Serge's Services: Difference between revisions
(adding some guidelines) |
No edit summary |
||
| (12 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
== These are the services that I'm running as of September 2nd, 2026. == | == These are the services that I'm running as of September 2nd, 2026. == | ||
* qsar - cluster2(epyc: | * qsar - cluster2(epyc:3000). http://10.20.1.42:3000 | ||
* datagrok - cluster2(epyc: | * datagrok - cluster2(epyc-a40:8090). | ||
* cdk/depict cluster2(abacus:8180). http://10.20.0.5:8180/ | |||
* jupyter-hub cluster7(https://cpu02:8888). <a target=_blank href="https://10.70.53.183:8888">cpu02</a> | |||
== How to start a service manually == | == How to start a service manually == | ||
Do this if you can't perform it through gitlab automation tools. | Do this if you can't perform it through gitlab automation tools or if the tools don't exist yet. | ||
These steps may require sudo or root permissions as it uses docker!! | |||
=== Jupyter-Hub === | === Jupyter-Hub === | ||
1. Log into cluster 7 (login01) and SSH into a cpu node: | |||
<syntaxhighlight lang="bash"> | |||
ssh <username>@169.230.26.173 | |||
ssh cpu02 | |||
</syntaxhighlight> | |||
You can use a different node than cpu02. Discover nodes by using sinfo. | |||
2. Create a screen so that the software works when you log off. | |||
<syntaxhighlight lang="bash"> | |||
screen -S jupyter-hub | |||
</syntaxhighlight> | |||
3. Navigate to the following folder while in the screen (you should automatically be in it): | |||
<syntaxhighlight lang="bash"> | |||
/nfs/home/sstrizhov/jupyter-hub | |||
</syntaxhighlight> | |||
4. Run the following script (may need sudo): | |||
<syntaxhighlight lang="bash"> | |||
./startup.sh | |||
</syntaxhighlight> | |||
This should automatically build and get the docker container running for you at "https://cpu02:8888" unless 8888 as a socket is already in use. | |||
=== cdk/depict === | === cdk/depict === | ||
1. Log into cluster 2. | |||
2. Create a screen so that the software works when you log off. | |||
<syntaxhighlight lang="bash"> | |||
screen -S cdkdepict | |||
</syntaxhighlight> | |||
3. Navigate the following folder: | |||
<syntaxhighlight lang="bash"> | |||
/nfs/home/sstrizhov/code/depict/docker | |||
</syntaxhighlight> | |||
4. Run the following script (may need sudo): | |||
<syntaxhighlight lang="bash"> | |||
./startup.sh | |||
</syntaxhighlight> | |||
=== Qsar === | |||
Fresh installation guide for cluster 7: https://wiki.docking.org/index.php?title=Qsar | |||
[[Category:Serge]] | [[Category:Serge]] | ||
Latest revision as of 21:08, 11 September 2026
These are the services that I'm running as of September 2nd, 2026.
- qsar - cluster2(epyc:3000). http://10.20.1.42:3000
- datagrok - cluster2(epyc-a40:8090).
- cdk/depict cluster2(abacus:8180). http://10.20.0.5:8180/
- jupyter-hub cluster7(https://cpu02:8888). <a target=_blank href="https://10.70.53.183:8888">cpu02</a>
How to start a service manually
Do this if you can't perform it through gitlab automation tools or if the tools don't exist yet.
These steps may require sudo or root permissions as it uses docker!!
Jupyter-Hub
1. Log into cluster 7 (login01) and SSH into a cpu node:
ssh <username>@169.230.26.173
ssh cpu02
You can use a different node than cpu02. Discover nodes by using sinfo.
2. Create a screen so that the software works when you log off.
screen -S jupyter-hub
3. Navigate to the following folder while in the screen (you should automatically be in it):
/nfs/home/sstrizhov/jupyter-hub
4. Run the following script (may need sudo):
./startup.sh
This should automatically build and get the docker container running for you at "https://cpu02:8888" unless 8888 as a socket is already in use.
cdk/depict
1. Log into cluster 2.
2. Create a screen so that the software works when you log off.
screen -S cdkdepict
3. Navigate the following folder:
/nfs/home/sstrizhov/code/depict/docker
4. Run the following script (may need sudo):
./startup.sh
Qsar
Fresh installation guide for cluster 7: https://wiki.docking.org/index.php?title=Qsar