underlying operating system runs Unix (CentOS).
It consists of several parts including:</p>
<ul>
<li><b>BIL Login Node:</b> <i>login.brainimagelibrary.org</i> The BIL login node provides command line, secure
shell (ssh) terminal access to BIL data. The login node is
not intended for heavy computation or visualization, but rather to provide convenient shell access to
BIL data and enable interactive and batch use of the <u>BIL Computational Cluster</u>. </li>
<li><b>BIL Workshop Node:</b> <i>workshop.brainimagelibrary.org</i> The BIL Workshop node is a 56 multi-core VM
node with about 2.6Tb of memory. It is intended to run desktop-visualization applications. It also acts as a
front-end running <i>SLURM and interact</i> to enable interactive and batch use of the
<u>BIL Computational Cluster</u>. Some notes about using the workshop node:
<ul>
<li>Any number of users can ssh in for command-line access</li>
<li>To run graphical applications, we recommend that you
connect using <a href="software/X2Go.html">X2Go</a> . Any number of users can use X2Go
to get a remote desktop. Many
graphical applications will work fine in X2Go (e.g. Matlab, Fiji, vaa3d) and
some applications that require advanced OpenGL features will not
(e.g. NVidia sample CUDA applications such as volumeRender, marchingCubes,
and Mandelbrot; VisRTX interactive viewer)</li>
<li>A single user can use TGX Receiver (available in /bil/packages/TGX/2.2/)
to get a remote desktop with GPU capabilities. All graphical
applications should work fine in this except for ones requiring
the most recent NVidia driver (e.g. VisRTX interactive viewer) </li>
<li>Multiple processes can make use of the GPU at the same time</li>
<li>Please note that if a user has an X2Go session open, they should
not try to start a TGX session and vice versa. This will
cause your logins to get wedged, and it is a hassle to clean
this up.</li>
</ul>
<li><b>BIL Workshop GPU Node:</b> <i>workshop2.brainimagelibrary.org</i> The BIL Workshop GPU node is a
multi-core (72 physical, 144 virtual cores) machine containing two V100
(32Gb) GPUS and about 1.5Tb of memory. Some notes about using the workshop node:
<ul>
<li>Any number of users can ssh in for command-line access</li>
<li>Any number of users can use <a href="software/X2Go.html">X2Go</a> to get a remote desktop.
Many graphical applications will work fine in this (e.g. Matlab,
Fiji, vaa3d) and some applications that require advanced OpenGL
features will not (e.g. NVidia sample CUDA applications such
as volumeRender, marchingCubes, and Mandelbrot; VisRTX
interactive viewer).</li>
<li>A single user can use TGX Receiver (available in /bil/packages/TGX/2.2/)
to get a remote desktop with GPU capabilities, all graphical
applications should work fine in this (including the VisRTX
interactive viewer)</li>
<li>multiple processes can make use of the GPUs at the same time</li>
<li>Please note that if a user has an X2Go session open, they should
not try to start a TGX session and vice versa. This will
cause your logins to get wedged, and it is a hassle to clean
this up.</li>
</ul>
To run graphical applications that use the GPUs, we
recommend that you connect using <a href="software/X2Go.html">X2Go</a> The workshop GPU node can
also be connected to using secure shell (ssh) for users that do not need to run graphical applications.
<li><b>BIL Computational Cluster:</b> The BIL Computational cluster consists of eight nodes
with 3 terabytes of RAM each. The computational cluster runs SLURM and is avaiable for
both batch and interactive access from the <u>BIL login node</u> and the <u>BIL Workshop Node</u>.
SLURM jobs can request a maximum of 75 cores per node</li>
<li><b>Other Systems:</b> The BIL analysis ecosystem includes other machines that we will
not be using for the hackathon including personalized VMs, the Bridges2 HPC system, and Neocortex a
Cerebras advanced AI computing system. If you are interested in these systems, please see
<a href="https://www.brainimagelibrary.org/computevisual.html"> this page</a> or contact
bil-support@psc.edu
<li><b>Data Transfer Nodes:</b> <i>upload.brainimagelibrary.org</i> and <i>download.brainimagelibrary.org</i>
These dedicated data transfer nodes can be used to move files to and from the BIL filesystem.
There are many supported ways to upload and download files into your /bil home directory from
the data transfer nodes, including rsync, Globus, sftp, and scp. </li>
</ul>
<H2>MATLAB Software</H2>
<p> If you desire to use MATLAB on the BIL analysis ecosystem please be sure to
<a href="https://www.psc.edu/resources/software/matlab/permission-form">fill out this form</a> prior to the hackathon if possible.
</p>
<H2>Hackathon Tutorial</H2>
<p> The Hackathon tutorial presented by Ivan can be <a href="https://hackmd.io/@biomed-apps/ryuab8M79#/">found here</a>
</p>
<p> </p>
</div>
<!--Bottom Footer -->
<div id="bottomfooter">
</div>
<script>
$(function(){
$("#bottomfooter").load("footer.html");
});
</script>
<!--end of bottomfooter-->
<script src="/js/jquery.min.js"></script>
<script src="/js/popper.min.js"></script>
<script src="/js/bootstrap.min.js"></script>
</div>
</body>
</html>